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.
72 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
73 related applications. Previously an "H" was used where available info
74 says that "M" should be, so change to match.
76 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
77 as arguments, so an implementation trying to copy these into a local
78 buffer was taking a taint-enforcement trap. Fix by using dynamically
79 created buffers. Similar fix for radius expansion condition.
81 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
82 reasonable, eg. to count headers. Fix by using dynamically created
83 buffers rather than a local. Do similar fixes for ACL actions "dcc",
84 "log_reject_target", "malware" and "spam"; the arguments are expanded
85 so could be handling tainted values.
87 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
88 broken the (no-op) support for this sendmail command. Restore it
89 to doing nothing, silently, and returning good status.
91 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
92 record path was given (or the default used) without a leading directory
93 path, an error occurred on trying to open it. Use the transport's working
96 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
97 transport. Previously it was the name on the DNS A-record; use instead
98 the head of the CNAME chain leading there (if there is one). This seems
99 to align better with RFC 6125.
101 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
102 smtp_accept_max_per_host allocated resources which were not released
103 when the limit was exceeded. This eventually crashed the daemon. Fix
104 by adding a release action in that path.
106 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
107 expanded; previously using tainted values was rejected. Fix by using
108 dynamically-created buffers.
110 JH/09 Relax restrictions on ACL verify condition needing access to message
111 headers. Previously they were only permitted in data and non-smtp ACLs;
112 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
113 not_blind, header_sender and header_names_ascii verification.
115 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
116 Previously a macro used one argument twice; when called with the
117 argument as an expression having side-effects, incorrect operation
118 resulted. Use an inlineable function.
120 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
121 held open for a verify callout. Previously this wan not accounted for
122 and a corrupt onward SMTP conversation resulted.
124 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
125 Previously they were embedded naively in the constructed address; when
126 needed, strip the quoting and quote the entire local_part.
127 Also make the inbound_srs expansion condition handle quoting.
129 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
130 excluded, not matching the documentation.
132 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
133 was given for the sqlite_dbfile a trap resulted.
135 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
136 "name" argument resulted in a trap. There is no reason to disallow such;
137 this was a coding error.
139 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
140 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
141 spent suspended, ignoring the POSIX definition. Previously we assumed
142 it did and a constant offset from real time could be used as a correction.
143 Change to using the same clock source for the start-of-message and the
144 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
145 exists, just to get a clock slightly more aligned to reality.
147 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
148 RFC says it is optional some validators care. The missing char was not
149 intended but triggered by a line-wrap alignment. Discovery and fix by
150 Guillaume Outters, hacked on by JH.
152 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
153 name being quoted was tainted a trap would be taken. Fix by using
154 dynamically created buffers. The routine could have been called by a
155 rewrite with the "h" flag, by using the "-F" command-line option, or
156 by using a "name=" option on a control=submission ACL modifier.
158 JH/19 SPF: change the Authentication-Results expansion component to give
159 smtp.helo when the sender domain is empty. Previously it gave
162 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
163 not in 127.0.0.0/8 to help in spotting list domains taken over by a
164 domain-parking registrar.
166 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
167 Previously when a whitespace character was specified it was not inserted
168 after removing the newline.
170 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
171 the domain part of the recipient address. This overrides any tls_sni
172 option set, which was previously used.
174 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
177 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
178 is_tainted() had an off-by-one error in the overenthusiastic direction.
179 Find and fix by Gavan. Although NetBSD is not a supported platform for
180 4.94 this bug could affect other platforms.
182 PP/01 Fix default prime selection to be consistent.
183 One path used ike23 still, instead of exim.dev.20160529.3; now both
184 execution flows will use the same DH primes (currently
185 exim.dev.20160529.3).
187 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
188 option in smtp transport, to match the documentation. Previously
189 verification was not being done in some cases where it should have been.
191 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
192 than one server was defined and depending on the platform memory layout
193 details, an internal consistency trap could be hit while walking the list
196 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
197 files. The value is used by the authresults expansion item. Previously
198 if this was used in a router or transport, a crash could result.
200 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
201 have a sufficient incidence of EINTR returns from select that an
202 interaction with other operations done by the main daemon loop exposed
203 a bug in the error-handling. This was benign apart from the log
206 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
207 for consistency with delivered "=>" and failed "**" lines. While we're
208 there, handle PRX and TFO.
210 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
211 applied. This resulted, if any header-line rewrite rules were configured,
212 in a panic-log triggerable by sending a message with a long address in
213 a header. Fix by increasing the arbitrary limit to larger than a single
214 (dewrapped) 5322 header line maximum size.
216 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
217 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
218 RFC 6648 which deprecates X- options in protocols as a general practice.
219 Changeover between the implementations is handled by the mechanisms
222 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
223 a remote which uses callout/hold. Previously the local was queued.
225 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
228 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
229 log line. The asterisk is now consistently appended to the remote IP
230 (and port, if given), and will also be provided on defer and fail log
231 lines. Previously it could be placed on the local IP if that was being
232 logged, and was only provided on delivery lines.
234 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
236 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
237 authenticator client_send option. Previously the next char, after a pair
238 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
239 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
240 documentation. There is still no way to get a leading ^ immediately
241 after a NUL (ie. for the password of a PLAIN method authenticator.
243 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
244 files. For bad sizes read, delete the record and whine to paniclog.
246 JH/38 When logging an AUTH failure, as server, do not include sensitive
247 information. Previously, the credentials would be included if given
248 as part of the AUTH command line and an ACL denied authentication.
250 JH/39 Bug 2691: fix $local_part_data. When the matching list element
251 referred to a file, bad data was returned. This likely also affected
254 JH/40 The gsasl authenticator now supports caching of the salted password
255 generated by the client-side implementation. This required the addition
256 of a new variable: $auth4.
258 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
259 left undeleted; the attempt to re-create it then failed - resulting in
260 the usual "SIGHUP tp have daemon reload configuration" to not work.
261 This affected any platform not supporting "abstract" Unix-domain
262 sockets (i.e. not Linux).
264 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
265 recipients" response to RCPT in a later response, with a 250. The
266 previous coding assumed this would not happen, and under PIPELINING
267 would result in both lost and duplicate recipients for a message.
269 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
270 Previously the weighting was incorrectly applied. Similar fix for socks
271 proxies. Found and fixed by Heiko Schlichting.
273 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
274 not handle sub-lists included using the +namedlist syntax. While
275 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
276 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
279 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
280 Previously the permanent pool was used, so the sore could not be freed.
281 This meant a connection with many messages would use continually-growing
284 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
285 per-pool so as not to waste too much space. Previously a constant size
286 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
287 DOS attacks harder. The cost is wasted memory use in the larger blocks.
289 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
290 same space repeatedly, and a smaller process footprint.
292 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
293 for GnuTLS. Previously the name started "%s" which, while not a bug,
294 looked as if if might be one.
296 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
297 "continued connection") make the $proxy_* variables available. Previously
298 the information was not passed across the exec() call for subsequent
299 transport executions. This also mean that the log lines for the
300 messages can show the proxy information.
302 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
303 receive time. With modern systems the difference is significant.
304 The historical behaviour can be restored by disabling (a new) log_selector
305 "queue_time_exclusive".
307 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
308 content) were specifically tested for. Now, also cover expansions
309 resulting in ACL names and inline ACL content.
311 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
312 making it unusable in complex expressions.
314 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
315 fail with a taint-check on the spoolfile name, and leave the message
318 HS/01 Enforce absolute PID file path name.
320 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
322 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
324 PP/02 Bug 2643: Correct TLS DH constants.
325 A missing NUL termination in our code-generation tool had led to some
326 incorrect Diffie-Hellman constants in the Exim source.
327 Reported by kylon94, code-gen tool fix by Simon Arlott.
329 PP/03 Impose security length checks on various command-line options.
330 Fixes CVE-2020-SPRSS reported by Qualys.
332 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
333 better. Reported by Qualys.
335 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
336 providing a particularly obnoxious sender full name.
339 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
341 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
344 PP/08 Change default for recipients_max from unlimited to 50,000.
346 PP/09 Fix security issue with too many recipients on a message (to remove a
347 known security problem if someone does set recipients_max to unlimited,
348 or if local additions add to the recipient list).
349 Fixes CVE-2020-RCPTL reported by Qualys.
351 PP/10 Fix security issue in SMTP verb option parsing
352 Fixes CVE-2020-EXOPT reported by Qualys.
354 PP/11 Fix security issue in BDAT state confusion.
355 Ensure we reset known-good where we know we need to not be reading BDAT
356 data, as a general case fix, and move the places where we switch to BDAT
357 mode until after various protocol state checks.
358 Fixes CVE-2020-BDATA reported by Qualys.
360 HS/03 Die on "/../" in msglog file names
362 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
363 the Exim runtime user are allowed to create files.
365 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
368 QS/03 When reading the output from interpreted forward files we do not
369 pass the pipe between the parent and the interpreting process to
370 executed child processes (if any).
372 QS/04 Always die if requested from internal logging, even is logging is
375 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
376 the API; compilation noo longer completes with DMARC support included.
377 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
378 been reported on other platforms.
380 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
382 JH/56 Make the majority of info read from config files readonly, for defence-in-
383 depth against exploits. Suggestion by Qualys.
384 Not supported on Solaris 10.
386 JH/57 Fix control=fakreject for a custom message containing tainted data.
387 Previously this resulted in a log complaint, due to a re-expansion present
388 since fakereject was originally introduced.
390 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
391 modifier was given, a loop resulted.
393 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
394 pipelined SMTP command followed the BDAT LAST then it would be
395 incorrectly treated as part of the message body, causing a verification
398 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
399 require looser validation rules than those for 821-level addresses,
400 which only permit IP addresses.
406 JH/01 Avoid costly startup code when not strictly needed. This reduces time
407 for some exim process initialisations. It does mean that the logging
408 of TLS configuration problems is only done for the daemon startup.
410 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
412 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
413 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
416 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
417 DKIM signing is being done. Previously a transport_filter would always
418 disable CHUNKING, falling back to traditional DATA.
420 JH/05 Regard command-line recipients as tainted.
422 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
424 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
425 PAM library frees one of the arguments given to it, despite the
426 documentation. Therefore a plain malloc must be used.
428 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
429 on-stack buffers were used, resulting in a taint trap when DSN information
430 copied from a received message was written into the buffer.
432 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
433 the ordering of its ARC headers. This caused a crash.
435 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
436 a new record was being constructed with information from the peer, a trap
439 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
440 installation would get error messages from DMARC verify, when it hit the
441 nonexistent file indicated by the default. Distros wanting DMARC enabled
442 should both provide the file and set the option.
443 Also enforce no DMARC verification for command-line sourced messages.
445 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
446 could, depending on the platform, hang at the STARTTLS response.
448 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
449 message on a connection. Previously if one message had ARC headers and
450 the following one did not, a crash could result when adding an
451 Authentication-Results: header.
453 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
454 Exim main code and Exim-related utities. The introduction of taint
455 tracking also did many adjustments to string handling. Since then, eximon
456 frequently terminated with an assert failure.
458 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
459 check for 452 responses. This slightly helps the inefficieny of doing
460 a large alias-expansion into a recipient-limited target. The max_rcpt
461 transport option still applies (and at the current default, will override
462 the new feature). The check is done for either cause of synch, and forces
463 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
464 connection. The new facility is not tunable at this time.
466 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
467 library live data was being used, so the results became garbage. Make
468 copies while it is still usable.
470 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
471 on delivery deferred (==) and failed (**) lines (if a delivery was
472 attemtped). Previously it was only on completion (=>) lines.
474 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
475 for the expansion of the server_scram_iter and server_scram_salt options.
477 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
478 are now specifically given a NO_DATA response without hitting the system
479 resolver. The library goes on to do the now-standard TXT lookup.
480 Use of dnsdb lookups is not affected.
482 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
483 only retrieve the errormessage once. Previously two calls to dlerror()
484 were used, and the second one (for mainlog/paniclog) retrieved null
487 JH/20 Taint checking: disallow use of tainted data for
488 - the appendfile transport file and directory options
489 - the pipe transport command
490 - the autoreply transport file, log and once options
491 - file names used by the redirect router (including filter files)
493 - paths used by single-key lookups
494 Previously this was permitted.
496 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
497 adjusted the size of a major service buffer; this failed because the
498 buffer was in use at the time. Change to a compile-time increase in the
499 buffer size, when this authenticator is compiled into exim.
501 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
502 previous fast-mode was untenable in the face of glibs using mmap to
503 support larger malloc requests.
505 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
506 New values supported, if defined on system where compiled:
507 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
508 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
510 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
511 running a limited number of proceses in parallel, a benefit is gained. The
512 amount varies with the platform hardware and load. The use of the option
513 queue_run_in_order means we cannot do this, as ordering becomes
516 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
517 had introduced a string-copy (for ensuring NUL-termination) which was not
518 appropriate for that case, which can include embedded NUL bytes in the
519 block of data. Investigation showed the copy to actually be needless, the
520 data being length-specified.
522 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
523 done during a receiving connection, and both used TLS, global info was
524 used rather than per-connection info for tracking the state of data
525 queued for transmission. This could result in a connection hang.
527 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
528 Previously, when delivering serveral messages down a single connection
529 only the first would provide a SIZE. This was due to the size information
530 not being properly tracked.
532 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
533 TAI (at 37 seconds currently), pretend to be in UTC for time-related
534 expansion and logging. Previously, spurious values such as a future
535 minute could be seen.
537 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
538 it could crash from a null-deref. This could also affect the
539 ${addresses: } operator and ${readsock } item.
541 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
542 message following a mime one, the variable was not reset.
544 JH/30 When an pipelined-connect fails at the first response, assume incorrect
545 cached capability (perhaps the peer reneged?) and immediately retry in
548 JH/31 Fix spurious detection of timeout while writing to transport filter.
550 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
551 an attempt to copy the string was made before checking it.
553 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
554 taint of the lookup key was maintained; we now regard the presence in the
555 filesystem as sufficient validation.
557 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
558 argument is supplied.
560 JH/35 The dsearch lookup now requires that the directory is an absolute path.
561 Previously this was not checked, and nonempty relative paths made an
562 access under Exim's current working directory.
564 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
565 Previously no event was raised.
567 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
568 parameter supplied by the sender MAIL FROM command. Previously it was
569 ignored, and only the check_spool_space option value for the required
572 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
573 the size of the signing public-key. Previously it was instead giving
574 the size of the signature hash.
576 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
577 the default. See the (new) dkim_verify_min_keysizes option.
579 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
580 and an ACL use a lookup for checking either the local_part or domain,
581 stale data could be accessed. Ensure that variable references are
582 dropped between messages.
584 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
585 by the client was not checked as pointing within response data before
586 being used. A malicious client could thus cause an out-of-bounds read and
587 possibly gain authentication. Fix by adding the check.
589 JH/42 Internationalisation: change the default for downconversion in the smtp
590 transport to be "if needed". Previously it was "as previously set" for
591 the message, which usually meant "if needed" for message-submission but
592 "no" for everything else. However, MTAs have been seen using SMTPUTF8
593 even when the envelope addresses did not need it, resulting in forwarding
594 failures to non-supporting MTAs. A downconvert in such cases will be
595 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
596 transport. The change does mean that addresses needing conversion will
597 be converted when previously a delivery failure would occur.
599 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
600 response was received it would be used unchecked in a fail-DSN, violating
601 standards on line-length limits. Truncate if needed.
603 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
604 doesn't work. (Additionally add a new main config option to configure the
611 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
612 side, to decode a TLS 1.3 packet capture.
614 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
615 Previously the default library behaviour applied, sending two, each in
618 JH/03 Debug output for ACL now gives the config file name and line number for
621 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
623 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
625 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
626 buffer overrun for (non-chunking) other transports.
628 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
629 TLS1.3, means that a server rejecting a client certificate is not visible
630 to the client until the first read of encrypted data (typically the
631 response to EHLO). Add detection for that case and treat it as a failed
632 TLS connection attempt, so that the normal retry-in-clear can work (if
633 suitably configured).
635 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
636 and/or domain. Found and fixed by Jason Betts.
638 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
639 configuration). If a CNAME target was not a wellformed name pattern, a
642 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
643 the OS reports them interleaved with other addresses.
645 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
646 used both for input and for a verify callout, both encrypted, SMTP
647 responses being sent by the server could be lost. This resulted in
648 dropped connections and sometimes bounces generated by a peer sending
651 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
652 string. Previously it was possible to cause undefined behaviour in a
653 library routine (usually a crash). Found by "zerons".
655 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
658 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
659 API was removed, so update to use the newer ones.
661 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
662 any timeout set, is taking a long time. Previously we would hang on to a
663 rotated logfile "forever" if the input was arriving with long gaps
664 (a previous attempt to fix addressed lack, for a long time, of initial
667 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
668 shared (NFS) environment. The length of the tempfile name is now
669 4 + 16 ("hdr.$message_exim_id") which might break on file
670 systems which restrict the file name length to lower values.
673 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
674 shared (NFS) environment.
676 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
677 did for all versions <4.90). Notably -M, -m, --invert, -I may be
680 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
681 on some platforms for bit 31.
683 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
684 to changes apparently associated with TLS1.3 handling some of the APIs
685 previously used were either nonfunctional or inappropriate. Strings
686 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
687 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
688 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
689 This affects log line X= elements, the $tls_{in,out}_cipher variables,
690 and the use of specific cipher names in the encrypted= ACL condition.
692 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
694 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
695 verification result was not updated unless hosts_require_ocsp applied.
697 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
698 queue_list_requires_admin set to false, non-admin users were denied the
701 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
702 directory-of-certs mode. Previously they were advertised despite the
705 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
706 A single TCP connection by a client will now hold a TLS connection open
707 for multiple message deliveries, by default. Previously the default was to
710 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
711 default. If built with the facility, DANE will be used. The facility
712 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
714 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
715 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
716 must be defined and you must still, unless you define DISABLE_TLS, manage
717 the the include-dir and library-file requirements that go with that
718 choice. Non-TLS builds are still supported.
720 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
723 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
724 default. If the platform supports and has the facility enabled, it will
725 be requested on all coneections.
727 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
728 controlled by the build-time option SUPPORT_PIPE_CONNECT.
730 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
732 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
733 success-DSN messages. Previously the From: header was always the default
734 one for these; the option was ignored.
736 JH/28 Fix the timeout on smtp response to apply to the whole response.
737 Previously it was reset for every read, so a teergrubing peer sending
738 single bytes within the time limit could extend the connection for a
739 long time. Credit to Qualsys Security Advisory Team for the discovery.
741 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
742 delivery address, which leaked information of the results of local
743 forwarding. Change to the original envelope recipient address, per
746 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
747 requested. Previously not bounce was generated and a log entry of
748 error ignored was made.
750 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
752 JH/32 Introduce a general tainting mechanism for values read from the input
753 channel, and values derived from them. Refuse to expand any tainted
754 values, to catch one form of exploit.
756 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
757 was unused and the unexpanded text used for the test. Found and
758 fixed by Ruben Jenster.
760 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
761 an attempt to use a TLS library read routine dereffed a nul pointer,
764 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
765 them in our smtp response.
767 JH/36 Have the general router option retry_use_local_part default to true when
768 any of the restrictive preconditions are set (to anything). Previously it
769 was only for check_local user. The change removes one item of manual
770 configuration which is required for proper retries when a remote router
771 handles a subset of addresses for a domain.
773 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
774 link count into consideration.
776 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
777 caused the extension of big_buffer, the following lines were ignored.
779 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
780 accordance with RFC 2308. Previously there was no expiry, so a longlived
781 receive process (eg. due to ACL delays) versus a short SOA value could
784 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
786 JH/39 Promote DMARC support to mainline.
788 JH/40 Bug 2452: Add a References: header to DSNs.
790 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
791 parameters. The relevant library call is documented as "Deprecated: This
792 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
793 3.6.0, DH parameters are negotiated following RFC7919."
795 HS/06 Change the default of dnssec_request_domains to "*"
797 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
798 carried on and emitted a BDAT command, even when PIPELINING was not
801 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
802 buffer was used for the filename, resulting in a trap when tainted
803 arguments (eg. $domain) were used.
805 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
806 recommended to avoid a possible server-load attack. The feature can be
807 re-enabled via the openssl_options main cofiguration option.
809 JH/45 local_scan API: documented the current smtp_printf() call. This changed
810 for version 4.90 - adding a "more data" boolean to the arguments.
811 Bumped the ABI version number also, this having been missed previously;
812 release versions 4.90 to 4.92.3 inclusive were effectively broken in
813 respect of usage of smtp_printf() by either local_scan code or libraries
814 accessed via the ${dlfunc } expansion item. Both will need coding
815 adjustment for any calls to smtp_printf() to match the new function
816 signature; a FALSE value for the new argument is always safe.
818 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
819 the file-offset (which the Linux syscall does, and exim expects); this
820 resulted in an indefinite loop.
822 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
823 to do ARC verification. The Authentication-Results: header line added
824 by the configuration then had no ARC item.
826 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
827 an incorrect mode was used for reading the result, resulting in it being
834 JH/01 Remove code calling the customisable local_scan function, unless a new
835 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
837 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
838 non-signal-safe functions being used.
840 JH/03 Bug 2269: When presented with a received message having a stupidly large
841 number of DKIM-Signature headers, disable DKIM verification to avoid
842 a resource-consumption attack. The limit is set at twenty.
844 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
845 report of oldest_pass in ${authres } in consequence, and separate out
846 some descriptions of reasons for verification fail.
848 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
849 files in the spool were present and unlocked. A queue-runner could spot
850 them, resulting in a duplicate delivery. Fix that by doing the unlock
851 after the unlink. Investigation by Tim Stewart. Take the opportunity to
852 add more error-checking on spoolfile handling while that code is being
855 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
856 No known attacks, no CVE, this is defensive hardening.
858 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
859 a queue-runner could start a delivery while other operations were ongoing.
860 Cutthrough delivery was a common victim, resulting in duplicate delivery.
861 Found and investigated by Tim Stewart. Fix by using the open message data
862 file handle rather than opening another, and not locally closing it (which
863 releases a lock) for that case, while creating the temporary .eml format
864 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
866 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
867 $sender_verify_failure/$recipient_verify_failure to "random".
869 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
872 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
873 Previously this would segfault.
875 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
878 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
879 like zero, since the resolver should be doing this for us, But we need one
880 as a CNAME but no MX presence gets the CNAME returned; we need to check
881 that doesn't point to an MX to declare it "no MX returned" rather than
882 "error, loop". A new main option is added so the older capability of
883 following some limited number of chain links is maintained.
885 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
887 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
888 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
889 now-deprecated earlier definitions used only specified the range up to TLS
890 1.2 (in the older-version library docs).
892 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
894 JH/15 Rework TLS client-side context management. Stop using a global, and
895 explicitly pass a context around. This enables future use of TLS for
896 connections to service-daemons (eg. malware scanning) while a client smtp
897 connection is using TLS; with cutthrough connections this is quite likely.
899 JH/16 Fix ARC verification to do AS checks in reverse order.
901 JH/17 Support a "tls" option on the ${readsocket } expansion item.
903 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
904 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
905 Previously the "utf8" would be re-prepended for every additional message.
907 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
908 Previously thery were accepted, resulting in issues when attempting to
909 forward messages to a non-supporting MTA.
911 PP/02 Let -n work with printing macros too, not just options.
913 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
914 one parent address was copied, and bogus data was used at delivery-logging
915 time. Either a crash (after delivery) or bogus log data could result.
916 Discovery and analysis by Tim Stewart.
918 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
919 Previously if the string ended mid-character, we did not insert the
920 promised '?' replacement.
922 PP/04 Documentation: current string operators work on bytes, not codepoints.
924 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
925 should pack well giving a smaller memory footprint so better caching and
926 therefore performance. Group the declarations where this can't be done so
927 that the byte-sized flag variables are not interspersed among pointer
928 variables, giving a better chance of good packing by the compiler.
930 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
931 non-null, to avoid issues with sites running BATV. Previously reports were
932 sent with an empty envelope sender so looked like bounces.
934 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
935 The ignore_error flag wasn't being returned from the filter subprocess so
936 was not set for later routers. Investigation and fix by Matthias Kurz.
938 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
939 and a msg:complete for the whole, when a message is manually removed using
940 -Mrm. Developement by Matthias Kurz, hacked on by JH.
942 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
943 a "Gnu special" function, asprintf() in the DB utility binary builds; I
944 hope that is portable enough.
946 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
947 requiring a known-CA anchor certificate; make it now rely entirely on the
948 TLSA as an anchor. Checking the name on the leaf cert against the name
949 on the A-record for the host is still done for TA (but not for EE mode).
951 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
952 included in delivery lines for non-proxied connections, when compiled with
953 SUPPORT_SOCKS and running with proxy logging enabled.
955 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
956 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
957 move the existing event to fire before the normal logging of message
958 failure so that custom logging is bracketed by normal logging.
960 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
961 msg:fail:internal event. Developement by Matthias Kurz.
963 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
964 far too small for todays use of crypto signatures stored there. Go all
965 the way to the max DNS message size of 64kB, even though this might be
966 overmuch for IOT constrained device use.
968 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
969 copy a string over itself. The library routine is documented as not
970 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
972 JH/32 For main options check_spool_space and check_inode_space, where the
973 platform supports 64b integers, support more than the previous 2^31 kB
974 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
975 the previous G, M, k.
977 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
978 $authenticated_fail_id variable on authentication failure. Previously
981 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
982 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
983 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
984 GNUTLS_SEC_PARAM_MEDIUM.
986 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
987 side. Previously we would continue as if no SNI had been received.
989 JH/36 Harden the handling of string-lists. When a list consisted of a sole
990 "<" character, which should be a list-separator specification, we walked
991 off past the nul-terimation.
993 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
994 causes) even when the retry time is not yet met. Previously they were
995 not, meaning that when (say) an account was over-quota and temp-rejecting,
996 and multiple senders' messages were queued, only one sender would get
997 notified on each configured delay_warning cycle.
999 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1001 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1002 cipher-suites, an error can be left on the stack even for a succeeding
1003 accept; this results in impossible error messages when a later operation
1006 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1007 return error codes indicating retry. Under TLS1.3 this becomes required.
1009 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1010 it only wrote the new authenticators, resulting in a lack of tracking of
1011 peer changes of ESMTP extensions until the next cache flush.
1013 JH/41 Fix the loop reading a message header line to check for integer overflow,
1014 and more-often against header_maxsize. Previously a crafted message could
1015 induce a crash of the recive process; now the message is cleanly rejected.
1017 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1018 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1024 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1025 When redis_servers is set to a list of > 1 element, and the Redis servers
1026 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1027 case of MOVED into a DEFER case instead, thus moving the query onto the
1028 next server in the list. For a cluster of N elements, all N servers must
1029 be defined in redis_servers.
1031 GF/02 Catch and remove uninitialized value warning in exiqsumm
1032 Check for existence of @ARGV before looking at $ARGV[0]
1034 JH/01 Replace the store_release() internal interface with store_newblock(),
1035 which internalises the check required to safely use the old one, plus
1036 the allocate and data copy operations duplicated in both (!) of the
1037 extant use locations.
1039 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1040 modifier. This matches the restriction on the commandline.
1042 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1043 Previously only the last row was returned.
1045 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1046 we assumed that tags in the header were well-formed, and parsed the
1047 element content after inspecting only the first char of the tag.
1048 Assumptions at that stage could crash the receive process on malformed
1051 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1052 While running the DKIM ACL we operate on the Permanent memory pool so that
1053 variables created with "set" persist to the DATA ACL. Also (at any time)
1054 DNS lookups that fail create cache records using the Permanent pool. But
1055 expansions release any allocations made on the current pool - so a dnsdb
1056 lookup expansion done in the DKIM ACL releases the memory used for the
1057 DNS negative-cache, and bad things result. Solution is to switch to the
1058 Main pool for expansions.
1059 While we're in that code, add checks on the DNS cache during store_reset,
1060 active in the testsuite.
1061 Problem spotted, and debugging aided, by Wolfgang Breyha.
1063 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1064 When none of the hosts presented to a transport match an already-open
1065 connection, close it and proceed with the list. Previously we would
1066 queue the message. Spotted by Lena with Yahoo, probably involving
1069 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1070 Previously a spurious "250 OK id=" response was appended to the proper
1073 JH/08 The "support for" informational output now, which built with Content
1074 Scanning support, has a line for the malware scanner interfaces compiled
1075 in. Interface can be individually included or not at build time.
1077 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1078 by the template makefile "src/EDITME". The "STREAM" support for an older
1079 ClamAV interface method is removed.
1081 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1082 rows affected is given instead).
1084 JH/11 The runtime Berkeley DB library version is now additionally output by
1085 "exim -d -bV". Previously only the compile-time version was shown.
1087 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1088 SMTP connection. Previously, when one had more recipients than the
1089 first, an abortive onward connection was made. Move to full support for
1090 multiple onward connections in sequence, handling cutthrough connection
1091 for all multi-message initiating connections.
1093 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1094 routers. Previously, a multi-recipient message would fail to match the
1095 onward-connection opened for the first recipient, and cause its closure.
1097 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1098 a timeout on read on a GnuTLS initiating connection, resulting in the
1099 initiating connection being dropped. This mattered most when the callout
1100 was marked defer_ok. Fix to keep the two timeout-detection methods
1103 JH/15 Relax results from ACL control request to enable cutthrough, in
1104 unsupported situations, from error to silently (except under debug)
1105 ignoring. This covers use with PRDR, frozen messages, queue-only and
1108 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1110 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1111 metadata, resulting in a crash in free().
1113 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1114 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1115 Broken also in d185889f4, with init system revamp.
1117 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1118 Previously we abruptly closed the connection after reading a malware-
1119 found indication; now we go on to read the "scan ok" response line,
1122 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1123 ACL. Previously, a crash would result.
1125 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1126 macro text after a replacement (previously it was only once per line) and
1127 by skipping builtin macros when searching for an uppercase lead character.
1129 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1130 for the build is renamed.
1132 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1133 was allocated for every new TLS startup, meaning one per message. Fix
1134 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1136 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1137 reported the original. Fix to report (as far as possible) the ACL
1138 result replacing the original.
1140 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1141 OpenSSL. Certificate information is loaded for every new TLS startup,
1142 and the resources needed to be freed.
1144 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1146 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1149 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1150 DATA response info to the (existing) per-recipient response info for
1151 the "C=" log element. It can have useful tracking info from the
1152 destination system. Patch from Simon Arlott.
1154 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1155 length value. Previously this would segfault.
1157 HS/02 Support Avast multiline protoocol, this allows passing flags to
1158 newer versions of the scanner.
1160 JH/28 Ensure that variables possibly set during message acceptance are marked
1161 dead before release of memory in the daemon loop. This stops complaints
1162 about them when the debug_store option is enabled. Discovered specifically
1163 for sender_rate_period, but applies to a whole set of variables.
1164 Do the same for the queue-runner and queue-list loops, for variables set
1165 from spool message files. Do the same for the SMTP per-message loop, for
1166 certain variables indirectly set in ACL operations.
1168 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1169 as a multi-recipient message from a mailinglist manager). The coding had
1170 an arbitrary cutoff number of characters while checking for more input;
1171 enforced by writing a NUL into the buffer. This corrupted long / fast
1172 input. The problem was exposed more widely when more pipelineing of SMTP
1173 responses was introduced, and one Exim system was feeding another.
1174 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1175 receiving system, and refused recipients seen by the sending system
1176 (propating to people being dropped from mailing lists).
1177 Discovered and pinpointed by David Carter.
1179 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1180 replaced by the ${authresults } expansion.
1182 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1184 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1185 allows proper process termination in container environments.
1187 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1188 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1190 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1191 and "err_perm", deprecated since 4.83 when the RFC-defined words
1192 "temperror" and "permerror" were introduced.
1194 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1195 transport-filters or DKIM-signing. The restriction was lost in the
1196 consolidation of verify-callout and delivery SMTP handling.
1197 Extend the restriction to also cover ARC-signing.
1199 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1200 in defer=pass mode supply a 450 to the initiator. Previously the message
1203 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1204 tls_require_ciphers is used as before.
1206 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1207 "pass_unscanned". Only tmpfails from the scanner are written to
1208 the paniclog, as they may require admin intervention (permission
1209 denied, license issues). Other scanner errors (like decompression
1210 bombs) do not cause a paniclog entry.
1212 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1213 Previously it was possible to log spurious information in receive log
1216 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1217 triggered odd behaviour from Outlook Express clients.
1219 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1222 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1223 since the IETF WG has not yet settled on that versus the original
1224 "bare" representation.
1226 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1227 Previously the millisecond value corrupted the output.
1228 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1229 corrupted the output.
1235 JH/01 Rework error string handling in TLS interface so that the caller in
1236 more cases is responsible for logging. This permits library-sourced
1237 string to be attached to addresses during delivery, and collapses
1238 pairs of long lines into single ones.
1240 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1241 during configuration. Wildcards are allowed and expanded.
1243 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1244 This permits better logging.
1246 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1247 do not close it down and have the child transport start it up again on
1248 the passed-on TCP connection. Instead, proxy the child (and any
1249 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1250 affected: the continued delivery log lines do not have any DNSSEC, TLS
1251 Certificate or OCSP information. TLS cipher information is still logged.
1253 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1254 identical IP addresses on different listening ports. Will also affect
1257 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1258 add noisy ifdef guards to special-case this sillyness.
1259 Patch from Bernd Kuhls.
1261 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1262 than 255 are no longer allowed.
1264 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1265 Disable the session-cache too, which might reduce our load. Since we
1266 currrectly use a new context for every connection, both as server and
1267 client, there is no benefit for these.
1268 GnuTLS appears to not support tickets server-side by default (we don't
1269 call gnutls_session_ticket_enable_server()) but client side is enabled
1270 by default on recent versions (3.1.3 +) unless the PFS priority string
1273 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1274 <https://reproducible-builds.org/specs/source-date-epoch/>.
1276 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1277 the check for any unsuccessful recipients did not notice the limit, and
1278 erroneously found still-pending ones.
1280 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1281 MSG_MORE. Only in-clear (not on TLS connections).
1283 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1284 transport-filter is involved we can buffer the headers in memory for
1285 creating the signature, and read the spool data file once for the
1286 signature and again for transmission.
1288 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1289 4.77 as the kernel support then wasn't solid, having issues in 64bit
1290 mode. Now, it's been long enough. Add support for FreeBSD also.
1292 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1293 case where the routing stage had gathered several addresses to send to
1294 a host before calling the transport for the first, we previously failed
1295 to close down TLS in the old transport process before passing the TCP
1296 connection to the new process. The new one sent a STARTTLS command
1297 which naturally failed, giving a failed delivery and bloating the retry
1298 database. Investigation and fix prototype from Wolfgang Breyha.
1300 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1301 false-negatives in the check that the sender had not preempted a response
1302 or prompt from Exim (running as a server), due to that code's lack of
1303 awareness of the SMTP input buffering.
1305 PP/04 Add commandline_checks_require_admin option.
1306 Exim drops privileges sanely, various checks such as -be aren't a
1307 security problem, as long as you trust local users with access to their
1308 own account. When invoked by services which pass untrusted data to
1309 Exim, this might be an issue. Set this option in main configuration
1310 AND make fixes to the calling application, such as using `--` to stop
1313 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1314 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1315 responses to those, into a single TLS record each way (this usually means
1316 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1319 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1320 affects you only if you're dancing at the edge of the param size limits.
1321 If you are, and this message makes sense to you, then: raise the
1322 configured limit or use OpenSSL 1.1. Nothing we can do for older
1325 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1326 cmdline element to get the documented default one. Previously it was
1329 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1330 are unsuable for non-message-specific reasons - eg. connection timeout,
1331 banner-time rejection.
1333 JH/16 Fix logging of delivery remote port, when specified by router, under
1336 PP/06 Repair manualroute's ability to take options in any order, even if one
1337 is the name of a transport.
1340 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1342 JH/17 Change the list-building routines interface to use the expanding-string
1343 triplet model, for better allocation and copying behaviour.
1345 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1346 Previously it was constructed the first time a possibly-matching string
1347 was met in the configuration file input during startup; now it is done
1350 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1351 compatible one, to avoid the (poorly documented) possibility of a config
1352 file in the working directory redirecting the DB files, possibly correpting
1353 some existing file. CVE-2017-10140 assigned for BDB.
1355 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1356 cache-hot. Previously, although the result was properly cached, the
1357 initial verify call returned a defer.
1359 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1360 the main verify for receipient in uncached-mode.
1362 JH/22 Retire historical build files to an "unsupported" subdir. These are
1363 defined as "ones for which we have no current evidence of testing".
1365 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1366 if present. Previously it was ignored.
1368 JH/24 Start using specified-initialisers in C structure init coding. This is
1369 a C99 feature (it's 2017, so now considered safe).
1371 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1372 if was a fixed-sized field and bitmask ops via macros; it is now more
1375 PP/07 GitHub PR 56: Apply MariaDB build fix.
1376 Patch provided by Jaroslav Škarvada.
1378 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1379 during Coverity cleanups [4.87 JH/47]
1380 Diagnosis and fix provided by Michael Fischer v. Mollard.
1382 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1383 the right size to place the terminating semicolon on its own folded
1384 line, the header hash was calculated to an incorrect value thanks to
1385 the (relaxed) space the fold became.
1387 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1388 and confused the parent.
1390 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1391 which could crash as a result. This could lead to undeliverable messages.
1393 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1396 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1397 matters on fast-turnover and PID-randomising systems, which were getting
1398 out-of-order delivery.
1400 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1401 a possibly-overlapping copy. The symptom was that "Remote host closed
1402 connection in response to HELO" was logged instead of the actual 4xx
1405 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1406 Previously only that bufferd was discarded, resulting in SYMTP command
1409 JH/32 DKIM: when a message has multiple signatures matching an identity given
1410 in dkim_verify_signers, run the dkim acl once for each. Previously only
1411 one run was done. Bug 2189.
1413 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1414 "panic the current process" to "deliberately defer". The panic log is
1415 still written with the problem list name; the mail and reject logs now
1416 get a temp-reject line for the message that was being handled, saying
1417 something like "domains check lookup or other defer". The SMTP 451
1418 message is still "Temporary local problem".
1420 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1421 A crafted sequence of BDAT commands could result in in-use memory beeing
1422 freed. CVE-2017-16943.
1424 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1425 from SMTP input. Previously it was always done; now only done for DATA
1426 and not BDAT commands. CVE-2017-16944.
1428 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1429 to the message (such as an overlong header line). Previously this was
1430 not done and we did not exit BDAT mode. Followon from the previous item
1431 though a different problem.
1437 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1438 than -2003 did; needs libidn2 in addition to libidn.
1440 JH/02 The path option on a pipe transport is now expanded before use.
1442 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1443 Patch provided by "Björn", documentation fix added too.
1445 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1446 missing a wire-to-host endian conversion.
1448 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1449 close after a BDAT command line could be taken as a following command,
1450 giving a synch failure. Fix by only checking for synch immediately
1451 before acknowledging the chunk.
1453 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1454 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1455 macro. Patches provided by Josh Soref.
1457 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1458 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1459 by IANA as a service extension. However, John Klensin suggests that we
1462 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1463 the dkim code may be unix-mode line endings rather than smtp wire-format
1464 CRLF, so prepend a CR to any bare LF.
1466 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1467 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1469 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1470 the first were themselves being wrongly included in the feed into dkim
1471 processing; with most chunk sizes in use this resulted in an incorrect
1472 body hash calculated value.
1474 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1475 DKIM signature block, for verification. Although advised against by
1476 standards it is specifically not ruled illegal.
1478 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1480 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1481 missing a body hash (the bh= tag).
1483 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1484 It seems that HAProxy sends the Proxy Protocol information in clear and
1485 only then does a TLS startup, so do the same.
1487 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1488 TCP connections (such as for Spamd) unless the daemon successfully set
1489 Fast Open mode on its listening sockets. This fixes breakage seen on
1490 too-old kernels or those not configured for Fast Open, at the cost of
1491 requiring both directions being enabled for TFO, and TFO never being used
1492 by non-daemon-related Exim processes.
1494 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1495 endings, at least on the first header line. Try to canonify any that get
1496 past that check, despite the cost.
1498 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1499 now limited to an arbitrary five deep, while parsing addresses with the
1500 strip_excess_angle_brackets option enabled.
1502 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1503 instead leave the unprompted TLS handshake in socket buffer for the
1504 TLS library to consume.
1506 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1508 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1510 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1511 region in one operation, for speed, and this leaves assigned pointers
1512 dangling. Add checks run only under the testsuite which checks all
1513 variables at a store-reset and panics on a dangling pointer; add code
1514 explicitly nulling out all the variables discovered. Fixes one known
1515 bug: a transport crash, where a dangling pointer for $sending_ip_address
1516 originally assigned in a verify callout, is re-used.
1518 PP/06 Drop '.' from @INC in various Perl scripts.
1520 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1522 PP/08 Reduce a number of compilation warnings under clang; building with
1523 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1524 should be warning-free.
1526 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1528 HS/01 Fix portability problems introduced by PP/08 for platforms where
1529 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1531 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1532 chunk. This allows us to accept broken chunked messages. We need a more
1533 general solution here.
1535 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1536 already-broken messages in the queue.
1538 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1540 JH/19 Fix reference counting bug in routing-generated-address tracking.
1546 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1547 supports it and a size is available (ie. the sending peer gave us one).
1549 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1550 years of being deprecated). The replacements are the ACLs
1551 acl_smtp_mime and acl_not_smtp_mime.
1553 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1554 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1555 or even an in-clear connection were permitted. Now, if the host lookup
1556 was dnssec and dane was requested then the host is only used if the
1557 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1558 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1559 if one fails this test.
1560 This means that a poorly-configured remote DNS will make it incommunicado;
1561 but it protects against a DNS-interception attack on it.
1563 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1564 non-noisy when a race steals the message being considered.
1566 JH/05 If main configuration option tls_certificate is unset, generate a
1567 self-signed certificate for inbound TLS connections.
1569 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1570 in rewrites and routers.
1572 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1573 and logged a warning sing 4.83; now they are a configuration file error.
1575 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1576 (lacking @domain). Apply the same qualification processing as RCPT.
1578 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1580 JH/10 Support ${sha256:} applied to a string (as well as the previous
1583 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1584 a cutthrough deliver is pending, as we always want to make a connection.
1585 This also avoids re-routing the message when later placing the cutthrough
1586 connection after a verify cache hit.
1587 Do not update it with the verify result either.
1589 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1590 when routing results in more than one destination address.
1592 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1593 signing (which inhibits the cutthrough capability). Previously only
1594 the presence of an option was tested; now an expansion evaluating as
1595 empty is permissible (obviously it should depend only on data available
1596 when the cutthrough connection is made).
1598 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1599 the relevant preceding SMTP command did not note the pipelining mode.
1601 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1602 Previously they were not counted.
1604 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1605 as one having no matching records. Previously we deferred the message
1606 that needed the lookup.
1608 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1609 distinguished as "(=".
1611 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1612 for missing MX records. Previously it only worked for missing A records.
1614 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1616 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1617 after the data-go-ahead and data-ack. Patch from Jason Betts.
1619 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1620 even for a "none" policy. Patch from Tony Meyer.
1622 JH/22 Fix continued use of a connection for further deliveries. If a port was
1623 specified by a router, it must also match for the delivery to be
1626 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1627 When one of the recipients of a message was unsuitable for the connection
1628 (has no matching addresses), we lost track of needing to mark it
1629 deferred. As a result mail would be lost.
1631 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1633 JH/25 Decoding ACL controls is now done using a binary search; the source code
1634 takes up less space and should be simpler to maintain. Merge the ACL
1635 condition decode tables also, with similar effect.
1637 JH/26 Fix problem with one_time used on a redirect router which returned the
1638 parent address unchanged. A retry would see the parent address marked as
1639 delivered, so not attempt the (identical) child. As a result mail would
1642 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1643 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1644 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1647 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1648 Main config options check_{log,spool}_{inodes,space} are now
1649 100 inodes, 10MB unless set otherwise in the configuration.
1651 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1652 Previously it only applied to the main-section connection policy
1655 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1657 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1658 by me. Added RFC7919 DH primes as an alternative.
1660 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1661 are not in the system include path.
1663 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1664 GnuTLS, when a session startup failed (eg because the client disconnected)
1665 Exim did stdio operations after fclose. This was exposed by a recent
1666 change which nulled out the file handle after the fclose.
1668 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1669 signed directly by the cert-signing cert, rather than an intermediate
1670 OCSP-signing cert. This is the model used by LetsEncrypt.
1672 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1674 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1675 an incoming connection.
1677 HS/02 Bug 1802: Do not half-close the connection after sending a request
1680 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1681 fallback to "prime256v1".
1683 JH/34 SECURITY: Use proper copy of DATA command in error message.
1684 Could leak key material. Remotely exploitable. CVE-2016-9963.
1690 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1691 and 3.4.4 - once the server is enabled to respond to an OCSP request
1692 it does even when not requested, resulting in a stapling non-aware
1693 client dropping the TLS connection.
1695 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1696 support variable-length bit vectors. No functional change.
1698 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1699 The I= interface field on outgoing lines is now after the H= remote
1700 host field, same as incoming lines. There is a separate
1701 outgoing_interface log selector which allows you to disable the
1704 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1705 If not running log_selector +smtp_connection the mainlog would be held
1706 open indefinitely after a "too many connections" event, including to a
1707 deleted file after a log rotate. Leave the per net connection logging
1708 leaving it open for efficiency as that will be quickly detected by the
1709 check on the next write.
1711 HS/01 Bug 1671: Fix post transport crash.
1712 Processing the wait-<transport> messages could crash the delivery
1713 process if the message IDs didn't exist for some reason. When
1714 using 'split_spool_directory=yes' the construction of the spool
1715 file name failed already, exposing the same netto behaviour.
1717 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1718 mime_regex ACL conditions.
1720 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1721 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1722 to HELO, local diagnostic string.
1724 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1725 log line to debug. Even when configured with a tls authenticator many
1726 client connections are expected to not authenticate in this way, so
1727 an authenticate fail is not an error.
1729 HS/02 Add the Exim version string to the process info. This way exiwhat
1730 gives some more detail about the running daemon.
1732 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1733 matter for fast-change records such as DNSBLs.
1735 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1736 retry record, even if constant. There may be multiple transports with
1737 different interface settings and the retry behaviour needs to be kept
1740 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1742 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1744 JH/10 Bug 840: fix log_defer_output option of pipe transport
1746 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1747 hosts_randomize. This matters a lot when combined with mua_wrapper.
1749 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1750 ${quote_pgsql:<string>} operator.
1752 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1754 JH/14 Change header file name for freeradius-client. Relevant if compiling
1755 with Radius support; from the Gentoo tree and checked under Fedora.
1757 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1759 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1760 option transport_filter, meaning no filtering.
1762 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1764 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1765 defaults to "*" (all hosts). The variable is now available when not built
1766 with TLS, default unset, mainly to enable keeping the testsuite sane.
1767 If a server certificate is not supplied (via tls_certificate) an error is
1768 logged, and clients will find TLS connections fail on startup. Presumably
1769 they will retry in-clear.
1770 Packagers of Exim are strongly encouraged to create a server certificate
1771 at installation time.
1773 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1774 with the $config_file variable.
1776 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1777 in transport context, after the attempt, and per-recipient. The latter type
1778 is per host attempted. The event data is the error message, and the errno
1779 information encodes the lookup type (A vs. MX) used for the (first) host,
1780 and the trailing two digits of the smtp 4xx response.
1782 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1783 to write to mainlog (or rejectlog, paniclog) in the window between file
1784 creation and permissions/ownership being changed. Particularly affects
1785 installations where exicyclog is run as root, rather than exim user;
1786 result is that the running daemon panics and dies.
1788 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1790 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1791 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1792 "pri" and "weight". Note that the previous implicit priority given by the
1793 list order is no longer honoured.
1795 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1796 for DKIM processing.
1798 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1799 by defining SUPPORT_SOCKS.
1801 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1802 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1803 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1804 variables are renamed to proxy_{local,external}_{address,port}.
1806 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1807 for a build by defining SUPPORT_I18N
1809 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1810 of the query string, and make ${quote_redis:} do that quoting.
1812 JH/29 Move Events support from Experimental to mainline, enabled by default
1813 and removable for a build by defining DISABLE_EVENT.
1815 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1817 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1818 cached by the daemon.
1820 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1821 by defining LOOKUP_REDIS. The libhiredis library is required.
1823 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1824 keys are given for lookup.
1826 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1827 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1828 only supported when built with TLS support. The PolarSSL SHA routines
1829 are still used when the TLS library is too old for convenient support.
1831 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1832 openssl_options), for security. OpenSSL forces this from version 1.1.0
1833 server-side so match that on older versions.
1835 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1836 allocation for $value could be released as the expansion processing
1837 concluded, but leaving the global pointer active for it.
1839 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1840 and to use the domains and local_parts ACL conditions.
1842 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1843 incorrectly not doubled on cutthrough transmission, hence seen as a
1844 body-termination at the receiving system - resulting in truncated mails.
1845 Commonly the sender saw a TCP-level error, and retransmitted the message
1846 via the normal store-and-forward channel. This could result in duplicates
1847 received - but deduplicating mailstores were liable to retain only the
1848 initial truncated version.
1850 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1852 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1854 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1855 we're in there, support oversigning also; bug 1309.
1857 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1859 HS/04 Add support for keep_environment and add_environment options.
1861 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1862 either intentional arithmetic overflow during PRNG, or testing config-
1865 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1866 delivery resulted in actual delivery. Cancel cutthrough before DATA
1869 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1870 rejects a recipient: pass the reject to the originator.
1872 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1873 Many were false-positives and ignorable, but it's worth fixing the
1876 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1877 for the new environment-manipulation done at startup. Move the routines
1878 from being local to tls.c to being global via the os.c file.
1880 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1881 an extract embedded as result-arg for a map, the first arg for extract
1882 is unavailable so we cannot tell if this is a numbered or keyed
1883 extraction. Accept either.
1889 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1892 JH/02 The smtp transport option "multi_domain" is now expanded.
1894 JH/03 The smtp transport now requests PRDR by default, if the server offers
1897 JH/04 Certificate name checking on server certificates, when exim is a client,
1898 is now done by default. The transport option tls_verify_cert_hostnames
1899 can be used to disable this per-host. The build option
1900 EXPERIMENTAL_CERTNAMES is withdrawn.
1902 JH/05 The value of the tls_verify_certificates smtp transport and main options
1903 default to the word "system" to access the system default CA bundle.
1904 For GnuTLS, only version 3.0.20 or later.
1906 JH/06 Verification of the server certificate for a TLS connection is now tried
1907 (but not required) by default. The verification status is now logged by
1908 default, for both outbound TLS and client-certificate supplying inbound
1911 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1914 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1915 Status Notification (bounce) messages are now MIME format per RFC 3464.
1916 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1917 under the control of the dsn_advertise_hosts option, and routers may
1918 have a dsn_lasthop option.
1920 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1921 default, modifiable by a malware= option. The list separator for
1922 the options can now be changed in the usual way. Bug 68.
1924 JH/10 The smtp_receive_timeout main option is now expanded before use.
1926 JH/11 The incoming_interface log option now also enables logging of the
1927 local interface on delivery outgoing connections.
1929 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1930 if the interface and destination host and port all match.
1932 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1935 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1936 Patch from Andrew Lewis.
1938 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1939 now supports optional time-restrictions, weighting, and priority
1940 modifiers per server. Patch originally by <rommer@active.by>.
1942 JH/16 The spamd_address main option now supports a mixed list of local
1943 and remote servers. Remote servers can be IPv6 addresses, and
1944 specify a port-range.
1946 JH/17 Bug 68: The spamd_address main option now supports an optional
1947 timeout value per server.
1949 JH/18 Bug 1581: Router and transport options headers_add/remove can
1950 now have the list separator specified.
1952 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1955 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1958 JH/21 Support for the A6 type of dns record is withdrawn.
1960 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1961 rather than the verbs used.
1963 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1964 from 255 to 1024 chars.
1966 JH/24 Verification callouts now attempt to use TLS by default.
1968 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1969 are generic router options now. The defaults didn't change.
1971 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1972 Original patch from Alexander Shikoff, worked over by JH.
1974 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1975 files if ZCAT_COMMAND is not executable.
1977 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1979 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1981 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1982 Normally benign, it bites when the pair was led to by a CNAME;
1983 modern usage is to not canonicalize the domain to a CNAME target
1984 (and we were inconsistent anyway for A-only vs AAAA+A).
1986 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1988 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1989 when evaluating $sender_host_dnssec.
1991 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1992 $sender_helo_dnssec variable.
1994 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1996 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1998 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2000 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2001 documented as working, but never had. Support all but $spam_report.
2003 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2004 added for tls authenticator.
2006 HS/03 Add perl_taintmode main config option
2012 TL/01 When running the test suite, the README says that variables such as
2013 no_msglog_check are global and can be placed anywhere in a specific
2014 test's script, however it was observed that placement needed to be near
2015 the beginning for it to behave that way. Changed the runtest perl
2016 script to read through the entire script once to detect and set these
2017 variables, reset to the beginning of the script, and then run through
2018 the script parsing/test process like normal.
2020 TL/02 The BSD's have an arc4random API. One of the functions to induce
2021 adding randomness was arc4random_stir(), but it has been removed in
2022 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2023 function when detected.
2025 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2026 cause callback expansion.
2028 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2029 syntax errors in an expansion can be treated as a string instead of
2030 logging or causing an error, due to the internal use of bool_lax
2031 instead of bool when processing it.
2033 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2034 server certificates when making smtp deliveries.
2036 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2038 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2040 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2042 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2043 Merged patch from Sebastian Wiedenroth.
2045 JH/05 Fix results-pipe from transport process. Several recipients, combined
2046 with certificate use, exposed issues where response data items split
2047 over buffer boundaries were not parsed properly. This eventually
2048 resulted in duplicates being sent. This issue only became common enough
2049 to notice due to the introduction of connection certificate information,
2050 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2052 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2053 size buffer was used, resulting in syntax errors when an expansion
2056 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2057 version 3.3.6 or later.
2059 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2060 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2061 both become "event_action", the variables become $event_name, $event_data
2062 and $event_defer_errno. There is a new variable $verify_mode, usable in
2063 routers, transports and related events. The tls:cert event is now also
2064 raised for inbound connections, if the main configuration event_action
2067 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2068 early OCSP support, but no stapling (appears to be less than 1.0.0).
2070 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2071 server certificate names available under the smtp transport option
2072 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2075 JH/10 Time-related extraction expansions from certificates now use the main
2076 option "timezone" setting for output formatting, and are consistent
2077 between OpenSSL and GnuTLS compilations. Bug 1541.
2079 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2080 encoded parameter in the incoming message. Bug 1558.
2082 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2083 include certificate info, eximon was claiming there were spoolfile
2086 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2088 JH/14 Log delivery-related information more consistently, using the sequence
2089 "H=<name> [<ip>]" wherever possible.
2091 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2092 are problematic for Debian distribution, omit them from the release
2095 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2097 JH/16 Fix string representation of time values on 64bit time_t architectures.
2100 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2101 output list separator was used.
2106 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2107 checkers that were complaining about end of non-void function with no
2110 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2111 This was a regression introduced in 4.83 by another bugfix.
2113 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2115 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2116 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2122 TF/01 Correctly close the server side of TLS when forking for delivery.
2124 When a message was received over SMTP with TLS, Exim failed to clear up
2125 the incoming connection properly after forking off the child process to
2126 deliver the message. In some situations the subsequent outgoing
2127 delivery connection happened to have the same fd number as the incoming
2128 connection previously had. Exim would try to use TLS and fail, logging
2129 a "Bad file descriptor" error.
2131 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2132 utilities have not been installed.
2134 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2135 temporary space as the ACL may create new global variables.
2137 TL/01 LDAP support uses per connection or global context settings, depending
2138 upon the detected version of the libraries at build time.
2140 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2141 to extract and use the src ip:port in logging and expansions as if it
2142 were a direct connection from the outside internet. PPv2 support was
2143 updated based on HAProxy spec change in May 2014.
2145 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2147 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2148 Properly escape header and check for NULL return.
2150 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2153 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2155 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2156 characters in header names, implemented as a verify condition.
2157 Contributed by Michael Fischer v. Mollard.
2159 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2160 results permerror and temperror. Previous values are deprecated but
2161 still accepted. In a future release, err_perm and err_temp will be
2162 completely removed, which will be a backward incompatibility if the
2163 ACL tests for either of these two old results. Patch contributed by
2164 user bes-internal on the mailing list.
2166 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2168 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2169 selectors, in both main and reject logs.
2171 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2174 JH/07 Add malware type "sock" for talking to simple daemon.
2176 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2178 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2179 routers/transports under cutthrough routing.
2181 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2182 numbers. Touch up "bool" conditional to keep the same definition.
2184 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2186 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2188 JH/12 Expand items in router/transport headers_add or headers_remove lists
2189 individually rather than the list as a whole. Bug 1452.
2191 Required for reasonable handling of multiple headers_ options when
2192 they may be empty; requires that headers_remove items with embedded
2193 colons must have them doubled (or the list-separator changed).
2195 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2196 view the policy declared in the DMARC record. Currently, $dmarc_status
2197 is a combined value of both the record presence and the result of the
2200 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2202 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2203 dnslookup router and the smtp transport (applying to the forward
2206 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2207 of ldap servers used for a specific lookup. Patch provided by Heiko
2210 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2211 New variable $lookup_dnssec_authenticated for observability.
2213 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2214 Patch submitted by Lars Timman.
2216 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2218 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2219 Requires trusted mode and valid format message id, aborts otherwise.
2220 Patch contributed by Heiko Schlichting.
2222 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2223 certextract with support for various fields. Bug 1358.
2225 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2226 is requested by default, modifiable by smtp transport option
2229 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2230 operate on certificate variables to give certificate fingerprints
2231 Also new ${sha256:cert_variable}.
2233 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2235 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2236 Christian Aistleitner.
2238 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2240 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2241 file. Patch from Wolfgang Breyha.
2243 JH/25 Expand the coverage of the delivery $host and $host_address to
2244 client authenticators run in verify callout. Bug 1476.
2246 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2247 align with daemon_smtp_ports. Bug 72.
2249 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2250 support and error reporting did not work properly.
2252 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2253 and is readable. Patch from Andrew Colin Kissa.
2255 TL/14 Enhance documentation of ${run expansion and how it parses the
2256 commandline after expansion, particularly in the case when an
2257 unquoted variable expansion results in an empty value.
2259 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2261 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2262 by renaming away from it. Addresses GH issue 15, reported by
2265 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2267 TL/15 SECURITY: prevent double expansion in math comparison functions
2268 (can expand unsanitized data). Not remotely exploitable.
2275 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2277 PP/02 Make -n do something, by making it not do something.
2278 When combined with -bP, the name of an option is not output.
2280 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2283 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2284 $sender_host_name and config options to manage this, and basic check
2287 PP/05 DSCP support for outbound connections and control modifier for inbound.
2289 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2290 (Only plugin which currently uses this is kerberos4, which nobody should
2291 be using, but we should make it available and other future plugins might
2292 conceivably use it, even though it would break NAT; stuff *should* be
2293 using channel bindings instead).
2295 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2296 name; added for Sendmail compatibility; requires admin caller.
2297 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2298 just ignore it); requires trusted caller.
2299 Also parse but ignore: -Ac -Am -X<logfile>
2302 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2304 TL/02 Add +smtp_confirmation as a default logging option.
2306 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2307 Patch by Magnus Holmgren from 2007-02-20.
2309 TL/04 Bugzilla 1281 - Spec typo.
2310 Bugzilla 1283 - Spec typo.
2311 Bugzilla 1290 - Spec grammar fixes.
2313 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2315 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2317 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2318 mailing list by by Dmitry Isaikin.
2320 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2322 JH/02 Support "G" suffix to numbers in ${if comparisons.
2324 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2326 NM/01 Bugzilla 1197 - Spec typo
2327 Bugzilla 1196 - Spec examples corrections
2329 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2331 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2332 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2335 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2336 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2338 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2339 "acl {{name}{arg}...}", and optional args on acl condition
2342 JH/05 Permit multiple router/transport headers_add/remove lines.
2344 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2346 JH/07 Avoid using a waiting database for a single-message-only transport.
2347 Performance patch from Paul Fisher. Bugzilla 1262.
2349 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2352 JH/09 Add $headers_added variable, with content from use of ACL modifier
2353 add_header (but not yet added to the message). Bugzilla 199.
2355 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2356 Pulled from Bugzilla 817 by Wolfgang Breyha.
2358 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2360 (nb: this is the same fix as in Exim 4.80.1)
2362 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2365 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2368 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2370 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2371 advertise SMTP AUTH mechanism to us, instead of a generic
2372 protocol violation error. Also, make Exim more robust to bad
2373 data from the Dovecot auth socket.
2375 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2377 When a queue runner is handling a message, Exim first routes the
2378 recipient addresses, during which it prunes them based on the retry
2379 hints database. After that it attempts to deliver the message to
2380 any remaining recipients. It then updates the hints database using
2383 So if a recipient address works intermittently, it can get repeatedly
2384 deferred at routing time. The retry hints record remains fresh so the
2385 address never reaches the final cutoff time.
2387 This is a fairly common occurrence when a user is bumping up against
2388 their storage quota. Exim had some logic in its local delivery code
2389 to deal with this. However it did not apply to per-recipient defers
2390 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2392 This change adds a proper retry rule check during routing so that the
2393 final cutoff time is checked against the message's age. We only do
2394 this check if there is an address retry record and there is not a
2395 domain retry record; this implies that previous attempts to handle
2396 the address had the retry_use_local_parts option turned on. We use
2397 this as an approximation for the destination being like a local
2398 delivery, as in LMTP.
2400 I suspect this new check makes the old local delivery cutoff check
2401 redundant, but I have not verified this so I left the code in place.
2403 TF/02 Correct gecos expansion when From: is a prefix of the username.
2405 Test 0254 submits a message to Exim with the header
2409 When I ran the test suite under the user fanf2, Exim expanded
2410 the header to contain my full name, whereas it should have added
2411 a Resent-Sender: header. It erroneously treats any prefix of the
2412 username as equal to the username.
2414 This change corrects that bug.
2416 GF/01 DCC debug and logging tidyup
2417 Error conditions log to paniclog rather than rejectlog.
2418 Debug lines prefixed by "DCC: " to remove any ambiguity.
2420 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2422 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2423 Bug spotted by Jeremy Harris; was flawed since initial commit.
2424 Would have resulted in OCSP responses post-SNI triggering an Exim
2425 NULL dereference and crash.
2427 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2429 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2430 Bug detection, analysis and fix by Samuel Thibault.
2431 Bugzilla 1331, Debian bug #698092.
2433 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2435 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2436 Server implementation by Todd Lyons, client by JH.
2437 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2438 config variable "prdr_enable" controls whether the server
2439 advertises the facility. If the client requests PRDR a new
2440 acl_data_smtp_prdr ACL is called once for each recipient, after
2441 the body content is received and before the acl_smtp_data ACL.
2442 The client is controlled by both of: a hosts_try_prdr option
2443 on the smtp transport, and the server advertisement.
2444 Default client logging of deliveries and rejections involving
2445 PRDR are flagged with the string "PRDR".
2447 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2448 fclose(). Diagnosis by Todd Lyons.
2450 PP/17 Update configure.default to handle IPv6 localhost better.
2451 Patch by Alain Williams (plus minor tweaks).
2454 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2455 This is now consistent with GnuTLS, and is now documented: the
2456 previous undocumented portable approach to treating the option as
2457 unset was to force an expansion failure. That still works, and
2458 an empty string is now equivalent.
2460 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2461 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2462 not performing validation itself.
2464 PP/20 Added force_command boolean option to pipe transport.
2465 Patch from Nick Koston, of cPanel Inc.
2467 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2470 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2472 PP/21 Fix eximon continuous updating with timestamped log-files.
2473 Broken in a format-string cleanup in 4.80, missed when I repaired the
2474 other false fix of the same issue.
2475 Report and fix from Heiko Schlichting.
2478 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2479 Report from Prashanth Katuri.
2481 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2482 It's SecureTransport, so affects any MacOS clients which use the
2483 system-integrated TLS libraries, including email clients.
2485 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2486 using a MIME ACL for non-SMTP local injection.
2487 Report and assistance in diagnosis by Warren Baker.
2489 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2491 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2493 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2494 last id that failed so it may be referenced in subsequent ACL's.
2496 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2499 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2500 ldap library initialization, allowing self-signed CA's to be
2501 used. Also properly sets require_cert option later in code by
2502 using NULL (global ldap config) instead of ldap handle (per
2503 session). Bug diagnosis and testing by alxgomz.
2505 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2506 the src/util/ subdirectory.
2508 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2509 renamed to Transport Post Delivery Action by Jeremy Harris, as
2512 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2513 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2514 redis_servers = needs to be configured which will be used by the redis
2515 lookup. Patch from Warren Baker, of The Packet Hub.
2517 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2519 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2520 hostname or reverse DNS when processing a host list. Used suggestions
2521 from multiple comments on this bug.
2523 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2525 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2526 Missed a few lines, added it to make the runtest require no keyboard
2529 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2530 contains upper case chars. Make router use caseful_local_part.
2532 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2533 support when GnuTLS has been built with p11-kit.
2539 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2541 This, or similar/improved, will also be change PP/11 of 4.82.
2547 PP/01 Handle short writes when writing local log-files.
2548 In practice, only affects FreeBSD (8 onwards).
2549 Bugzilla 1053, with thanks to Dmitry Isaikin.
2551 NM/01 Bugzilla 949 - Documentation tweak
2553 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2556 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2558 PP/02 Implemented gsasl authenticator.
2560 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2562 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2563 `pkg-config foo` for cflags/libs.
2565 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2566 with rest of GSASL and with heimdal_gssapi.
2568 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2569 `pkg-config foo` for cflags/libs for the TLS implementation.
2571 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2572 properties get this fed in as external SSF. A number of robustness
2573 and debugging improvements to the cyrus_sasl authenticator.
2575 PP/08 cyrus_sasl server now expands the server_realm option.
2577 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2578 Patch by Jeremy Harris.
2580 PP/10 Added dbmjz lookup type.
2582 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2584 PP/12 MAIL args handles TAB as well as SP, for better interop with
2585 non-compliant senders.
2586 Analysis and variant patch by Todd Lyons.
2588 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2589 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2590 Patch from Dirk Mueller <dmueller@suse.com>
2592 PP/13 tls_peerdn now print-escaped for spool files.
2593 Observed some $tls_peerdn in wild which contained \n, which resulted
2594 in spool file corruption.
2596 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2597 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2598 or write after TLS renegotiation, which otherwise led to messages
2601 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2602 as a tracking header (ie: a signed header comes before the signature).
2603 Patch from Wolfgang Breyha.
2605 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2606 comma-sep list; embedded commas doubled.
2608 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2610 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2612 Report and patch from Dmitry Banschikov.
2614 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2615 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2616 failures. We appear to now support TLS1.1+ with Exim.
2618 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2619 lets Exim select keys and certificates based upon TLS SNI from client.
2620 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2621 before an outbound SMTP session. New log_selector, +tls_sni.
2623 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2624 NULL dereference. Report and patch from Alun Jones.
2626 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2627 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2628 Not seeing resolver debug output on NetBSD, but suspect this is a
2629 resolver implementation change.
2631 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2632 Left warnings. Added "eximon gdb" invocation mode.
2634 PP/21 Defaulting "accept_8bitmime" to true, not false.
2636 PP/22 Added -bw for inetd wait mode support.
2638 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2639 locate the relevant includes and libraries. Made this the default.
2641 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2642 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2644 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2645 This may cause build issues on older platforms.
2647 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2648 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2649 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2650 Added SNI support via GnuTLS too.
2651 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2653 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2655 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2656 Applied second patch from Janne, implementing suggestion to default
2657 multiple-strings-in-record handling to match SPF spec.
2659 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2661 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2662 read-only, out of scope).
2663 Patch from Wolfgang Breyha, report from Stuart Northfield.
2665 PP/29 Fix three issues highlighted by clang analyser static analysis.
2666 Only crash-plausible issue would require the Cambridge-specific
2667 iplookup router and a misconfiguration.
2668 Report from Marcin Mirosław.
2670 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2672 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2673 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2674 As part of this, removing so much warning spew let me fix some minor
2675 real issues in debug logging.
2677 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2678 assignment on my part. Fixed.
2680 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2681 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2682 Janne Snabb (who went above and beyond: thank you).
2684 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2685 string otherwise requires a connection and a bunch more work and it's
2686 relatively easy to get wrong. Should also expose TLS library linkage
2689 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2690 64-bit ${eval} (JH/03).
2692 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2693 GNU libc to support some of the 64-bit stuff, should not lead to
2694 conflicts. Defined before os.h is pulled in, so if a given platform
2695 needs to override this, it can.
2697 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2698 protection layer was required, which is not implemented.
2699 Bugzilla 1254, patch from Wolfgang Breyha.
2701 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2702 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2703 tls_dhparam take prime identifiers. Also unbreak combination of
2704 OpenSSL+DH_params+TLSSNI.
2706 PP/39 Disable SSLv2 by default in OpenSSL support.
2712 PP/01 Solaris build fix for Oracle's LDAP libraries.
2713 Bugzilla 1109, patch from Stephen Usher.
2715 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2717 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2720 TF/02 Fix a couple more cases where we did not log the error message
2721 when unlink() failed. See also change 4.74-TF/03.
2723 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2724 lock up or crash if it happened to be inside a call to libc when it
2725 got a SIGUSR1 from exiwhat.
2727 The SIGUSR1 handler appends the current process status to the process
2728 log which is later printed by exiwhat. It used to use the general
2729 purpose logging code to do this, but several functions it calls are
2730 not safe for signals.
2732 The new output code in the SIGUSR1 handler is specific to the process
2733 log, and simple enough that it's easy to inspect for signal safety.
2734 Removing some special cases also simplifies the general logging code.
2735 Removing the spurious timestamps from the process log simplifies
2738 TF/04 Improved ratelimit ACL condition.
2740 The /noupdate option has been deprecated in favour of /readonly which
2741 has clearer semantics. The /leaky, /strict, and /readonly update modes
2742 are mutually exclusive. The update mode is no longer included in the
2743 database key; it just determines when the database is updated. (This
2744 means that when you upgrade Exim will forget old rate measurements.)
2746 Exim now checks that the per_* options are used with an update mode that
2747 makes sense for the current ACL. For example, when Exim is processing a
2748 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2749 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2750 must specify per_mail/readonly. If you omit the update mode it defaults to
2751 /leaky where that makes sense (as before) or /readonly where required.
2753 The /noupdate option is now undocumented but still supported for
2754 backwards compatibility. It is equivalent to /readonly except that in
2755 ACLs where /readonly is required you may specify /leaky/noupdate or
2756 /strict/noupdate which are treated the same as /readonly.
2758 A useful new feature is the /count= option. This is a generalization
2759 of the per_byte option, so that you can measure the throughput of other
2760 aggregate values. For example, the per_byte option is now equivalent
2761 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2763 The per_rcpt option has been generalized using the /count= mechanism
2764 (though it's more complicated than the per_byte equivalence). When it is
2765 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2766 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2767 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2768 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2769 Note that using per_rcpt with a non-readonly update mode in more than
2770 one ACL will cause the recipients to be double-counted. (The per_mail
2771 and per_byte options don't have this problem.)
2773 The handling of very low rates has changed slightly. If the computed rate
2774 is less than the event's count (usually one) then this event is the first
2775 after a long gap. In this case the rate is set to the same as this event's
2776 count, so that the first message of a spam run is counted properly.
2778 The major new feature is a mechanism for counting the rate of unique
2779 events. The new per_addr option counts the number of different
2780 recipients that someone has sent messages to in the last time period. It
2781 behaves like per_rcpt if all the recipient addresses are different, but
2782 duplicate recipient addresses do not increase the measured rate. Like
2783 the /count= option this is a general mechanism, so the per_addr option
2784 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2785 example, measure the rate that a client uses different sender addresses
2786 with the options per_mail/unique=$sender_address. There are further
2787 details in the main documentation.
2789 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2791 TF/06 Removed a few PCRE remnants.
2793 TF/07 Automatically extract Exim's version number from tags in the git
2794 repository when doing development or release builds.
2796 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2797 Bugzilla 879. Patch from Paul Fisher.
2799 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2800 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2803 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2805 PP/05 Variable $av_failed, true if the AV scanner deferred.
2806 Bugzilla 1078. Patch from John Horne.
2808 PP/06 Stop make process more reliably on build failure.
2809 Bugzilla 1087. Patch from Heiko Schlittermann.
2811 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2812 Bugzilla 1089. Patch from Heiko Schlittermann.
2814 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2815 Bugzilla 1131. Patch from Holger Weiß.
2817 PP/09 Handle IPv6 addresses with SPF.
2818 Bugzilla 860. Patch from Wolfgang Breyha.
2820 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2822 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2825 PP/11 match_* no longer expand right-hand-side by default.
2826 New compile-time build option, EXPAND_LISTMATCH_RHS.
2827 New expansion conditions, "inlist", "inlisti".
2829 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2831 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2833 PP/14 fix log_write() format string regression from TF/03.
2834 Bugzilla 1152. Patch from Dmitry Isaikin.
2840 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2842 PP/02 Harmonised TLS library version reporting; only show if debugging.
2843 Layout now matches that introduced for other libraries in 4.74 PP/03.
2845 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2847 PP/04 New "dns_use_edns0" global option.
2849 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2852 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2853 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2855 TK/01 Updated PolarSSL code to 0.14.2.
2856 Bugzilla 1097. Patch from Andreas Metzler.
2858 PP/07 Catch divide-by-zero in ${eval:...}.
2859 Fixes bugzilla 1102.
2861 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2864 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2865 format-string attack -- SECURITY: remote arbitrary code execution.
2867 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2868 time unintentionally subject to list matching rules, letting the header
2869 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2870 arbitrary string expansion). This allowed for information disclosure.
2872 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2873 INT_MIN/-1 -- value coerced to INT_MAX.
2879 NM/01 Workaround for PCRE version dependency in version reporting
2882 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2883 This fixes portability to compilers other than gcc, notably
2884 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2886 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2887 makefiles for portability to HP-UX and POSIX correctness.
2889 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2890 Also via indented variable definition in the Makefile.
2891 (Debugging by Oliver Heesakkers).
2893 PP/02 Restore caching of spamd results with expanded spamd_address.
2894 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2896 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2897 Improves build reliability. Fix from: Frank Elsner
2899 NM/02 Fix wide character breakage in the rfc2047 coding
2900 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2902 NM/03 Allow underscore in dnslist lookups
2903 Fixes bug 1026. Patch from Graeme Fowler
2905 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2906 Code patches from Adam Ciarcinski of NetBSD.
2908 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2911 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2912 is logged, to avoid truncation. Patch from John Horne.
2914 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2915 Patch from Jakob Hirsch.
2917 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2918 SQL string expansion failure details.
2919 Patch from Andrey Oktyabrski.
2921 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2922 Patch from Simon Arlott.
2924 PP/09 New lookups functionality failed to compile on old gcc which rejects
2925 extern declarations in function scope.
2926 Patch from Oliver Fleischmann
2928 PP/10 Use sig_atomic_t for flags set from signal handlers.
2929 Check getgroups() return and improve debugging.
2930 Fixed developed for diagnosis in bug 927 (which turned out to be
2933 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2934 Patch from Mark Zealey.
2936 PP/12 Bugzilla 1056: Improved spamd server selection.
2937 Patch from Mark Zealey.
2939 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2940 Based on patch from Heiko Schlittermann.
2942 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2943 Patch from Uwe Doering, sign-off by Michael Haardt.
2945 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2946 variable declaration deep within a block. Bug and patch from
2949 PP/15 lookups-Makefile IRIX compatibility coercion.
2951 PP/16 Make DISABLE_DKIM build knob functional.
2953 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2954 Patch by Simon Arlott
2956 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2957 variable argument macros. Our copy now differs from upstream.
2963 TF/01 Failure to get a lock on a hints database can have serious
2964 consequences so log it to the panic log.
2966 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2967 controlled using the smtp_confirmation log selector.
2969 TF/03 Include the error message when we fail to unlink a spool file.
2971 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2972 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2973 for maintaining out-of-tree patches for some time.
2975 PP/01 Bugzilla 139: Documentation and portability issues.
2976 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2977 Handle per-OS dynamic-module compilation flags.
2979 PP/02 Let /dev/null have normal permissions.
2980 The 4.73 fixes were a little too stringent and complained about the
2981 permissions on /dev/null. Exempt it from some checks.
2982 Reported by Andreas M. Kirchwitz.
2984 PP/03 Report version information for many libraries, including
2985 Exim version information for dynamically loaded libraries. Created
2986 version.h, now support a version extension string for distributors
2987 who patch heavily. Dynamic module ABI change.
2989 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2990 privilege escalation vulnerability whereby the Exim run-time user
2991 can cause root to append content of the attacker's choosing to
2994 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2997 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2998 If dropping privileges for untrusted macros, we disabled normal logging
2999 on the basis that it would fail; for the Exim run-time user, this is not
3000 the case, and it resulted in successful deliveries going unlogged.
3001 Fixed. Reported by Andreas Metzler.
3007 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3008 only prepend for the Resent-* case. Fixes regression introduced in
3009 Exim 4.70 by NM/22 for Bugzilla 607.
3011 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3012 increasing numbers of administrators be bitten by this.
3014 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3016 PP/03 Bugzilla 994: added openssl_options main configuration option.
3018 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3020 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3022 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3024 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3025 without a peer certificate, leading to a segfault because of an
3026 assumption that peers always have certificates. Be a little more
3027 paranoid. Problem reported by Martin Tscholak.
3029 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3030 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3031 NB: ClamAV planning to remove STREAM in "middle of 2010".
3032 CL also introduces -bmalware, various -d+acl logging additions and
3033 more caution in buffer sizes.
3035 PP/09 Implemented reverse_ip expansion operator.
3037 PP/10 Bugzilla 937: provide a "debug" ACL control.
3039 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3041 PP/12 Bugzilla 973: Implement --version.
3043 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3045 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3047 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3049 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3050 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3051 ignore trailing whitespace.
3053 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3055 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3058 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3059 Notification from Dr Andrew Aitchison.
3061 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3062 ExtendedDetectionInfo response format.
3063 Notification from John Horne.
3065 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3068 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3069 XSL and documented dependency on system catalogs, with examples of how
3072 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3075 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3076 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3077 configuration file which is writeable by the Exim user or group.
3079 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3080 of configuration files to cover files specified with the -C option if
3081 they are going to be used with root privileges, not just the default
3084 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3085 option (effectively making it always true).
3087 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3088 files to be used while preserving root privileges.
3090 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3091 that rogue child processes cannot use them.
3093 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3094 run-time user, instead of root.
3096 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3097 Exim run-time user without dropping privileges.
3099 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3100 result string, instead of calling string_vformat() twice with the same
3103 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3104 for other users. Others should always drop root privileges if they use
3105 -C on the command line, even for a whitelisted configure file.
3107 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3109 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3115 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3116 $data_path, and $header_path variables; fixed documentation bugs and
3119 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3120 exipick to access non-standard spools, including the "frozen" queue
3123 NM/01 Bugzilla 965: Support mysql stored procedures.
3124 Patch from Alain Williams
3126 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3128 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3129 Patch from Andreas Metzler
3131 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3132 Patch from Kirill Miazine
3134 NM/05 Bugzilla 671: Added umask to procmail example.
3136 JJ/03 installed exipick 20100323.0, fixing doc bug
3138 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3139 directory. Notification and patch from Dan Rosenberg.
3141 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3143 TK/02 Improve log output when DKIM signing operation fails.
3145 MH/01 Treat the transport option dkim_domain as a colon separated
3146 list, not as a single string, and sign the message with each element,
3147 omitting multiple occurences of the same signer.
3149 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3150 Bugzilla 985, 986. Patch by Simon Arlott
3152 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3153 Patch by Simon Arlott
3155 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3156 MBX locking. Notification from Dan Rosenberg.
3162 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3164 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3166 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3168 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3170 NM/04 Bugzilla 915: Flag broken perl installation during build.
3176 TK/01 Added patch by Johannes Berg that expands the main option
3177 "spamd_address" if it starts with a dollar sign.
3179 TK/02 Write list of recipients to X-Envelope-Sender header when building
3180 the mbox-format spool file for content scanning (suggested by Jakob
3183 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3184 (http://www.dcc-servers.net/) support via dccifd. Activated by
3185 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3187 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3188 by Mark Daniel Reidel <mr@df.eu>.
3190 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3191 When building exim an external PCRE library is now needed -
3192 PCRE is a system library on the majority of modern systems.
3193 See entry on PCRE_LIBS in EDITME file.
3195 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3196 conversation. Added nologin parameter to request.
3197 Patch contributed by Kirill Miazine.
3199 TF/01 Do not log submission mode rewrites if they do not change the address.
3201 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3203 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3204 log files in place. Contributed by Roberto Lima.
3206 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3208 TF/03 Bugzilla 615: When checking the local_parts router precondition
3209 after a local_part_suffix or local_part_prefix option, Exim now
3210 does not use the address's named list lookup cache, since this
3211 contains cached lookups for the whole local part.
3213 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3214 Robert Millan. Documentation is in experimental-spec.txt.
3216 TF/04 Bugzilla 668: Fix parallel build (make -j).
3218 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3220 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3221 Patch provided by Jan Srzednicki.
3223 TF/05 Leading white space used to be stripped from $spam_report which
3224 wrecked the formatting. Now it is preserved.
3226 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3227 that they are available at delivery time.
3229 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3231 TF/08 TLS error reporting now respects the incoming_interface and
3232 incoming_port log selectors.
3234 TF/09 Produce a more useful error message if an SMTP transport's hosts
3235 setting expands to an empty string.
3237 NM/06 Bugzilla 744: EXPN did not work under TLS.
3238 Patch provided by Phil Pennock.
3240 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3241 Patch provided by Richard Godbee.
3243 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3244 acl_smtp_notquit, added index entry.
3246 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3247 Patch provided by Eugene Bujak.
3249 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3250 accept(). Patch provided by Maxim Dounin.
3252 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3253 Patch provided by Phil Pennock.
3255 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3257 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3258 Patch provided by Brad "anomie" Jorsch.
3260 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3261 Patch provided by Dean Brooks.
3263 TK/05 Add native DKIM support (does not depend on external libraries).
3265 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3266 Patch provided by Graeme Fowler.
3268 NM/16 Bugzilla 851: Documentation example syntax fix.
3270 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3272 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3275 NM/19 Bugzilla 745: TLS version reporting.
3276 Patch provided by Phil Pennock.
3278 NM/20 Bugzilla 167: bool: condition support.
3279 Patch provided by Phil Pennock.
3281 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3282 clients. Patch provided by Phil Pennock.
3284 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3285 Patch provided by Brad "anomie" Jorsch.
3287 NM/23 Bugzilla 687: Fix misparses in eximstats.
3288 Patch provided by Heiko Schlittermann.
3290 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3291 Patch provided by Heiko Schlittermann.
3293 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3294 plus update to original patch.
3296 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3298 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3299 Patch provided by David Brownlee.
3301 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3303 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3305 NM/30 Bugzilla 888: TLS documentation bugfixes.
3307 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3309 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3310 Unlike the original bugzilla I have changed all shell scripts in src tree.
3312 NM/33 Bugzilla 898: Transport filter timeout fix.
3313 Patch by Todd Rinaldo.
3315 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3316 Patch by Serge Demonchaux.
3318 NM/35 Bugzilla 39: Base64 decode bug fixes.
3319 Patch by Jakob Hirsch.
3321 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3323 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3325 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3327 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3333 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3334 ALT-N's libdkim that I have put here:
3335 http://duncanthrax.net/exim-experimental/
3337 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3338 were called 'true' and it seems that C99 defines that as a reserved
3339 keyword to be used with 'bool' variable types. That means you could
3340 not include C99-style headers which use bools without triggering
3341 build errors in sieve.c.
3343 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3344 as mailq or other aliases. Changed the --help handling significantly
3345 to do whats expected. exim_usage() emits usage/help information.
3347 SC/01 Added the -bylocaldomain option to eximstats.
3349 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3351 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3353 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3359 PH/01 Another patch from the Sieve maintainer.
3361 PH/02 When an IPv6 address is converted to a string for single-key lookup
3362 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3363 dots are used instead of colons so that keys in lsearch files need not
3364 contain colons. This was done some time before quoting was made available
3365 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3366 (notated using the quote facility) so as to distinguish them from IPv4
3367 keys. This meant that lookups for IP addresses in host lists did not work
3368 for iplsearch lookups.
3370 This has been fixed by arranging for IPv6 addresses to be expressed with
3371 colons if the lookup type is iplsearch. This is not incompatible, because
3372 previously such lookups could never work.
3374 The situation is now rather anomalous, since one *can* have colons in
3375 ordinary lsearch keys. However, making the change in all cases is
3376 incompatible and would probably break a number of configurations.
3378 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3381 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3382 conversion specification without a maximum field width, thereby enabling
3383 a rogue spamd server to cause a buffer overflow. While nobody in their
3384 right mind would setup Exim to query an untrusted spamd server, an
3385 attacker that gains access to a server running spamd could potentially
3386 exploit this vulnerability to run arbitrary code as the Exim user.
3388 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3389 $primary_hostname instead of what libspf2 thinks the hosts name is.
3391 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3392 a directory entry by the name of the lookup key. Previously, if a
3393 symlink pointed to a non-existing file or a file in a directory that
3394 Exim lacked permissions to read, a lookup for a key matching that
3395 symlink would fail. Now it is enough that a matching directory entry
3396 exists, symlink or not. (Bugzilla 503.)
3398 PH/03 The body_linecount and body_zerocount variables are now exported in the
3401 PH/04 Added the $dnslist_matched variable.
3403 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3404 This means they are set thereafter only if the connection becomes
3407 PH/06 Added the client_condition to authenticators so that some can be skipped
3408 by clients under certain conditions.
3410 PH/07 The error message for a badly-placed control=no_multiline_responses left
3411 "_responses" off the end of the name.
3413 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3415 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3416 (without spaces) instead of just copying the configuration text.
3418 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3420 PH/11 Added $max_received_linelength.
3422 PH/12 Added +ignore_defer and +include_defer to host lists.
3424 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3425 way in which PCRE > 7.0 is built.
3427 PH/14 Implemented queue_only_load_latch.
3429 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3430 MAIL command. The effect was to mangle the value on 64-bit systems.
3432 PH/16 Another patch from the Sieve maintainer.
3434 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3436 PH/18 If a system quota error occurred while trying to create the file for
3437 a maildir delivery, the message "Mailbox is full" was not appended to the
3438 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3439 only to a quota excession during the actual writing of the file.
3441 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3442 characters?) which causes problems in log lines. The DN values are now
3443 passed through string_printing() before being added to log lines.
3445 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3446 and InterBase are left for another time.)
3448 PH/21 Added message_body_newlines option.
3450 PH/22 Guard against possible overflow in moan_check_errorcopy().
3452 PH/23 POSIX allows open() to be a macro; guard against that.
3454 PH/24 If the recipient of an error message contained an @ in the local part
3455 (suitably quoted, of course), incorrect values were put in $domain and
3456 $local_part during the evaluation of errors_copy.
3462 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3463 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3466 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3467 issue a MAIL command.
3469 PH/02 In an ACL statement such as
3471 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3473 if a client was not listed at all, or was listed with a value other than
3474 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3475 the condition was not true (as it should be), so access was not denied.
3476 The bug was that the ! inversion was incorrectly passed on to the second
3477 item. This has been fixed.
3479 PH/03 Added additional dnslists conditions == and =& which are different from
3480 = and & when the dns lookup returns more than one IP address.
3482 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3483 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3485 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3486 FSYNC, which compiles an option called disable_fsync that allows for
3487 bypassing fsync(). The documentation is heavily laced with warnings.
3489 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3491 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3492 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3493 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3494 including adding "make clean"; (3) Added -fPIC when compiling the test
3495 dynamically loaded module, to get rid of a warning.
3497 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3498 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3499 = 0s. The bug is otherwise harmless.
3501 PH/07 There was a bug in the dovecot authenticator such that the value of
3502 $auth1 could be overwritten, and so not correctly preserved, after a
3503 successful authentication. This usually meant that the value preserved by
3504 the server_setid option was incorrect.
3506 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3508 PH/09 Installed PCRE release 7.0.
3510 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3511 run for batched SMTP input. It is now run at the start of every message
3512 in the batch. While fixing this I discovered that the process information
3513 (output by running exiwhat) was not always getting set for -bs and -bS
3514 input. This is fixed, and it now also says "batched" for BSMTP.
3516 PH/11 Added control=no_pipelining.
3518 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3519 patch, slightly modified), and move the expansion of helo_data till after
3520 the connection is made in the smtp transport (so it can use these
3523 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3525 PH/14 Added log_selector = +pid.
3527 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3529 PH/16 Add ${if forany and ${if forall.
3531 PH/17 Added dsn_from option to vary the From: line in DSNs.
3533 PH/18 Flush SMTP output before performing a callout, unless control =
3534 no_callout_flush is set.
3536 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3537 was true (the default) a successful delivery failed to delete the retry
3538 item, thus causing premature timeout of the address. The bug is now
3541 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3543 PH/21 Long custom messages for fakedefer and fakereject are now split up
3544 into multiline responses in the same way that messages for "deny" and
3545 other ACL rejections are.
3547 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3548 with slight modification.
3550 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3551 draft, changing the syntax and factoring some duplicate code".
3553 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3554 for deliveries of the second and subsequent messages over the same SMTP
3557 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3558 ${reduce, with only minor "tidies".
3560 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3562 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3563 expansion side effects.
3565 PH/27 When a message times out after an over-quota error from an Exim-imposed
3566 quota, the bounce message says "mailbox is full". This message was not
3567 being given when it was a system quota that was exceeded. It now should
3570 MH/03 Made $recipients available in local_scan(). local_scan() already has
3571 better access to the recipient list through recipients_list[], but
3572 $recipients can be useful in postmaster-provided expansion strings.
3574 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3575 in the case of a MAIL command with additional options following the
3576 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3577 were accidentally chopped off.
3579 PH/29 SMTP synchronization checks are implemented when a command is read -
3580 there is a check that no more input is waiting when there shouldn't be
3581 any. However, for some commands, a delay in an ACL can mean that it is
3582 some time before the response is written. In this time, more input might
3583 arrive, invalidly. So now there are extra checks after an ACL has run for
3584 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3585 pipelining has not been advertised.
3587 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3589 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3590 returned in the SMTP response when smtp_return_error_details was false.
3591 This has been fixed.
3593 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3594 instead of the C I/O that was originally supplied, because problems were
3595 reported on Solaris.
3597 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3598 Exim which did not show up earlier: it was assuming that a call to
3599 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3600 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3601 was a macro that became an assignment, so it seemed to work. This has
3602 changed to a proper function call with a void return, hence the compile
3603 error. Exim's code has been fixed.
3605 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3608 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3610 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3612 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3613 because it turns out to be too big or there is a timeout) there is no
3614 "Completed" line in the log. When some messages of this type were
3615 selected by exigrep, they were listed as "not completed". Others were
3616 picked up by some special patterns. I have improved the selection
3617 criteria to be more general.
3619 PH/38 The host_find_failed option in the manualroute router can now be set
3620 to "ignore", to completely ignore a host whose IP address cannot be
3621 found. If all hosts are ignored, the behaviour is controlled by the new
3622 host_all_ignored option.
3624 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3625 homing or because of multiple MX records with /mx) generated more than
3626 one IP address, and the following item turned out to be the local host,
3627 all the secondary addresses of the first item were incorrectly removed
3628 from the list, along with the local host and any following hosts (which
3629 is what is supposed to happen).
3631 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3632 whoever called Exim into the -H file. In the case of the daemon it was
3633 behaving confusingly. When first started, it used values for whoever
3634 started the daemon, but after a SIGHUP it used the Exim user (because it
3635 calls itself on a restart). I have changed the code so that it now always
3638 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3639 message are rejected with the same error (e.g. no authentication or bad
3640 sender address), and a DATA command is nevertheless sent (as can happen
3641 with PIPELINING or a stupid MUA), the error message that was given to the
3642 RCPT commands is included in the rejection of the DATA command. This is
3643 intended to be helpful for MUAs that show only the final error to their
3646 PH/42 Another patch from the Sieve maintainer.
3648 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3649 Eximstats - Fixed some broken HTML links and added missing column headers
3651 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3652 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3654 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3656 PH/43 Yet another patch from the Sieve maintainer.
3658 PH/44 I found a way to check for a TCP/IP connection going away before sending
3659 the response to the final '.' that terminates a message, but only in the
3660 case where the client has not sent further data following the '.'
3661 (unfortunately, this is allowed). However, in many cases there won't be
3662 any further data because there won't be any more messages to send. A call
3663 to select() can be used: if it shows that the input is "ready", there is
3664 either input waiting, or the socket has been closed. An attempt to read
3665 the next input character can distinguish the two cases. Previously, Exim
3666 would have sent an OK response which the client would never have see.
3667 This could lead to message repetition. This fix should cure that, at
3668 least in a lot of common cases.
3670 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3671 advertised in response to EHLO.
3677 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3678 fixed by 4.65/MH/01 (is this a record?) are fixed:
3680 (i) An empty string was always treated as zero by the numeric comparison
3681 operators. This behaviour has been restored.
3683 (ii) It is documented that the numeric comparison operators always treat
3684 their arguments as decimal numbers. This was broken in that numbers
3685 starting with 0 were being interpreted as octal.
3687 While fixing these problems I realized that there was another issue that
3688 hadn't been noticed. Values of message_size_limit (both the global option
3689 and the transport option) were treated as octal if they started with 0.
3690 The documentation was vague. These values are now always treated as
3691 decimal, and I will make that clear in the documentation.
3697 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3698 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3701 MH/01 Don't check that the operands of numeric comparison operators are
3702 integers when their expansion is in "skipping" mode (fixes bug
3703 introduced by 4.64-PH/07).
3705 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3706 child addresses, Exim now panics and dies. Previously, because the count
3707 is held in a short int, deliveries were likely to be lost. As such a
3708 large number of recipients for a single message is ridiculous
3709 (performance will be very, very poor), I have chosen to impose a limit
3710 rather than extend the field.
3716 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3717 leftover -K file (the existence of which was triggered by #402).
3718 While we were at it, introduced process PID as part of the -K
3719 filename. This should rule out race conditions when creating
3722 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3723 processing considerably. Previous code took too long for large mails,
3724 triggering a timeout which in turn triggers #401.
3726 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3727 in the DK code in transports.c. sendfile() is not really portable,
3728 hence the _LINUX specificness.
3730 TF/01 In the add_headers option to the mail command in an Exim filter,
3731 there was a bug that Exim would claim a syntax error in any
3732 header after the first one which had an odd number of characters
3735 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3736 callout verification, Exim cached a "reject" for the entire domain. This
3737 is correct for most verifications, but it is not correct for a recipient
3738 verification with use_sender or use_postmaster set, because in that case
3739 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3740 case of MAIL FROM:<> rejection from other early rejections (e.g.
3741 rejection of HELO). When verifying a recipient using a non-null MAIL
3742 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3743 Whatever the result of the callout, the value of the domain cache is
3744 left unchanged (for any other kind of callout, getting as far as trying
3745 RCPT means that the domain itself is ok).
3747 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3750 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3751 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3752 session, the daemon ignores ECONNECT errors and logs others; it now
3753 ignores EPIPE as well.
3755 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3756 (quoted-printable decoding).
3758 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3759 later the small subsequent patch to fix an introduced bug.
3761 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3763 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3765 PH/08 An error is now given if message_size_limit is specified negative.
3767 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3768 to be given (somewhat) arbitrary names.
3770 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3773 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3774 miscellaneous code fixes
3776 PH/10 Added the log_reject_target ACL modifier to specify where to log
3779 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3780 hostname. This is wrong, because it relates to the incoming message (and
3781 probably the interface on which it is arriving) and not to the outgoing
3782 callout (which could be using a different interface). This has been
3783 changed to use the value of the helo_data option from the smtp transport
3784 instead - this is what is used when a message is actually being sent. If
3785 there is no remote transport (possible with a router that sets up host
3786 addresses), $smtp_active_hostname is used.
3788 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3789 tweaks were necessary in order to get it to work (see also 21 below):
3790 (a) The code assumed that strncpy() returns a negative number on buffer
3791 overflow, which isn't the case. Replaced with Exim's string_format()
3793 (b) There were several signed/unsigned issues. I just did the minimum
3794 hacking in of casts. There is scope for a larger refactoring.
3795 (c) The code used strcasecmp() which is not a standard C function.
3796 Replaced with Exim's strcmpic() function.
3797 (d) The code set only $1; it now sets $auth1 as well.
3798 (e) A simple test gave the error "authentication client didn't specify
3799 service in request". It would seem that Dovecot has changed its
3800 interface. Fortunately there's a specification; I followed it and
3801 changed what the client sends and it appears to be working now.
3803 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3806 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3807 address A is aliased to B and C, where B exists and C does not. Without
3808 -v the output is "A verified" because verification stops after a
3809 successful redirection if more than one address is generated. However,
3810 with -v the child addresses are also verified. Exim was outputting "A
3811 failed to verify" and then showing the successful verification for C,
3812 with its parentage. It now outputs "B failed to verify", showing B's
3813 parentage before showing the successful verification of C.
3815 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3816 look up a TXT record in a specific list after matching in a combined
3819 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3820 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3821 they consult the DNS. I had assumed they would set it the way they
3822 wanted; and indeed my experiments on Linux seem to show that in some
3823 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3824 To be on the safe side, however, I have now made the interface to
3825 host_find_byname() similar to host_find_bydns(), with an argument
3826 containing the DNS resolver options. The host_find_byname() function now
3827 sets these options at its start, just as host_find_bydns() does. The smtp
3828 transport options dns_qualify_single and dns_search_parents are passed to
3829 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3830 of host_find_byname() use the default settings of RES_DEFNAMES
3831 (qualify_single) but not RES_DNSRCH (search_parents).
3833 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3834 spool_read_header() do less string testing, by means of a preliminary
3835 switch on the second character of optional "-foo" lines. (This is
3836 overdue, caused by the large number of possibilities that now exist.
3837 Originally there were few.) While I was there, I also converted the
3838 str(n)cmp tests so they don't re-test the leading "-" and the first
3839 character, in the hope this might squeeze out yet more improvement.
3841 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3842 flag allowing group syntax was set by the header_syntax check but not
3843 turned off, possible causing trouble later; (2) The flag was not being
3844 set at all for the header_verify test, causing "group"-style headers to
3845 be rejected. I have now set it in this case, and also caused header_
3846 verify to ignore an empty address taken from a group. While doing this, I
3847 came across some other cases where the code for allowing group syntax
3848 while scanning a header line wasn't quite right (mostly, not resetting
3849 the flag correctly in the right place). These bugs could have caused
3850 trouble for malformed header lines. I hope it is now all correct.
3852 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3853 with the "reply" argument non-NULL. The code, however (which originally
3854 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3855 but it didn't always do it. This confused somebody who was copying the
3856 code for some other use. I have removed all the tests.
3858 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3859 feature that was used to support insecure browsers during the U.S. crypto
3860 embargo. It requires special client support, and Exim is probably the
3861 only MTA that supported it -- and would never use it because real RSA is
3862 always available. This code has been removed, because it had the bad
3863 effect of slowing Exim down by computing (never used) parameters for the
3864 RSA_EXPORT functionality.
3866 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3867 authenticator to fail if there's a tab character in the incoming data
3868 (there should never be unless someone is messing about, as it's supposed
3869 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3870 if the connection is using TLS or if the remote IP is the same as the
3871 local IP, and the "valid-client-cert option" if a client certificate has
3874 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3875 authenticators. This can be used for authorization after authentication
3876 succeeds. (In the case of plaintext, it servers for both authentication
3879 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3880 if any retry times were supplied.
3882 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3883 connection, where there is no client IP address to check. In this
3884 situation, the verify now always succeeds.
3886 PH/25 Applied John Jetmore's -Mset patch.
3888 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3890 PH/27 In a string expansion for a processed (not raw) header when multiple
3891 headers of the same name were present, leading whitespace was being
3892 removed from all of them, but trailing whitespace was being removed only
3893 from the last one. Now trailing whitespace is removed from each header
3894 before concatenation. Completely empty headers in a concatenation (as
3895 before) are ignored.
3897 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3898 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3900 PH/29 [Removed. This was a change that I later backed out, and forgot to
3901 correct the ChangeLog entry (that I had efficiently created) before
3902 committing the later change.]
3904 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3905 address errors (4xx response to RCPT) over the same connection as other
3906 messages routed to the same hosts. Such deliveries are always "forced",
3907 so retry times are not inspected. This resulted in far too many retries
3908 for the affected addresses. The effect occurred only when there were more
3909 hosts than the hosts_max_try setting in the smtp transport when it had
3910 the 4xx errors. Those hosts that it had tried were not added to the list
3911 of hosts for which the message was waiting, so if all were tried, there
3912 was no problem. Two fixes have been applied:
3914 (i) If there are any address or message errors in an SMTP delivery, none
3915 of the hosts (tried or untried) are now added to the list of hosts
3916 for which the message is waiting, so the message should not be a
3917 candidate for sending over the same connection that was used for a
3918 successful delivery of some other message. This seems entirely
3919 reasonable: after all the message is NOT "waiting for some host".
3920 This is so "obvious" that I'm not sure why it wasn't done
3921 previously. Hope I haven't missed anything, but it can't do any
3922 harm, as the worst effect is to miss an optimization.
3924 (ii) If, despite (i), such a delivery is accidentally attempted, the
3925 routing retry time is respected, so at least it doesn't keep
3926 hammering the server.
3928 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3929 in ${readsocket because some servers need this prod.
3931 PH/32 Added some extra debug output when updating a wait-xxx database.
3933 PH/33 The hint "could be header name not terminated by colon", which has been
3934 given for certain expansion errors for a long time, was not being given
3935 for the ${if def:h_colon_omitted{... case.
3937 PH/34 The spec says: "With one important exception, whenever a domain list is
3938 being scanned, $domain contains the subject domain." There was at least
3939 one case where this was not true.
3941 PH/35 The error "getsockname() failed: connection reset by peer" was being
3942 written to the panic log as well as the main log, but it isn't really
3943 panic-worthy as it just means the connection died rather early on. I have
3944 removed the panic log writing for the ECONNRESET error when getsockname()
3947 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3948 runs only) independently of the message's sender address. This meant
3949 that, if the 4xx error was in fact related to the sender, a different
3950 message to the same recipient with a different sender could confuse
3951 things. In particular, this can happen when sending to a greylisting
3952 server, but other circumstances could also provoke similar problems.
3953 I have changed the default so that the retry time for these errors is now
3954 based a combination of the sender and recipient addresses. This change
3955 can be overridden by setting address_retry_include_sender=false in the
3958 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3959 remote server are returned as part of bounce messages. This was not
3960 happening for LMTP over a pipe (the lmtp transport), but now it is the
3961 same for both kinds of LMTP.
3963 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3964 in header lines that were in fact CNAMEs. This is no longer the case.
3966 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3967 and queue runs started by the daemon processed all messages. This has
3968 been fixed so that -R and -S can now usefully be given with -q<time>.
3970 PH/40 Import PCRE release 6.7 (fixes some bugs).
3972 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3974 PH/42 Give an error if -q is specified more than once.
3976 PH/43 Renamed the variables $interface_address and $interface_port as
3977 $received_ip_address and $received_port, to make it clear that these
3978 values apply to message reception, and not to the outgoing interface when
3979 a message is delivered. (The old names remain recognized, of course.)
3981 PH/44 There was no timeout on the connect() call when using a Unix domain
3982 socket in the ${readsocket expansion. There now is.
3984 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3985 be meaningful with "accept".
3987 SC/01 Eximstats V1.43
3988 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3990 SC/02 Eximstats V1.44
3991 Use a glob alias rather than an array ref in the generated
3992 parser. This improves both readability and performance.
3994 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3995 Collect SpamAssassin and rejection statistics.
3996 Don't display local sender or destination tables unless
3997 there is data to show.
3998 Added average volumes into the top table text output.
4000 SC/04 Eximstats V1.46
4001 Collect data on the number of addresses (recipients)
4002 as well as the number of messages.
4004 SC/05 Eximstats V1.47
4005 Added 'Message too big' to the list of mail rejection
4006 reasons (thanks to Marco Gaiarin).
4008 SC/06 Eximstats V1.48
4009 Mainlog lines which have GMT offsets and are too short to
4010 have a flag are now skipped.
4012 SC/07 Eximstats V1.49 (Alain Williams)
4013 Added the -emptyok flag.
4015 SC/08 Eximstats V1.50
4016 Fixes for obtaining the IP address from reject messages.
4018 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4019 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4020 whitespace changes from 4.64-PH/27
4022 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4025 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4026 are found, allow negative numbers in numeric criteria)
4028 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4030 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4031 to match changes made in 4.64-PH/43
4033 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4035 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4036 "domain" is actually an IP address, and force a failure. This locks out
4037 those revolvers/nameservers that support "A-for-A" lookups, in
4038 contravention of the specifications.
4040 PH/48 When a host name was looked up from an IP address, and the subsequent
4041 forward lookup of the name timed out, the host name was left in
4042 $sender_host_name, contrary to the specification.
4044 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4045 restricted to single-key lookups, Exim was not diagnosing an error if
4046 * or *@ was used with a query-style lookup.
4048 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4050 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4051 long ago, but noone interested enough thought of it. Let's just say that
4052 the "1.1" means that there are some new functions that weren't there at
4053 some point in the past.
4055 PH/51 Error processing for expansion failure of helo_data from an smtp
4056 transport during callout processing was broken.
4058 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4059 tested/used via the -bh/-bhc/-bs options.
4061 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4062 bug, fixed in subsequent PCRE releases).
4064 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4065 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4067 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4073 SC/01 Use a glob alias rather than an array ref in eximstats generated
4074 parser. This improves both readability and performance.
4076 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4077 Don't display local sender or destination tables in eximstats unless
4078 there is data to show.
4079 Added average volumes into the eximstats top table text output.
4081 SC/03 Collect data on the number of addresses (recipients) as well
4082 as the number of messages in eximstats.
4084 TF/01 Correct an error in the documentation for the redirect router. Exim
4085 does (usually) call initgroups() when daemonizing.
4087 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4088 with consistent privilege compared to when running as a daemon.
4090 TF/03 Note in the spec that $authenticated_id is not set for local
4091 submissions from trusted users.
4093 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4094 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4096 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4097 by adding some example configuration directives to the default
4098 configuration file. A little bit of work is required to uncomment the
4099 directives and define how usernames and passwords are checked, but
4100 there is now a framework to start from.
4102 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4103 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4104 without this. I don't know how relevant this is to other LDAP libraries.
4106 PH/02 Add the verb name to the "unknown ACL verb" error.
4108 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4110 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4112 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4113 directory not expanded when it should be if an expanded home directory
4114 was set for the address (which is overridden by the transport).
4116 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4119 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4120 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4121 because it is too late at that time, and has no effect.
4123 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4124 security issue with \' (bugzilla #107). I could not use the
4125 PQescapeStringConn() function, because it needs a PGconn value as one of
4128 PH/08 When testing addresses using -bt, indicate those final addresses that
4129 are duplicates that would not cause an additional delivery. At least one
4130 person was confused, thinking that -bt output corresponded to deliveries.
4131 (Suppressing duplicates isn't a good idea as you lose the information
4132 about possibly different redirections that led to the duplicates.)
4134 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4135 systems where poll() doesn't work, in particular OS X.
4137 PH/10 Added more information to debugging output for retry time not reached.
4139 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4140 operations in malware.c.
4142 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4145 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4146 syslog_duplication set false, log lines that would normally be written
4147 both the the main log and to the reject log were not written to syslog at
4150 PH/14 In the default configuration, change the use of "message" in ACL warn
4151 statements to "add_header".
4153 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4154 not followed by a command (e.g. "seen endif").
4156 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4157 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4160 PH/17 Added extra conditions to the default value of delay_warning_condition
4164 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4165 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4166 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4169 The Auto-Submitted: and various List- headers are standardised, whereas I
4170 don't think Precedence: ever was.
4172 PH/18 Refactored debugging code in route_finduser() to show more information,
4173 in particular, the error code if getpwnam() issues one.
4175 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4176 This is apparently needed in addition to the PH/07 change above to avoid
4177 any possible encoding problems.
4179 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4180 but not after initializing Perl.
4182 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4183 output them only if debugging. By default they are written stderr,
4184 apparently, which is not desirable.
4186 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4189 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4192 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4194 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4195 authenticated or an ident call has been made. Suppress the default
4196 values for $authenticated_id and $authenticated_sender (but permit -oMai
4197 and -oMas) when testing with -bh.
4199 PH/24 Re-jigged the order of the tests in the default configuration so that the
4200 tests for valid domains and recipients precede the DNS black list and CSA
4201 tests, on the grounds that those ones are more expensive.
4203 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4204 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4205 This bug exists in every version of Exim that I still have, right back to
4208 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4209 However, an attempt to turn on case-sensitivity in a regex key by
4210 including (?-i) didn't work because the subject string was already
4211 lowercased, and the effects were non-intuitive. It turns out that a
4212 one-line patch can be used to allow (?-i) to work as expected.
4218 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4219 other effects) broke the use of negated acl sub-conditions.
4221 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4224 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4225 "Deny" causes Exim to reject the incoming connection with a 554 error.
4226 Unfortunately, if there is a major crisis, such as a disk failure,
4227 tcp-wrappers gives "deny", whereas what one would like would be some
4228 kind of temporary error. A kludge has been added to help with this.
4229 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4230 554 error is used if errno is still zero or contains ENOENT (which occurs
4231 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4234 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4236 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4237 errors. Otherwise a message that provokes a temporary error (when other
4238 messages do not) can cause a whole host to time out.
4240 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4241 addresses were routed directly to files or pipes from a redirect router.
4242 File deliveries just didn't batch; pipe deliveries might have suffered
4245 PH/06 A failure to get a lock for a hints database would erroneously always say
4246 "Failed to get write lock", even when it was really a read lock.
4248 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4249 of 0600. This has been changed to use the value of the lockfile_mode
4250 option (which defaults to 0600).
4252 PH/08 Applied small patch from the Sieve maintainer.
4254 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4255 folder from quota calculations, a direct delivery into this folder messed
4256 up the contents of the maildirsize file. This was because the regex was
4257 used only to exclude .Trash (or whatever) when the size of the mailbox
4258 was calculated. There was no check that a delivery was happening into an
4259 excluded directory. This bug has been fixed by ignoring all quota
4260 processing for deliveries into excluded directories.
4262 PH/10 Added the maildirfolder_create_regex option to appendfile.
4268 PH/01 The code for finding all the local interface addresses on a FreeBSD
4269 system running IPv6 was broken. This may well have applied to all BSD
4270 systems, as well as to others that have similar system calls. The broken
4271 code found IPv4 interfaces correctly, but gave incorrect values for the
4272 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4273 that it would not match correctly against @[] and not recognize the IPv6
4276 PH/02 The ipliteral router was not recognizing addresses of the form user@
4277 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4279 PH/03 Added disable_ipv6.
4281 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4282 decoded form, because it is most often used to construct To: headers
4283 lines in autoreplies, and the decoded form may well be syntactically
4284 invalid. However, $reply_address has leading white space removed, and all
4285 newlines turned into spaces so that the autoreply transport does not
4288 PH/05 If group was specified without a user on a router, and no group or user
4289 was specified on a transport, the group from the router was ignored.
4291 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4292 for visible compile-time settings that can be used to change these
4293 numbers, for those that want even more. Backwards compatibility with old
4294 spool files has been maintained. However, going back to a previous Exim
4295 release will lost any variables that are in spool files.
4297 PH/07 Two small changes when running in the test harness: increase delay when
4298 passing a TCP/IP connection to a new process, in case the original
4299 process has to generate a bounce, and remove special handling of
4300 127.0.0.2 (sic), which is no longer necessary.
4302 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4303 be the same on different OS.
4305 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4308 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4309 whether --show-vars was specified or not
4311 JJ/02 exipick: Added support for new ACL variable spool format introduced
4314 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4315 syntactically invalid From: or Reply-to: line, and a filter used this to
4316 generate an autoreply, and therefore failed to obtain an address for the
4317 autoreply, Exim could try to deliver to a non-existent relative file
4318 name, causing unrelated and misleading errors. What now happens is that
4319 it logs this as a hard delivery error, but does not attempt to create a
4322 PH/11 The exinext utility has a -C option for testing purposes, but although
4323 the given file was scanned by exinext itself; it wasn't being passed on
4324 when Exim was called.
4326 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4327 an end-of-file indication when reading a command response.
4329 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4330 compiled. In many other places in Exim, IPv6 addresses are always
4331 recognized, so I have changed this. It also means that IPv4 domain
4332 literals of the form [IPV4:n.n.n.n] are now always recognized.
4334 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4335 used if the router is not running as root, for example, when verifying at
4336 ACL time, or when using -bh. The debugging output from this situation was
4337 non-existent - all you got was a failure to exec. I have made two
4340 (a) Failures to set uid/gid, the current directory, or a process leader
4341 in a subprocess such as that created by queryprogram now generate
4342 suitable debugging output when -d is set.
4344 (b) The queryprogram router detects when it is not running as root,
4345 outputs suitable debugging information if -d is set, and then runs
4346 the subprocess without attempting to change uid/gid.
4348 PH/15 Minor change to Makefile for building test_host (undocumented testing
4351 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4352 additional section of a DNS packet that returns MX or SRV records.
4353 Instead, it always explicitly searches for A/AAAA records. This avoids
4354 major problems that occur when a DNS server includes only records of one
4355 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4356 fixed another bug: if SRV records were looked up and the corresponding
4357 address records were *not* found in the additional section, the port
4358 values from the SRV records were lost.
4360 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4361 using the correct key (the original address) when searching the retry
4362 rules in order to find which one to use for generating the retry hint.
4364 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4365 adding the default one. Similarly, if it contains a Reply-To: header, the
4366 errors_reply_to option, if set, is not used.
4368 PH/19 When calculating a retry time, Exim used to measure the "time since
4369 failure" by looking at the "first failed" field in the retry record. Now
4370 it does not use this if it is later than than the arrival time of the
4371 message. Instead it uses the arrival time. This makes for better
4372 behaviour in cases where some deliveries succeed, thus re-setting the
4373 "first failed" field. An example is a quota failure for a huge message
4374 when small messages continue to be delivered. Without this change, the
4375 "time since failure" will always be short, possible causing more frequent
4376 delivery attempts for the huge message than are intended.
4377 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4379 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4380 $1, $2, $3) because the numerical variables can be reset during some
4381 expansion items (e.g. "match"), thereby losing the authentication data.
4383 PH/21 Make -bV show the size of off_t variables so that the test suite can
4384 decide whether to run tests for quotas > 2G.
4386 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4387 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4388 filecount value is greater than 2G or if a quota value is greater than 2G
4389 on a system where the size of off_t is not greater than 4, a panic error
4392 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4393 never match. The debug and -bh output now contains an explicit error
4394 message indicating a malformed IPv4 address or mask.
4396 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4397 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4398 PH/23 above applies.
4400 PH/25 Do not write to syslog when running in the test harness. The only
4401 occasion when this arises is a failure to open the main or panic logs
4402 (for which there is an explicit test).
4404 PH/26 Added the /no_tell option to "control=freeze".
4406 PH/27 If a host name lookup failed very early in a connection, for example, if
4407 the IP address matched host_lookup and the reverse lookup yielded a name
4408 that did not have a forward lookup, an error message of the form "no IP
4409 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4410 could be logged. Now it outputs the IP address instead of "NULL".
4412 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4413 allows the sender and the authenticated sender to be set when
4414 submitting a message from within Exim. Since child_open_exim() is
4415 documented for local_scan(), the new function should be too.
4417 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4418 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4419 results in an empty string is now treated as unset.
4421 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4423 PH/31 Added sender_verify_fail logging option.
4425 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4426 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4427 tidied the source and removed it altogether.
4429 PH/33 When a queue run was abandoned because the load average was too high, a
4430 log line was always written; now it is written only if the queue_run log
4431 selector is set. In addition, the log line for abandonment now contains
4432 information about the queue run such as the pid. This is always present
4433 in "start" and "stop" lines but was omitted from the "abandon" line.
4435 PH/34 Omit spaces between a header name and the colon in the error message that
4436 is given when verify = headers_syntax fails (if there are lots of them,
4437 the message gets confusing).
4439 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4440 names, as there are now some PTR records that contain slashes. This check
4441 is only to protect against broken name servers that fall over on strange
4442 characters, so the fact that it applies to all lookups doesn't matter.
4444 PH/36 Now that the new test suite is complete, we can remove some of the
4445 special code in Exim that was needed for the old test suite. For example,
4446 sorting DNS records because real resolvers return them in an arbitrary
4447 order. The new test suite's fake resolver always returns records in the
4450 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4451 bounces) except when queue_only is set, to avoid logging races between
4452 the different processes.
4454 PH/38 Panic-die if .include specifies a non-absolute path.
4456 PH/39 A tweak to the "H" retry rule from its user.
4458 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4459 a label. They prevented compilation on older perls.
4461 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4462 a warning to be raised on newish perls.
4464 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4465 on queue. Changes to match documented behaviour of showing count of
4466 messages matching specified criteria.
4468 PH/40 Changed the default ident timeout from 30s to 5s.
4470 PH/41 Added support for the use of login_cap features, on those BSD systems
4471 that have them, for controlling the resources used by pipe deliveries.
4473 PH/42 The content-scanning code uses fopen() to create files in which to put
4474 message data. Previously it was not paying any attention to the mode of
4475 the files. Exim runs with umask(0) because the rest of the code creates
4476 files with open(), and sets the required mode explicitly. Thus, these
4477 files were ending up world-writeable. This was not a big issue, because,
4478 being within the spool directory, they were not world-accessible. I have
4479 created a function called modefopen, which takes an additional mode
4480 argument. It sets umask(777), creates the file, chmods it to the required
4481 mode, then resets the umask. All the relevant calls to fopen() in the
4482 content scanning code have been changed to use this function.
4484 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4485 to 24 hours. This avoids potential overflow problems when processing G
4486 and H retry rules. I suspect nobody ever tinkers with this value.
4488 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4490 PH/45 When the plaintext authenticator is running as a client, the server's
4491 challenges are checked to ensure they are valid base64 strings. By
4492 default, the authentication attempt is cancelled if an invalid string is
4493 received. Setting client_ignore_invalid_base64 true ignores these errors.
4494 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4495 they are received. Thus, the responses can be made to depend on the
4496 challenges. If an invalid string is ignored, an empty string is placed in
4499 PH/46 Messages that are created by the autoreply transport now contains a
4500 References: header, in accordance with RFCs 2822 and 3834.
4502 PH/47 Added authenticated_sender_force to the smtp transport.
4504 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4506 PH/49 Installed latest patch from the Sieve maintainer.
4508 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4509 was also set, the appendfile transport was unnecessarily scanning a
4510 directory of message files (e.g. for maildir delivery) to find the count
4511 of files (along with the size), even though it did not need this
4512 information. It now does the scan only if it needs to find either the
4513 size of the count of files.
4515 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4517 PH/52 Two bugs concerned with error handling when the smtp transport is
4520 (i) Exim was not creating retry information for temporary errors given
4521 for individual recipients after the DATA command when the smtp transport
4522 was used in LMTP mode. This meant that they could be retried too
4523 frequently, and not timed out correctly.
4525 (ii) Exim was setting the flag that allows error details to be returned
4526 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4527 recipients that were returned after the DATA command.
4529 PH/53 This is related to PH/52, but is more general: for any failing address,
4530 when detailed error information was permitted to be returned to the
4531 sender, but the error was temporary, then after the final timeout, only
4532 "retry timeout exceeded" was returned. Now it returns the full error as
4533 well as "retry timeout exceeded".
4535 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4536 do this, and (what is worse) MTAs that accept it.
4538 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4539 will now be deprecated.
4541 PH/56 New os.c-cygwin from the Cygwin maintainer.
4543 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4544 formats (previously only available in exim formats via -bpr, -bpru,
4545 and -bpra. Now also available in native and exiqgrep formats)
4547 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4548 with very large, slow to parse queues
4550 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4552 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4554 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4555 responses to authentication challenges, though it was showing the
4556 challenges; (ii) I've removed the CR characters from the debug output for
4559 PH/58 Allow for the insertion of a newline as well as a space when a string
4560 is turned into more than one encoded-word during RFC 2047 encoding. The
4561 Sieve code now uses this.
4563 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4564 data_4xx, lost_connection, tls_required.
4566 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4567 message was being sent as an SMTP response.
4569 PH/61 Add -l and -k options to exicyclog.
4571 PH/62 When verifying, if an address was redirected to one new address, so that
4572 verification continued, and the new address failed or deferred after
4573 having set something in $address_data, the value of $address_data was not
4574 passed back to the ACL. This was different to the case when no
4575 redirection occurred. The value is now passed back in both cases.
4577 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4578 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4579 use login_cap.h, so on its own it isn't the distinguishing feature. The
4580 new name refers directly to the setclassresources() function.
4582 PH/65 Added configuration files for NetBSD3.
4584 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4586 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4587 is preferred over IPv4.
4589 PH/68 The bounce_return_message and bounce_return_body options were not being
4590 honoured for bounces generated during the reception of non-SMTP messages.
4591 In particular, this applied to messages rejected by the ACL. This bug has
4592 been fixed. However, if bounce_return_message is true and bounce_return_
4593 body is false, the headers that are returned for a non-SMTP message
4594 include only those that have been read before the error was detected.
4595 (In the case of an ACL rejection, they have all been read.)
4597 PH/69 The HTML version of the specification is now built in a directory called
4598 spec_html instead of spec.html, because the latter looks like a path with
4599 a MIME-type, and this confuses some software.
4601 PH/70 Catch two compiler warnings in sieve.c.
4603 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4604 function verify_get_ident() calls ip_connect() to connect a socket, but
4605 if the "connect()" function timed out, ip_connect() used to close the
4606 socket. However, verify_get_ident() also closes the socket later, and in
4607 between Exim writes to the log, which may get opened at this point. When
4608 the socket was closed in ip_connect(), the log could get the same file
4609 descriptor number as the socket. This naturally causes chaos. The fix is
4610 not to close the socket in ip_connect(); the socket should be closed by
4611 the function that creates it. There was only one place in the code where
4612 this was missing, in the iplookup router, which I don't think anybody now
4613 uses, but I've fixed it anyway.
4615 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4616 well as to direct DNS lookups. Otherwise the handling of names in host
4617 lists is inconsistent and therefore confusing.
4623 PH/01 Two changes to the default runtime configuration:
4625 (1) Move the checks for relay_from_hosts and authenticated clients from
4626 after to before the (commented out) DNS black list checks.
4628 (2) Add control=submission to the relay_from_hosts and authenticated
4629 clients checks, on the grounds that messages accepted by these
4630 statements are most likely to be submissions.
4632 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4634 (1) Generate an error if the third argument for the ${prvs expansion is
4637 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4640 (3) Reset the variables that are obtained from the first argument of
4641 ${prvscheck and used in the second argument before leaving the code,
4642 because their memory is reclaimed, so using them afterwards may do
4645 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4646 one (it's much easier than Tom thought :-).
4648 (5) Because of (4), we can now allow for the use of $prvscheck_result
4649 inside the third argument.
4651 PH/03 For some reason, the default setting of PATH when running a command from
4652 a pipe transport was just "/usr/bin". I have changed it to
4655 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4656 anything to be listed in the output from -bV.
4658 PH/05 When a filter generated an autoreply, the entire To: header line was
4659 quoted in the delivery log line, like this:
4661 => >A.N.Other <ano@some.domain> <original@ddress> ...
4663 This has been changed so that it extracts the operative address. There
4664 may be more than one such address. If so, they are comma-separated, like
4667 => >ano@some.domain,ona@other.domain <original@ddress> ...
4669 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4670 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4671 not being looked up in the rDNS to get a host name, Exim was showing the
4672 IP address twice in Received: lines, even though the IP addresses were
4673 identical. For example:
4675 Received: from [1.2.3.4] (helo=[1.2.3.4])
4677 However, if the real host name was known, it was omitting the HELO data
4678 if it matched the actual IP address. This has been tidied up so that it
4679 doesn't show the same IP address twice.
4681 PH/07 When both +timestamp and +memory debugging was on, the value given by
4682 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4683 called by the debug printing, thereby overwriting the timestamp buffer.
4684 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4686 PH/08 When the original message was included in an autoreply transport, it
4687 always said "this is a copy of the message, including all the headers",
4688 even if body_only or headers_only was set. It now gives an appropriate
4691 PH/09 Applied a patch from the Sieve maintainer which:
4693 o fixes some comments
4694 o adds the (disabled) notify extension core
4695 o adds some debug output for the result of if/elsif tests
4696 o points to the current vacation draft in the documentation
4697 and documents the missing references header update
4701 o fixes a bug in processing the envelope test (when testing
4702 multiple envelope elements, the last element determined the
4705 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4706 Electronic Mail") by including:
4708 Auto-submitted: auto-generated
4710 in the messages that it generates (bounce messages and others, such as
4711 warnings). In the case of bounce messages for non-SMTP messages, there was
4712 also a typo: it was using "Auto_submitted" (underscore instead of
4713 hyphen). Since every message generated by Exim is necessarily in response
4714 to another message, thes have all been changed to:
4716 Auto-Submitted: auto-replied
4718 in accordance with these statements in the RFC:
4720 The auto-replied keyword:
4722 - SHOULD be used on messages sent in direct response to another
4723 message by an automatic process,
4725 - MUST NOT be used on manually-generated messages,
4727 - MAY be used on Delivery Status Notifications (DSNs) and Message
4728 Disposition Notifications (MDNs),
4730 - MUST NOT be used on messages generated by automatic or periodic
4731 processes, except for messages which are automatic responses to
4734 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4735 to the default Received: header definition.
4737 PH/12 Added log selector acl_warn_skipped (default on).
4739 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4740 variables because (a) they are in the wrong storage pool and (b) even if
4741 they were copied, it wouldn't work properly because of the caching.
4743 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4744 checking when decoding. Apparently there are clients that generate
4745 overlong encoded strings. Why am I not surprised?
4747 PH/15 If the first argument of "${if match_address" was not empty, but did not
4748 contain an "@" character, Exim crashed. Now it writes a panic log message
4749 and treats the condition as false.
4751 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4753 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4754 extension "envelope-auth". The code is finished and in agreement with
4755 other implementations, but there is no documentation so far and in fact,
4756 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4757 not changing the active code.
4759 Print executed "if" and "elsif" statements when debugging is used. This
4760 helps a great deal to understand what a filter does.
4762 Document more things not specified clearly in RFC3028. I had all this
4763 sorted out, when out of a sudden new issues came to my mind. Oops."
4765 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4768 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4769 canonical form (as documented). However, after a host name lookup from
4770 the IP address, check_host() was doing a simple string comparison with
4771 addresses acquired from the DNS when checking that the found name did
4772 have the original IP as one of its addresses. Since any found IPv6
4773 addresses are likely to be in abbreviated form, the comparison could
4774 fail. Luckily, there already exists a function for doing the comparison
4775 by converting both addresses to binary, so now that is used instead of
4776 the text comparison.
4778 PH/20 There was another similar case to PH/19, when a complete host name was
4779 given in a host list; looking up its IP address could give an abbreviated
4780 form, whereas the current host's name might or might not be abbreviated.
4781 The same fix has been applied.
4787 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4788 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4791 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4792 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4794 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4796 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4797 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4798 Also, for "Auto-Submitted", treat anything other than "no" as
4799 non-personal, in accordance with RFC 3834. (Previously it treated
4800 anything starting "auto-" as non-personal.)
4802 TF/01 The control=submission/name=... option had a problem with syntax
4803 errors if the name included a slash character. The /name= option
4804 now slurps the rest of the string, so it can include any characters
4805 but it must come last in the list of options (after /sender_retain
4808 PH/05 Some modifications to the interface to the fake nameserver for the new
4816 TK/01 Added the "success_on_redirect" address verification option. See
4817 NewStuff for rationale and an example.
4819 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4821 PH/02 Patch to exigrep to allow it to work on syslog lines.
4823 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4824 fread() to skip over the body file's header line, because in Cygwin the
4825 header line is locked and is inaccessible.
4827 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4828 co-exist for some time) to make it clear that it is the Exim ID that is
4829 referenced, not the Message-ID: header line.
4831 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4832 string_format() function, because snprintf() does not exist on all
4835 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4836 ${stat: expansion item.
4838 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4839 protocol synchronization error", to keep the pedants happy.
4841 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4842 well as for IRIX systems, when gcc is being used. See the host.c source
4845 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4847 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4850 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4851 transport and to the smtp transport in LMTP mode.
4853 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4855 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4856 run a filter in a subprocess. This could lead to confusion in subsequent
4857 lookups in the parent process. There should also be a search_tidyup() at
4858 the end of the subprocess.
4860 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4861 only if the host matched helo_try_verify_hosts, which caused the
4862 verification to occur when the EHLO/HELO command was issued. The ACL just
4863 tested the remembered result. Now, if a previous verification attempt has
4864 not happened, "verify = helo" does it there and then.
4866 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4868 TK/03 Fix log output including CR from clamd.
4870 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4871 exist provoked a memory error which could cause a segfault.
4873 PH/15 Installed PCRE 6.2
4875 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4877 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4878 of the problem. Specifically, suggested +O2 rather than +O1 for the
4881 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4883 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4884 record, the port was not correctly logged when the outgoing_port log
4885 selector was set (it logged the transort's default port).
4887 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4888 fallback_hosts, and "hosts" in the smtp transport.
4890 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4891 host errors such as "Connection refused".
4893 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4894 authentication with radiusclient 0.4.9:
4896 - Error returned from rc_read_config was caught wrongly
4897 - Username/password not passed on to radius server due to wrong length.
4899 The presumption is that some radiusclient API changes for 4.51/PH/17
4900 were not taken care of correctly. The code is still untested by me (my
4901 Linux distribution still has 0.3.2 of radiusclient), but it was
4902 contributed by a Radius user.
4904 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4905 expanding the "port" option of the smtp transport.
4907 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4908 while reading a MIME header. Thanks to Tom Hughes for a patch.
4910 PH/24 Include config.h inside local_scan.h so that configuration settings are
4913 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4914 that in an ACL for RCPT (for example), you can examine exactly what was
4917 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4918 commands, but it was not correctly comparing the address with the actual
4919 client host address. Thus, it would show the EHLO address in Received:
4920 header lines when this was not necessary.
4922 PH/27 Added the % operator to ${eval:}.
4924 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4925 it should be ignoring failures (because with -C, for example, it has lost
4926 privilege). It wasn't ignoring creation failures other than "already
4929 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4932 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4933 because an input line was too long, either on its own, or by virtue of
4934 too many continuations, the temporary file was not being removed, and the
4935 return code was incorrect.
4937 PH/31 Missing "BOOL" in function definition in filtertest.c.
4939 PH/32 Applied Sieve patches from the maintainer.
4941 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4943 PH/33 Added "verify = not_blind".
4945 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4946 Local/Makefile (with some defaults set). These are used in built scripts
4947 such as exicyclog, but they have never been used in the exim_install
4948 script (though there are many overriding facilities there). I have
4949 arranged that the exim_install script now takes note of these two
4952 PH/35 Installed configuration files for Dragonfly.
4954 PH/36 When a locally submitted message by a trusted user did not contain a
4955 From: header, and the sender address was obtained from -f or from an SMTP
4956 MAIL command, and the trusted user did not use -F to supply a sender
4957 name, $originator_name was incorrectly used when constructing a From:
4958 header. Furthermore, $originator_name was used for submission mode
4959 messages from external hosts without From: headers in a similar way,
4960 which is clearly wrong.
4962 PH/37 Added control=suppress_local_fixups.
4964 PH/38 When log_selector = +received_sender was set, and the addition of the
4965 sender made the log line's construction buffer exactly full, or one byte
4966 less than full, an overflow happened when the terminating "\n" was
4969 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4970 when the result of a list match is failure because a DNS lookup failed.
4972 PH/40 RM_COMMAND is now used in the building process.
4974 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4975 the "build-* directories that it finds.
4977 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4978 address in a domain literal was a prefix of an interface address.
4980 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4981 when verifying a sender address, unless rewrite_headers is false.
4983 PH/44 Wrote a long comment about why errors_to addresses are verified as
4984 recipients, not senders.
4986 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4987 the ratelimit ACL was added.
4989 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4991 PH/46 Added extra information about PostgreSQL errors to the error string.
4993 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4994 avoiding the need to install special zones in a real server. This is
4995 backwards compatible; if it can't find the fake resolver, it drops back.
4996 Thus, both old and new test suites can be run.
4998 TF/02 Added util/ratelimit.pl
5000 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5001 clock is set back in time.
5003 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5004 Candler <B.Candler@pobox.com>.
5006 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5007 OK for addresses that are the result of redirections.
5009 PH/48 A number of further additions for the benefit of the new test suite,
5010 including a fake gethostbyname() that interfaces to the fake DNS resolver
5013 TF/06 The fix for widen_domains has also been applied to qualify_single and
5014 search_parents which are the other dnslookup options that can cause
5017 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5020 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5022 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5023 service" if it encountered an MX record with an empty target hostname.
5024 The message is now "an MX or SRV record indicated no SMTP service".
5026 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5027 if the DNS of the sending site is misconfigured. This is quite a
5028 common situation. This change restores the behaviour of treating a
5029 helo verification defer as a failure.
5031 PH/51 If self=fail was set on a router, the bounce message did not include the
5032 actual error message.
5038 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5040 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5041 command itself ended in error, the underlying message about the transport
5042 filter timeout was being overwritten with the pipe command error. Now the
5043 underlying error message should be appended to the second error message.
5045 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5047 PH/02 Reduce the amount of output that "make" produces by default. Full output
5048 can still be requested.
5050 PH/03 The warning log line about a condition test deferring for a "warn" verb
5051 was being output only once per connection, rather than after each
5052 occurrence (because it was using the same function as for successful
5053 "warn" verbs). This seems wrong, so I have changed it.
5055 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5056 it should not have, which might have caused a crash in the right
5057 circumstances, but probably never did.
5059 PH/04 Installed a modified version of Tony Finch's patch to make submission
5060 mode fix the return path as well as the Sender: header line, and to
5061 add a /name= option so that you can make the user's friendly name appear
5064 TF/03 Added the control = fakedefer ACL modifier.
5066 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5067 Mark Lowes for thorough testing.
5069 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5071 TK/03 Merged latest SRS patch from Miles Wilton.
5073 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5074 with the definition in sysexits.h (which is #included earlier).
5075 Fortunately, Exim does not actually use EX_OK. The code used to try to
5076 preserve the sysexits.h value, by assuming that macro definitions were
5077 scanned for macro replacements. I have been disabused of this notion,
5078 so now the code just undefines EX_OK before #including unistd.h.
5080 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5081 in the smtp transport. When a block could not be written in a single
5082 write() function, the timeout was being re-applied to each part-write.
5083 This seems wrong - if the receiver was accepting one byte at a time it
5084 would take for ever. The timeout is now adjusted when this happens. It
5085 doesn't have to be particularly precise.
5087 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5088 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5090 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5091 without a domain if the check to <postmaster@domain> fails.
5093 SC/01 Eximstats: added -xls and the ability to specify output files
5094 (patch written by Frank Heydlauf).
5096 SC/02 Eximstats: use FileHandles for outputting results.
5098 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5100 SC/04 Eximstats: fixed display of large numbers with -nvr option
5102 SC/05 Eximstats: fixed merging of reports with empty tables.
5104 SC/06 Eximstats: added the -include_original_destination flag
5106 SC/07 Eximstats: removed tabs and trailing whitespace.
5108 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5110 TK/06 MBOX spool code: Add real "From " MBOX separator line
5111 so the .eml file is really in mbox format (even though
5112 most programs do not really care). Patch from Alex Miller.
5114 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5115 The latter is generated from $received_to and is only set if the
5116 message has one envelope recipient. SA can use these headers,
5117 obviously out-of-the-box. Patch from Alex Miller.
5119 PH/08 The ${def test on a variable was returning false if the variable's
5120 value was "0", contrary to what the specification has always said!
5121 The result should be true unless the variable is empty.
5123 PH/09 The syntax error of a character other than { following "${if
5124 def:variable_name" (after optional whitespace) was not being diagnosed.
5125 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5126 accidental colon was present, for example, could give incorrect results.
5128 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5129 result is used (not including appendfile, where other changes are about
5132 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5133 This involved changing a lot of size variables from int to off_t. It
5134 should work with maildirs and everything.
5136 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5137 spamd dying while we are connected to it.
5139 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5142 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5143 function for BDB 4.3.
5145 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5147 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5148 the sources for PCRE have been re-arranged and more files are now
5151 PH/15 The code I had for printing potentially long long variables in PH/11
5152 above was not the best (it lost precision). The length of off_t variables
5153 is now inspected at build time, and an appropriate printing format (%ld
5154 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5155 to be "long long int" or "long int". This is needed for the internal
5156 formatting function string_vformat().
5158 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5159 the configuration file to be ":syslog", then the script "guesses" where
5160 the logs files are, rather than using the compiled in default. In our
5161 case the guess is not the same as the compiled default, so the script
5162 suddenly stopped working when I started to use syslog. The patch checks
5163 to see if log_file_path is "". If so, it attempts to read it from exim
5164 with no configuration file to get the compiled in version, before it
5165 falls back to the previous guessing code."
5167 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5168 implementing BATV in an Exim configuration. See NewStuff for the gory
5171 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5172 Makefile that are specific to HP-UX.
5174 PH/18 If the "use_postmaster" option was set for a recipient callout together
5175 with the "random" option, the postmaster address was used as the MAIL
5176 FROM address for the random test, but not for the subsequent recipient
5177 test. It is now used for both.
5179 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5180 patch removes a few documentation additions to RFC 3028, because the
5181 latest draft now contains them. It adds the new en;ascii-case comparator
5182 and a new error check for 8bit text in MIME parts. Comparator and
5183 require names are now matched exactly. I enabled the subaddress
5184 extension, but it is not well tested yet (read: it works for me)."
5186 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5187 rework some of the code of TK/09 above to avoid the hardwired use of
5188 "%lld" and "long long". Replaced the call to snprintf() with a call to
5191 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5192 records point to non-existent hosts", "retry timeout exceeded", and
5193 "retry time not reached for any host after a long failure period".
5195 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5196 experimental DomainKeys support:
5198 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5199 (2) On an error such as an illegally used "control", the wrong name for
5200 the control was given.
5202 These problems did NOT occur unless DomainKeys support was compiled.
5204 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5206 PH/24 Added ${if match_ip condition.
5208 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5209 databases so that it will be absolutely obvious if a crash occurs in the
5210 DB library. This is a regular occurrence (often caused by mis-matched
5213 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5214 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5215 picked up on a user's system that detects such things. There doesn't seem
5216 to be a gcc warning option for this - only an attribute that has to be
5217 put on the function's prototype. It seems that in Fedora Core 4 they have
5218 set this on a number of new functions. No doubt there will be more in due
5221 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5222 specify a transport. However, if an address that was verified by such a
5223 router was the subject of a callout, Exim crashed because it tried to
5224 read the rcpt_include_affixes from the non-existent transport. Now it
5225 just assumes that the setting of that option is false. This bug was
5226 introduced by 4.51/PH/31.
5228 PH/28 Changed -d+all to exclude +memory, because that information is very
5229 rarely of interest, but it makes the output a lot bigger. People tend to
5230 do -d+all out of habit.
5232 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5233 code in os-type was giving problems when libc.so lives in lib64, like on
5236 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5237 aren't the modern standard, and it seems that some systems' include files
5238 don't always have them. Exim was already checking for some of the newer
5239 ones like T_AAAA, and defining it itself. I've added checks for all the
5240 record types that Exim uses.
5242 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5243 not automatically generating a new one, as it is supposed to. This
5244 prevented TLS from working. If the file did exist, but contained invalid
5245 data, a new version was generated, as expected. It was only the case of a
5246 non-existent file that was broken.
5248 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5249 with a change in libdomainkeys > 0.64.
5251 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5252 from DNS. If the selector record carries the flag, it now has
5253 precedence over the domain-wide flag.
5255 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5257 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5258 the use of a port name that isn't defined in /etc/services) occurred, the
5259 message was deferred as in a normal delivery, and thus remained on the
5260 spool, instead of being failed because of the mua_wrapper setting. This
5261 is now fixed, and I tidied up some of the mua_wrapper messages at the
5264 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5265 the messages in a hash of arrays rather than using individual hashes.
5266 This is a bit cleaner and results in dramatic memory savings, albeit
5267 at a slight CPU cost.
5269 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5270 as requested by Marc Sherman.
5272 SC/10 Eximstats: added histograms for user specified patterns as requested
5275 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5277 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5278 fopen() in the content-scanning modules that did not already have it.
5284 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5285 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5287 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5289 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5291 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5292 to test Sieve filters that use "vacation".
5294 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5295 that changes the way the GnuTLS parameters are stored in the cache file.
5296 The new format can be generated externally. For backward compatibility,
5297 if the data in the cache doesn't make sense, Exim assumes it has read an
5298 old-format file, and it generates new data and writes a new file. This
5299 means that you can't go back to an older release without removing the
5302 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5303 work if there are any delivery delays because "one_time" forces the
5304 parent to be marked "delivered", so its unseen clone is never tried
5305 again. For this reason, Exim now forbids the simultaneous setting of
5308 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5309 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5310 entry if you want to know the details. The fix, however, neglected to
5311 consider the case when local delivery batching is involved. The test for
5312 "previously delivered" was not happening when checking to see if an
5313 address could be batched with a previous (undelivered) one; under
5314 certain circumstances this could lead to multiple deliveries to the same
5317 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5318 in its include files, and this causes problems building Exim.
5320 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5321 header_syntax) but Exim was just ignoring anything given after a slash.
5322 In particular, this caused confusion with an attempt to use "verify =
5323 reverse_host_lookup/defer_ok". An error is now given when options are
5324 supplied for verify items that do not have them. (Maybe reverse_host_
5325 lookup should have a defer_ok option, but that's a different point.)
5327 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5328 defined by RFC 821) to 2048, because there were problems with some AUTH
5329 commands, and RFC 1869 says the size should be increased for extended
5330 SMTP commands that take arguments.
5332 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5335 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5336 "unknown" error; now it says that the functionality isn't in the binary.
5338 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5339 an address' error message when a string expansion fails (syntax or
5340 whatever). Otherwise the password may appear in the log. Following change
5341 PH/42 below, there is no longer a chance of it appearing in a bounce
5344 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5346 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5347 of its addresses was ever tried. (Bugzilla bug #2.)
5349 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5350 the result incorrectly in the debug output. (It correctly added a newline
5351 to what was transported.)
5353 TF/01 Added $received_time.
5355 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5356 commented out examples of how to interface to a virus scanner and to
5357 SpamAssassin. Also added commented examples of av_scanner and
5358 spamd_address settings.
5360 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5361 and controls are allowed in which ACLs. There were a couple of minor
5362 errors. Some of the entries in the conditions table (which is a table of
5363 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5364 negation of where the condition IS allowed.
5366 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5368 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5369 header file does not have a version number, so I've had to invent a new
5370 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5371 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5372 radiusclient), but it was contributed by a Radius user.
5374 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5375 files or directories, for OpenSSL.
5377 PH/19 When an Exim process that is running as root has to create an Exim log
5378 file, it does so in a subprocess that runs as exim:exim so as to get the
5379 ownership right at creation (otherwise, other Exim processes might see
5380 the file with the wrong ownership). There was no test for failure of this
5381 fork() call, which would lead to the process getting stuck as it waited
5382 for a non-existent subprocess. Forks do occasionally fail when resources
5383 run out. I reviewed all the other calls to fork(); they all seem to check
5386 PH/20 When checking for unexpected SMTP input at connect time (before writing
5387 the banner), Exim was not dealing correctly with a non-positive return
5388 from the read() function. If the client had disconnected by this time,
5389 the result was a log entry for a synchronization error with an empty
5390 string after "input=" when read() returned zero. If read() returned -1
5391 (an event I could not check), uninitialized data bytes were printed.
5392 There were reports of junk text (parts of files, etc) appearing after
5395 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5397 PH/22 Added support for macro redefinition, and (re)definition in between
5398 driver and ACL definitions.
5400 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5401 forgetting to use the resulting value; it was using the unexpanded value.
5403 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5404 hadn't been configured. The fix is from Juergen Kreileder, who
5405 understands it better than I do:
5407 "Here's what I see happening with three configured cyrus_sasl
5408 authenticators configured (plain, login, cram-md5):
5410 On startup auth_cyrus_sasl_init() gets called for each of these.
5411 This means three calls to sasl_listmech() without a specified mech_list.
5412 => SASL tests which mechs of all available mechs actually work
5413 => three warnings about OTP not working
5414 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5416 With the patch, sasl_listmech() also gets called three times. But now
5417 SASL's mech_list option is set to the server_mech specified in the the
5418 authenticator. Or in other words, the answer from sasl_listmech()
5419 gets limited to just the mech you're testing for (which is different
5421 => the return list contains just 'plain' or 'login', 'cram-md5' or
5422 nothing depending on the value of ob->server_mech.
5424 I've just tested the patch: Authentication still works fine,
5425 unavailable mechs specified in the exim configuration are still
5426 caught, and the auth.log warnings about OTP are gone."
5428 PH/25 When debugging is enabled, the contents of the command line are added
5429 to the debugging output, even when log_selector=+arguments is not
5432 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5433 answer is "GNU", and only if the return is "GNU/something" is the answer
5436 PH/27 $acl_verify_message is now set immediately after the failure of a
5437 verification in an ACL, and so is available in subsequent modifiers. In
5438 particular, the message can be preserved by coding like this:
5440 warn !verify = sender
5441 set acl_m0 = $acl_verify_message
5443 Previously, $acl_verify_message was set only while expanding "message"
5444 and "log_message" when a very denied access.
5446 PH/28 Modified OS/os.c-Linux with
5448 -#ifndef OS_LOAD_AVERAGE
5449 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5451 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5452 nomenclature these days.)
5454 PH/29 Installed patch from the Sieve maintainer that adds the options
5455 sieve_useraddress and sieve_subaddress to the redirect router.
5457 PH/30 In these circumstances:
5458 . Two addresses routed to the same list of hosts;
5459 . First host does not offer TLS;
5460 . First host accepts first address;
5461 . First host gives temporary error to second address;
5462 . Second host offers TLS and a TLS session is established;
5463 . Second host accepts second address.
5464 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5465 and peerdn, if requested) that were in fact used only for the second
5468 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5469 attention to any local part prefix or suffix that was matched by the
5470 router that accepted the address. It now behaves in the same way as it
5471 does for delivery: the affixes are removed from the local part unless
5472 rcpt_include_affixes is set on the transport.
5474 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5475 timeout during the DATA phase of an incoming message.
5477 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5478 applied a patch sanctioned by the Sieve maintainer.
5480 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5481 the uid or gid is negative. A case of a negative gid caused this to be
5482 noticed. The fix allows for either to be negative.
5484 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5485 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5488 PH/36 The change PH/12 above was broken. Fixed it.
5490 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5491 the grounds that routing the same address twice would always produce the
5492 same answer. This might have been true once, but it is certainly no
5493 longer true now. Routing a child address may depend on the previous
5494 routing that produced that child. Some complicated redirection strategies
5495 went wrong when messages had multiple recipients, and made Exim's
5496 behaviour dependent on the order in which the addresses were given.
5498 I have moved the duplicate checking until after the routing is complete.
5499 Exim scans the addresses that are assigned to local and remote
5500 transports, and removes any duplicates. This means that more work will be
5501 done, as duplicates will always all be routed, but duplicates are
5502 presumably rare, so I don't expect this is of any significance.
5504 For deliveries to pipes, files, and autoreplies, the duplicate checking
5505 still happens during the routing process, since they are not going to be
5508 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5509 It corrects a timeout issue with spamd. This is Ian's comment: "The
5510 background is that sometimes spamd either never reads data from a
5511 connection it has accepted, or it never writes response data. The exiscan
5512 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5513 blindly assumes that writes won't block so it may never time out."
5515 PH/39 Allow G after quota size as well as K and M.
5517 PH/40 The value set for $authenticated_id in an authenticator may not contain
5518 binary zeroes or newlines because the value is written to log lines and
5519 to spool files. There was no check on this. Now the value is run through
5520 the string_printing() function so that such characters are converted to
5521 printable escape sequences.
5523 PH/41 $message_linecount is a new variable that contains the total number of
5524 lines in the message. Compare $body_linecount, which is the count for the
5527 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5528 bounce and delay warning messages, except in certain special cases, which
5531 (a) An SMTP error message from a remote host;
5532 (b) A message specified in a :fail: redirection;
5533 (c) A message specified in a "fail" command in a system filter;
5534 (d) A message specified in a FAIL return from the queryprogram router;
5535 (e) A message specified by the cannot_route_message router option.
5537 In these cases only, Exim does include the error details in bounce and
5538 warning messages. There are also a few cases where bland messages such
5539 as "unrouteable address" or "local delivery error" are given.
5541 PH/43 $value is now also set for the "else" part of a ${run expansion.
5543 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5544 being worked on, but at least Exim now implements the latest version to
5547 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5548 process to complete was treated as a delivery failure, a timeout while
5549 writing the message to the pipe was logged, but erroneously treated as a
5550 successful delivery. Such timeouts include transport filter timeouts. For
5551 consistency with the overall process timeout, these timeouts are now
5552 treated as errors, giving rise to delivery failures by default. However,
5553 there is now a new Boolean option for the pipe transport called
5554 timeout_defer, which, if set TRUE, converts the failures into defers for
5555 both kinds of timeout. A transport filter timeout is now identified in
5558 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5559 systems where "make" and "gmake" are different, calling "gmake" at top
5560 level broke things. I've arranged for the value of $(MAKE) to be passed
5561 from the Makefile to this script so that it can call the same version of
5565 A note about Exim versions 4.44 and 4.50
5566 ----------------------------------------
5568 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5569 changes of various kinds. As a consequence, a big documentation update was
5570 needed. This delayed the release for rather longer than seemed good, especially
5571 in the light of a couple of (minor) security issues. Therefore, the changes
5572 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5573 release. So 4.44 and 4.50 are in effect two different branches that both start
5576 I have left the 4.50 change log unchanged; it contains all the changes since
5577 4.43. The change log for 4.44 is below; many of its items are identical to
5578 those for 4.50. This seems to be the most sensible way to preserve the
5579 historical information.
5585 1. Minor wording change to the doc/README.SIEVE file.
5587 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5588 computation of the current number of files was incorrect.
5590 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5591 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5594 4. Give more explanation in the error message when the command for a transport
5595 filter fails to execute.
5597 5. There are several places where Exim runs a non-Exim command in a
5598 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5599 was being done only for the command run by the queryprogram router. It is
5600 now done for all such subprocesses. The other cases are: ${run, transport
5601 filters, and the commands run by the lmtp and pipe transports.
5603 6. Added CONFIGURE_GROUP build-time option.
5605 7. Some older OS have a limit of 256 on the maximum number of file
5606 descriptors. Exim was using setrlimit() to set 1000 as a large value
5607 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5608 systems. I've change it so that if it can't get 1000, it tries for 256.
5610 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5611 was an oversight, and furthermore, ever since the addition of extra
5612 controls (e.g. 4.43/32), the checks on when to allow different forms of
5613 "control" were broken. There should now be diagnostics for all cases when a
5614 control that does not make sense is encountered.
5616 9. Added the /retain_sender option to "control=submission".
5618 10. $recipients is now available in the predata ACL (oversight).
5620 11. Tidy the search cache before the fork to do a delivery from a message
5621 received from the command line. Otherwise the child will trigger a lookup
5622 failure and thereby defer the delivery if it tries to use (for example) a
5623 cached ldap connection that the parent has called unbind on.
5625 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5626 of $address_data from the recipient verification was clobbered by the
5627 sender verification.
5629 13. The value of address_data from a sender verification is now available in
5630 $sender_address_data in subsequent conditions in the ACL statement.
5632 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5634 15. Added a new option "connect=<time>" to callout options, to set a different
5637 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5638 was its contents. (It was OK if the option was not defined at all.)
5640 17. A "Completed" log line is now written for messages that are removed from
5641 the spool by the -Mrm option.
5643 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5644 information about exactly what failed.
5646 19. Added -dd to debug only the daemon process.
5648 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5649 handles timeouts, both on the server side and network timeouts. Renamed the
5650 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5652 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5653 It is now set to "smtps".
5655 22. $host_address is now set to the target address during the checking of
5656 ignore_target_hosts.
5658 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5659 being passed; this would have caused $sender_host_name to have been used if
5660 matching the list had actually called for a host name (not very likely,
5661 since this list is usually IP addresses). A host name is now passed as
5664 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5665 code that specifies a non-restarting handler (typically sigaction() in
5666 modern systems) in an attempt to fix a rare and obscure crash bug.
5668 25. Narrowed the window for a race in the daemon that could cause it to ignore
5669 SIGCHLD signals. This is not a major problem, because they are used only to
5670 wake it up if nothing else does.
5672 26. A malformed maildirsize file could cause Exim to calculate negative values
5673 for the mailbox size or file count. Odd effects could occur as a result.
5674 The maildirsize information is now recalculated if the size or filecount
5677 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5678 support for a long time. Removed HAVE_SYS_VFS_H.
5680 28. Installed the latest version of exipick from John Jetmore.
5682 29. In an address list, if the pattern was not a regular expression, an empty
5683 subject address (from a bounce message) matched only if the pattern was an
5684 empty string. Non-empty patterns were not even tested. This was the wrong
5685 because it is perfectly reasonable to use an empty address as part of a
5686 database query. An empty address is now tested by patterns that are
5687 lookups. However, all the other forms of pattern expect the subject to
5688 contain a local part and a domain, and therefore, for them, an empty
5689 address still always fails if the pattern is not itself empty.
5691 30. Exim went into a mad DNS loop when attempting to do a callout where the
5692 host was specified on an smtp transport, and looking it up yielded more
5693 than one IP address.
5695 31. Re-factored the code for checking spool and log partition space into a
5696 function that finds that data and another that does the check. The former
5697 is then used to implement four new variables: $spool_space, $log_space,
5698 $spool_inodes, and $log_inodes.
5700 32. The RFC2047 encoding function was originally intended for short strings
5701 such as real names; it was not keeping to the 75-character limit for
5702 encoded words that the RFC imposes. It now respects the limit, and
5703 generates multiple encoded words if necessary. To be on the safe side, I
5704 have increased the buffer size for the ${rfc2047: expansion operator from
5707 33. It is now permitted to omit both strings after an "if" condition; if the
5708 condition is true, the result is "true". As before, when the second string
5709 is omitted, a false condition yields an empty string. This makes it less
5710 cumbersome to write custom ACL and router conditions.
5712 34. Failure to deliver a bounce message always caused it to be frozen, even if
5713 there was an errors_to setting on the router. The errors_to setting is now
5716 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5717 canonical form (fully expanded) before being placed in
5718 $sender_host_address.
5720 36. The table in the code that translates DNS record types into text (T_A to
5721 "A" for instance) was missing entries for NS and CNAME. It is just possible
5722 that this could have caused confusion if both these types were looked up
5723 for the same domain, because the text type is used as part of Exim's
5724 per-process caching. But the chance of anyone hitting this buglet seems
5727 37. The dnsdb lookup has been extended in a number of ways.
5729 (1) There is a new type, "zns", which walks up the domain tree until it
5730 finds some nameserver records. It should be used with care.
5732 (2) There is a new type, "mxh", which is like "mx" except that it returns
5733 just the host names, not the priorities.
5735 (3) It is now possible to give a list of domains (or IP addresses) to be
5736 looked up. The behaviour when one of the lookups defers can be
5737 controlled by a keyword.
5739 (4) It is now possible to specify the separator character for use when
5740 multiple records are returned.
5742 38. The dnslists ACL condition has been extended: it is now possible to supply
5743 a list of IP addresses and/or domains to be looked up in a particular DNS
5746 39. Added log_selector=+queue_time_overall.
5748 40. When running the queue in the test harness, wait just a tad after forking a
5749 delivery process, to get repeatability of debugging output.
5751 41. Include certificate and key file names in error message when GnuTLS fails
5752 to set them up, because the GnuTLS error message doesn't include the name
5753 of the failing file when there is a problem reading it.
5755 42. Allow both -bf and -bF in the same test run.
5757 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5759 44. The "Exiscan patch" is now merged into the mainline Exim source.
5761 45. Sometimes the final signoff response after QUIT could fail to get
5762 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5763 < 0 before doing a fflush(). This bug looks as though it goes back to the
5764 introduction of TLS in release 3.20, but "sometimes" must have been rare
5765 because the tests only now provoked it.
5767 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5768 (this can affect the format of dates).
5770 47. exim_tidydb, when checking for the continued existence of a message for
5771 which it has found a message-specific retry record, was not finding
5772 messages that were in split spool directories. Consequently, it was
5773 deleting retry records that should have stayed in existence.
5775 48. Steve fixed some bugs in eximstats.
5777 49. The SPA authentication driver was not abandoning authentication and moving
5778 on to the next authenticator when an expansion was forced to fail,
5779 contradicting the general specification for all authenticators. Instead it
5780 was generating a temporary error. It now behaves as specified.
5782 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5783 (the order specifies the preference for clients). The order is now AES256,
5784 AES128, 3DES, ARCFOUR128.
5786 51. Small patch to Sieve code - explicitly set From: when generating an
5789 52. Exim crashed if a remote delivery caused a very long error message to be
5790 recorded - for instance if somebody sent an entire SpamAssassin report back
5791 as a large number of 550 error lines. This bug was coincidentally fixed by
5792 increasing the size of one of Exim's internal buffers (big_buffer) that
5793 happened as part of the Exiscan merge. However, to be on the safe side, I
5794 have made the code more robust (and fixed the comments that describe what
5797 53. Now that there can be additional text after "Completed" in log lines (if
5798 the queue_time_overall log selector is set), a one-byte patch to exigrep
5799 was needed to allow it to recognize "Completed" as not the last thing in
5802 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5803 patch that reportedly fixes this has been added. I am not expert enough to
5804 create a test for it. This is what the patch creator wrote:
5806 "I found a little strange behaviour of ldap code when working with
5807 Windows 2003 AD Domain, where users was placed in more than one
5808 Organization Units. When I tried to give exim partial DN, the exit code
5809 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5810 But simultaneously result of request was absolutely normal ldap result,
5811 so I produce this patch..."
5813 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5814 REFERENCE, so I have modified the code to exclude the patch when that macro
5817 55. Some experimental protocols are using DNS PTR records for new purposes. The
5818 keys for these records are domain names, not reversed IP addresses. The
5819 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5820 leaves it alone. Component reversal etc. now happens only for IP addresses.
5823 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5825 57. Double the size of the debug message buffer (to 2048) so that more of very
5826 long debug lines gets shown.
5828 58. The exicyclog utility now does better if the number of log files to keep
5829 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5831 59. Two changes related to the smtp_active_hostname option:
5833 (1) $smtp_active_hostname is now available as a variable.
5834 (2) The default for smtp_banner uses $smtp_active_hostname instead
5835 of $primary_hostname.
5837 60. The host_aton() function is supposed to be passed a string that is known
5838 to be a valid IP address. However, in the case of IPv6 addresses, it was
5839 not checking this. This is a hostage to fortune. Exim now panics and dies
5840 if the condition is not met. A case was found where this could be provoked
5841 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5842 components; fortuitously, this particular loophole had already been fixed
5843 by change 4.50/55 above.
5845 If there are any other similar loopholes, the new check in host_aton()
5846 itself should stop them being exploited. The report I received stated that
5847 data on the command line could provoke the exploit when Exim was running as
5848 exim, but did not say which command line option was involved. All I could
5849 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5850 running as the user.
5853 61. There was a buffer overflow vulnerability in the SPA authentication code
5854 (which came originally from the Samba project). I have added a test to the
5855 spa_base64_to_bits() function which I hope fixes it.
5858 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5859 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5861 63. The daemon start-up calls getloadavg() while still root for those OS that
5862 need the first call to be done as root, but it missed one case: when
5863 deliver_queue_load_max is set with deliver_drop_privilege. This is
5864 necessary for the benefit of the queue runner, because there is no re-exec
5865 when deliver_drop_privilege is set.
5867 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5868 This has been fixed.
5870 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5871 was in use, was not putting the data itself into the right store pool;
5872 consequently, it could be overwritten for a subsequent message in the same
5873 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5876 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5878 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5879 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5880 the function were treating the return as a boolean value, which happened to
5881 work because 0=false and not-0=true, but is not correct code.
5883 68. The host_aton() function was not handling scoped IPv6 addresses (those
5884 with, for example, "%eth0" on the end) correctly.
5886 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5887 negated items (that is, ~something) in unsigned ints. Some compilers
5888 apparently mutter when there is no cast.
5890 70. If an address verification called from an ACL failed, and did not produce a
5891 user-specific message (i.e. there was only a "system" message), nothing was
5892 put in $acl_verify_message. In this situation, it now puts the system
5895 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5896 session; change 4.31/43 added the unwanted input to the log line - except
5897 that it did not do this in the start of session case. It now does.
5899 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5900 This is wrong and can cause the other end to generate a synchronization
5901 error if it is another Exim or anything else that does the synchronization
5902 check. A QUIT command is no longer sent after a timeout.
5904 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5905 during host lookups.
5907 74. The defer_ok option of callout verification was not working if it was used
5908 when verifying addresses in header lines, that is, for this case:
5910 verify = header_sender/callout=defer_ok
5912 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5913 those file descriptors could be used for SMTP connections. If anything
5914 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5915 could be sent to the SMTP client, causing chaos. The daemon now opens
5916 stdin, stdout, and stderr to /dev/null when it puts itself into the
5919 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5920 log by default. The user can override this with suitable Perl magic.
5922 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5923 the log message when discard triggers, was not working for the DATA ACL or
5924 for the non-SMTP ACL.
5926 78. Error message wording change in sieve.c.
5928 79. If smtp_accept_max_per_host was set, the number of connections could be
5929 restricted to fewer than expected, because the daemon was trying to set up
5930 a new connection before checking whether the processes handling previous
5931 connections had finished. The check for completed processes is now done
5932 earlier. On busy systems, this bug wouldn't be noticed because something
5933 else would have woken the daemon, and it would have reaped the completed
5936 80. If a message was submitted locally by a user whose login name contained one
5937 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5938 It caused a spool format error. I have fixed the spool reading code. A
5939 related problem was that the "from" clause in the Received: line became
5940 illegal because of the space(s). It is now covered by ${quote_local_part.
5942 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5945 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5947 83. Patch from Sieve maintainer to fix unterminated string problem in
5948 "vacation" handling.
5950 84. Some minor changes to the Linux configuration files to help with other
5951 OS variants using glibc.
5953 85. One more patch for Sieve to update vacation handling to latest spec.
5956 ----------------------------------------------------
5957 See the note above about the 4.44 and 4.50 releases.
5958 ----------------------------------------------------
5964 1. Change 4.43/35 introduced a bug that caused file counts to be
5965 incorrectly computed when quota_filecount was set in an appendfile
5968 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5969 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5972 3. Give more explanation in the error message when the command for a transport
5973 filter fails to execute.
5975 4. There are several places where Exim runs a non-Exim command in a
5976 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5977 was being done only for the command run by the queryprogram router. It is
5978 now done for all such subprocesses. The other cases are: ${run, transport
5979 filters, and the commands run by the lmtp and pipe transports.
5981 5. Some older OS have a limit of 256 on the maximum number of file
5982 descriptors. Exim was using setrlimit() to set 1000 as a large value
5983 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5984 systems. I've change it so that if it can't get 1000, it tries for 256.
5986 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5987 was an oversight, and furthermore, ever since the addition of extra
5988 controls (e.g. 4.43/32), the checks on when to allow different forms of
5989 "control" were broken. There should now be diagnostics for all cases when a
5990 control that does not make sense is encountered.
5992 7. $recipients is now available in the predata ACL (oversight).
5994 8. Tidy the search cache before the fork to do a delivery from a message
5995 received from the command line. Otherwise the child will trigger a lookup
5996 failure and thereby defer the delivery if it tries to use (for example) a
5997 cached ldap connection that the parent has called unbind on.
5999 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6000 of $address_data from the recipient verification was clobbered by the
6001 sender verification.
6003 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6004 was its contents. (It was OK if the option was not defined at all.)
6006 11. A "Completed" log line is now written for messages that are removed from
6007 the spool by the -Mrm option.
6009 12. $host_address is now set to the target address during the checking of
6010 ignore_target_hosts.
6012 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6013 being passed; this would have caused $sender_host_name to have been used if
6014 matching the list had actually called for a host name (not very likely,
6015 since this list is usually IP addresses). A host name is now passed as
6018 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6019 code that specifies a non-restarting handler (typically sigaction() in
6020 modern systems) in an attempt to fix a rare and obscure crash bug.
6022 15. Narrowed the window for a race in the daemon that could cause it to ignore
6023 SIGCHLD signals. This is not a major problem, because they are used only to
6024 wake it up if nothing else does.
6026 16. A malformed maildirsize file could cause Exim to calculate negative values
6027 for the mailbox size or file count. Odd effects could occur as a result.
6028 The maildirsize information is now recalculated if the size or filecount
6031 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6032 support for a long time. Removed HAVE_SYS_VFS_H.
6034 18. Updated exipick to current release from John Jetmore.
6036 19. Allow an empty sender to be matched against a lookup in an address list.
6037 Previously the only cases considered were a regular expression, or an
6040 20. Exim went into a mad DNS lookup loop when doing a callout where the
6041 host was specified on the transport, if the DNS lookup yielded more than
6044 21. The RFC2047 encoding function was originally intended for short strings
6045 such as real names; it was not keeping to the 75-character limit for
6046 encoded words that the RFC imposes. It now respects the limit, and
6047 generates multiple encoded words if necessary. To be on the safe side, I
6048 have increased the buffer size for the ${rfc2047: expansion operator from
6051 22. Failure to deliver a bounce message always caused it to be frozen, even if
6052 there was an errors_to setting on the router. The errors_to setting is now
6055 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6056 canonical form (fully expanded) before being placed in
6057 $sender_host_address.
6059 24. Updated eximstats to version 1.33
6061 25. Include certificate and key file names in error message when GnuTLS fails
6062 to set them up, because the GnuTLS error message doesn't include the name
6063 of the failing file when there is a problem reading it.
6065 26. Expand error message when OpenSSL has problems setting up cert/key files.
6068 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6069 (this can affect the format of dates).
6071 28. exim_tidydb, when checking for the continued existence of a message for
6072 which it has found a message-specific retry record, was not finding
6073 messages that were in split spool directories. Consequently, it was
6074 deleting retry records that should have stayed in existence.
6076 29. eximstats updated to version 1.35
6077 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6078 1.35 - bugfix such that pie charts by volume are generated correctly
6080 30. The SPA authentication driver was not abandoning authentication and moving
6081 on to the next authenticator when an expansion was forced to fail,
6082 contradicting the general specification for all authenticators. Instead it
6083 was generating a temporary error. It now behaves as specified.
6085 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6086 (the order specifies the preference for clients). The order is now AES256,
6087 AES128, 3DES, ARCFOUR128.
6089 31. Small patch to Sieve code - explicitly set From: when generating an
6092 32. Exim crashed if a remote delivery caused a very long error message to be
6093 recorded - for instance if somebody sent an entire SpamAssassin report back
6094 as a large number of 550 error lines. This bug was coincidentally fixed by
6095 increasing the size of one of Exim's internal buffers (big_buffer) that
6096 happened as part of the Exiscan merge. However, to be on the safe side, I
6097 have made the code more robust (and fixed the comments that describe what
6100 33. Some experimental protocols are using DNS PTR records for new purposes. The
6101 keys for these records are domain names, not reversed IP addresses. The
6102 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6103 leaves it alone. Component reversal etc. now happens only for IP addresses.
6106 34. The host_aton() function is supposed to be passed a string that is known
6107 to be a valid IP address. However, in the case of IPv6 addresses, it was
6108 not checking this. This is a hostage to fortune. Exim now panics and dies
6109 if the condition is not met. A case was found where this could be provoked
6110 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6111 components; fortuitously, this particular loophole had already been fixed
6112 by change 4.50/55 or 4.44/33 above.
6114 If there are any other similar loopholes, the new check in host_aton()
6115 itself should stop them being exploited. The report I received stated that
6116 data on the command line could provoke the exploit when Exim was running as
6117 exim, but did not say which command line option was involved. All I could
6118 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6119 running as the user.
6122 35. There was a buffer overflow vulnerability in the SPA authentication code
6123 (which came originally from the Samba project). I have added a test to the
6124 spa_base64_to_bits() function which I hope fixes it.
6127 36. The daemon start-up calls getloadavg() while still root for those OS that
6128 need the first call to be done as root, but it missed one case: when
6129 deliver_queue_load_max is set with deliver_drop_privilege. This is
6130 necessary for the benefit of the queue runner, because there is no re-exec
6131 when deliver_drop_privilege is set.
6133 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6134 was in use, was not putting the data itself into the right store pool;
6135 consequently, it could be overwritten for a subsequent message in the same
6136 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6139 38. Sometimes the final signoff response after QUIT could fail to get
6140 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6141 < 0 before doing a fflush(). This bug looks as though it goes back to the
6142 introduction of TLS in release 3.20, but "sometimes" must have been rare
6143 because the tests only now provoked it.
6149 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6150 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6151 Now it returns an integer. A number of calls were still expecting a T/F
6152 return. Fortuitously, in all cases, the tests worked in OK situations,
6153 which is the norm. However, things would have gone wrong on any write
6154 failures on the smtp file descriptor. This function is used when sending
6155 messages over SMTP and also when doing verify callouts.
6157 2. When Exim is called to do synchronous delivery of a locally submitted
6158 message (the -odf or -odi options), it no longer closes stderr before doing
6161 3. Implemented the mua_wrapper option.
6163 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6165 5. Implemented the functions header_remove(), header_testname(),
6166 header_add_at_position(), and receive_remove_recipient(), and exported them
6169 6. If an ACL "warn" statement specified the addition of headers, Exim already
6170 inserted X-ACL-Warn: at the start if there was no header name. However, it
6171 was not making this test for the second and subsequent header lines if
6172 there were newlines in the string. This meant that an invalid header could
6173 be inserted if Exim was badly configured.
6175 7. Allow an ACL "warn" statement to add header lines at the start or after all
6176 the Received: headers, as well as at the end.
6178 8. Added the rcpt_4xx retry error code.
6180 9. Added postmaster_mailfrom=xxx to callout verification option.
6182 10. Added mailfrom=xxxx to the callout verification option, for verify=
6185 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6186 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6188 12. Inserted some casts to stop certain compilers warning when using pointer
6189 differences as field lengths or precisions in printf-type calls (mostly
6190 affecting debugging statements).
6192 13. Added optional readline() support for -be (dynamically loaded).
6194 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6195 same clock tick as a message's arrival, so that its received time was the
6196 same as the "first fail" time on the retry record, and that message
6197 remained on the queue past the ultimate address timeout, every queue runner
6198 would try a delivery (because it was past the ultimate address timeout) but
6199 after another failure, the ultimate address timeout, which should have then
6200 bounced the address, did not kick in. This was a "< instead of <=" error;
6201 in most cases the first failure would have been in the next clock tick
6202 after the received time, and all would be well.
6204 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6205 being recognized when the domain list was tested by the match_domain
6206 condition in an expansion string.
6208 16. Added the ${str2b64: operator.
6210 17. Exim was always calling setrlimit() to set a large limit for the number of
6211 processes, without checking whether the existing limit was already
6212 adequate. (It did check for the limit on file descriptors.) Furthermore,
6213 errors from getrlimit() and setrlimit() were being ignored. Now they are
6214 logged to the main and panic logs, but Exim does carry on, to try to do its
6215 job under whatever limits there are.
6217 18. Imported PCRE 5.0.
6219 19. Trivial typo in log message " temporarily refused connection" (the leading
6222 20. If the log selector return_path_on_delivery was set and an address was
6223 redirected to /dev/null, the delivery process crashed because it assumed
6224 that a return path would always be set for a "successful" delivery. In this
6225 case, the whole delivery is bypassed as an optimization, and therefore no
6228 21. Internal re-arrangement: the function for sending a challenge and reading
6229 a response while authentication was assuming a zero-terminated challenge
6230 string. It's now changed to take a pointer and a length, to allow for
6231 binary data in such strings.
6233 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6235 23. Exim was not respecting finduser_retries when seeking the login of the
6236 uid under which it was called; it was always trying 10 times. (The default
6237 setting of finduser_retries is zero.) Also, it was sleeping after the final
6238 failure, which is pointless.
6240 24. Implemented tls_on_connect_ports.
6242 25. Implemented acl_smtp_predata.
6244 26. If the domain in control=submission is set empty, Exim assumes that the
6245 authenticated id is a complete email address when it generates From: or
6246 Sender: header lines.
6248 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6249 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6250 chown and chgrp in /bin and hostname in /usr/bin.
6252 28. Exim was keeping the "process log" file open after each use, just as it
6253 does for the main log. This opens the possibility of it remaining open for
6254 long periods when the USR1 signal hits a daemon. Occasional processlog
6255 errors were reported, that could have been caused by this. Anyway, it seems
6256 much more sensible not to leave this file open at all, so that is what now
6259 29. The long-running daemon process does not normally write to the log once it
6260 has entered its main loop, and it closes the log before doing so. This is
6261 so that log files can straightforwardly be renamed and moved. However,
6262 there are a couple of unusual error situations where the daemon does write
6263 log entries, and I had neglected to close the log afterwards.
6265 30. The text of an SMTP error response that was received during a remote
6266 delivery was being truncated at 512 bytes. This is too short for some of
6267 the long messages that one sometimes sees. I've increased the limit to
6270 31. It is now possible to make retry rules that apply only when a message has a
6271 specific sender, in particular, an empty sender.
6273 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6274 it possible to be selective about when SMTP synchronization is enforced.
6276 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6278 32. Implemented hosts_connection_nolog.
6280 33. Added an ACL for QUIT.
6282 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6285 35. Added mailbox_size and mailbox_filecount to appendfile.
6287 36. Added control = no_multiline_responses to ACLs.
6289 37. There was a bug in the logic of the code that waits for the clock to tick
6290 in the case where the clock went backwards by a substantial amount such
6291 that the microsecond fraction of "now" was more than the microsecond
6292 fraction of "then" (but the whole seconds number was less).
6294 38. Added support for the libradius Radius client library this is found on
6295 FreeBSD (previously only the radiusclient library was supported).
6301 1. When certain lookups returned multiple values in the form name=value, the
6302 quoting of the values was not always being done properly. Specifically:
6303 (a) If the value started with a double quote, but contained no whitespace,
6305 (b) If the value contained whitespace other than a space character (i.e.
6306 tabs or newlines or carriage returns) it was not quoted.
6307 This fix has been applied to the mysql and pgsql lookups by writing a
6308 separate quoting function and calling it from the lookup code. The fix
6309 should probably also be applied to nisplus, ibase and oracle lookups, but
6310 since I cannot test any of those, I have not disturbed their existing code.
6312 2. A hit in the callout cache for a specific address caused a log line with no
6313 reason for rejecting RCPT. Now it says "Previous (cached) callout
6314 verification failure".
6316 3. There was an off-by-one bug in the queryprogram router. An over-long
6317 return line was truncated at 256 instead of 255 characters, thereby
6318 overflowing its buffer with the terminating zero. As well as fixing this, I
6319 have increased the buffer size to 1024 (and made a note to document this).
6321 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6322 when Exim is waiting for an SMTP response from a remote server, Exim
6323 restarts its select() call on the socket, thereby resetting its timeout.
6324 This is not a problem when such interrupts are rare. Somebody set up a cron
6325 job to run exiwhat every 2 minutes, which is less than the normal select()
6326 timeout (5 or 10 minutes). This meant that the select() timeout never
6327 kicked in because it was always reset. I have fixed this by comparing the
6328 time when an interrupt arrives with the time at the start of the first call
6329 to select(). If more time than the timeout has elapsed, the interrupt is
6330 treated as a timeout.
6332 5. Some internal re-factoring in preparation for the addition of Sieve
6333 extensions (by MH). In particular, the "personal" test is moved to a
6334 separate function, and given an option for scanning Cc: and Bcc: (which is
6335 not set for Exim filters).
6337 6. When Exim created an email address using the login of the caller as the
6338 local part (e.g. when creating a From: or Sender: header line), it was not
6339 quoting the local part when it contained special characters such as @.
6341 7. Installed new OpenBSD configuration files.
6343 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6344 try to make them clearer.
6346 9. Callout options, other than the timeout value, were being ignored when
6347 verifying sender addresses in header lines. For example, when using
6349 verify = header_sender/callout=no_cache
6351 the cache was (incorrectly) being used.
6353 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6354 only the Cygwin environment.
6356 11. When return_path_on_delivery was set as a log selector, if different remote
6357 addresses in the same message used different return paths and parallel
6358 remote delivery occurred, the wrong values would sometimes be logged.
6359 (Whenever a remote delivery process finished, the return path value from
6360 the most recently started remote delivery process was logged.)
6362 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6363 lines when AUTH and/or TLS are in use. This is the "received protocol"
6364 field. Exim used to use "asmtp" for authenticated SMTP, without any
6365 indication (in the protocol name) for TLS use. Now it follows the RFC and
6366 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6367 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6368 names appear in log lines as well as in Received: header lines.
6370 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6371 extensions, and comparison tests, and to fix some bugs.
6373 14. Changes to the "personal" filter test:
6375 (1) The test was buggy in that it was just doing the equivalent of
6376 "contains" tests on header lines. For example, if a user's address was
6377 anne@some.where, the "personal" test would incorrectly be true for
6379 To: susanne@some.where
6381 This test is now done by extracting each address from the header in turn,
6382 and checking the entire address. Other tests that are part of "personal"
6383 are now done using regular expressions (for example, to check local parts
6384 of addresses in From: header lines).
6386 (2) The list of non-personal local parts in From: addresses has been
6387 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6388 taken from the Sieve specification recommendations.
6390 (3) If the message contains any header line starting with "List-" it is
6391 treated as non-personal.
6393 (4) The test for "circular" in the Subject: header line has been removed
6394 because it now seems ill-conceived.
6396 15. Minor typos in src/EDITME comments corrected.
6398 16. Installed latest exipick from John Jetmore.
6400 17. If headers_add on a router specified a text string that was too long for
6401 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6402 of string_sprintf() is now avoided.
6404 18. $message_body_size was not set (it was always zero) when running the DATA
6405 ACL and the local_scan() function.
6407 19. For the "mail" command in an Exim filter, no default was being set for
6408 the once_repeat time, causing a random time value to be used if "once" was
6409 specified. (If the value happened to be <= 0, no repeat happened.) The
6410 default is now 0s, meaning "never repeat". The "vacation" command was OK
6411 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6412 (I found it when inspecting the code).
6414 20. There is now an overall timeout for performing a callout verification. It
6415 defaults to 4 times the callout timeout, which applies to individual SMTP
6416 commands during the callout. The overall timeout applies when there is more
6417 than one host that can be tried. The timeout is checked before trying the
6418 next host. This prevents very long delays if there are a large number of
6419 hosts and all are timing out (e.g. when the network connections are timing
6420 out). The value of the overall timeout can be changed by specifying an
6421 additional sub-option for "callout", called "maxwait". For example:
6423 verify = sender/callout=5s,maxwait=20s
6425 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6426 to the end before writing, but this should make it even safer).
6428 22. Exim was forgetting that it had advertised PIPELINING for the second and
6429 subsequent messages on an SMTP connection. It was also not resetting its
6430 memory on STARTTLS and an internal HELO.
6432 23. When Exim logs an SMTP synchronization error within a session, it now
6433 records whether PIPELINING has been advertised or not.
6435 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6436 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6437 rather than long int.
6439 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6441 26. Added the never_mail option to autoreply.
6447 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6448 crash if the getsockname() call failed; this can happen if a connection is
6449 closed very soon after it is established. The problem was simply in the
6450 order in which certain operations were done, causing Exim to try to write
6451 to the SMTP stream before it had set up the file descriptor. The bug has
6452 been fixed by making things happen in the correct order.
6458 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6459 before the connection was closed, thus losing the rejection response.
6461 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6462 some early Solaris releases, but causes trouble in current releases where
6463 socklen_t is defined.
6465 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6468 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6471 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6472 that releasing the top of it at the end releases what was used for sub-
6473 expansions (unless the block got too big). However, discard this block if
6474 the first thing is a variable or header, so that we can use its block when
6475 it is dynamic (useful for very large $message_headers, for example).
6477 6. Lookups now cache *every* query, not just the most recent. A new, separate
6478 store pool is used for this. It can be recovered when all lookup caches are
6479 flushed. Lookups now release memory at the end of their result strings.
6480 This has involved some general refactoring of the lookup sources.
6482 7. Some code has been added to the store_xxx() functions to reduce the amount
6483 of flapping under certain conditions.
6485 8. log_incoming_interface used to affect only the <= reception log lines. Now
6486 it causes the local interface and port to be added to several more SMTP log
6487 lines, for example "SMTP connection from", and rejection lines.
6489 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6491 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6493 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6494 could be overwritten at the end of the current message (or the start of a
6495 new message if it was set in a HELO ACL). The value is now preserved for
6496 the duration of the SMTP connection.
6498 12. If a transport had a headers_rewrite setting, and a matching header line
6499 contained an unqualified address, that address was qualified, even if it
6500 did not match any rewriting rules. The underlying bug was that the values
6501 of the flags that permit the existence of unqualified sender and recipient
6502 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6503 non-local messages, and by -bnq for local messages) were not being
6504 preserved with the message after it was received.
6506 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6507 "next input=" as part of the text comprising the host identity instead of
6508 the correct text. The code was using the same buffer for two different
6509 strings. However, depending on which order the printing function evaluated
6510 its arguments, the bug did not always show up. Under Linux, for example, my
6511 test suite worked just fine.
6513 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6514 doesn't work with some older versions of Perl. It has been changed to "my",
6515 which in any case is probably the better facility to use.
6517 15. A really picky compiler found some instances of statements for creating
6518 error messages that either had too many or two few arguments for the format
6521 16. The size of the buffer for calls to the DNS resolver has been increased
6522 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6523 for addresses that have a lot of PTR records. This alleviates a problem; it
6524 does not fully solve it.
6526 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6527 buffer now truncates the list and logs the incident, which is the same
6528 action as happens when Exim is looking up a host name and its aliases.
6529 Previously in this situation something unpredictable would happen;
6530 sometimes it was "internal error: store_reset failed".
6532 18. If a server dropped the connection unexpectedly when an Exim client was
6533 using GnuTLS and trying to read a response, the client delivery process
6534 crashed while trying to generate an error log message.
6536 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6537 string, for example:
6539 warn message = H1: something\nH2: something
6541 the text was added as a single header line from Exim's point of view
6542 though it ended up OK in the delivered message. However, searching for the
6543 second and subsequent header lines using $h_h2: did not work. This has been
6544 fixed. Similarly, if a system filter added multiple headers in this way,
6545 the routers could not see them.
6547 20. Expanded the error message when iplsearch is called with an invalid key to
6548 suggest using net-iplsearch in a host list.
6550 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6551 ACLs are no longer actually imposed (and a message to that effect is
6554 22. If a "gecos" field in a passwd entry contained escaped characters, in
6555 particular, if it contained a \" sequence, Exim got it wrong when building
6556 a From: or a Sender: header from that name. A second bug also caused
6557 incorrect handling when an unquoted " was present following a character
6558 that needed quoting.
6560 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6561 was not being matched caselessly.
6563 24. Arranged for all hyphens in the exim.8 source to be escaped with
6566 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6567 properly. Recipient callout cache records were still being keyed to include
6568 the sender, even when use_sender was set false. This led to far more
6569 callouts that were necessary. The sender is no longer included in the key
6570 when use_sender is false.
6572 26. Added "control = submission" modifier to ACLs.
6574 27. Added the ${base62d: operator to decode base 62 numbers.
6576 28. dnsdb lookups can now access SRV records.
6578 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6579 the configuration file.
6581 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6582 (-v) mode. This makes the output for a verbose queue run more intelligible.
6584 31. Added a use_postmaster feature to recipient callouts.
6586 32. Added the $body_zerocount variable, containing the number of binary zero
6587 bytes in the message body.
6589 33. The time of last modification of the "new" subdirectory is now used as the
6590 "mailbox time last read" when there is a quota error for a maildir
6593 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6595 35. Added +ignore_unknown as a special item in host lists.
6597 36. Code for decoding IPv6 addresses in host lists is now included, even if
6598 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6599 address was recognized as an IP address, but was then not correctly decoded
6600 into binary, causing unexpected and incorrect effects when compared with
6607 1. Very minor rewording of debugging text in manualroute to say "list of
6608 hosts" instead of "hostlist".
6610 2. If verify=header_syntax was set, and a header line with an unqualified
6611 address (no domain) and a large number of spaces between the end of the
6612 name and the colon was received, the reception process suffered a buffer
6613 overflow, and (when I tested it) crashed. This was caused by some obsolete
6614 code that should have been removed. The fix is to remove it!
6616 3. When running in the test harness, delay a bit after writing a bounce
6617 message to get a bit more predictability in the log output.
6619 4. Added a call to search_tidyup() just before forking a reception process. In
6620 theory, someone could use a lookup in the expansion of smtp_accept_max_
6621 per_host which, without the tidyup, could leave open a database connection.
6623 5. Added the variables $recipient_data and $sender_data which get set from a
6624 lookup success in an ACL "recipients" or "senders" condition, or a router
6625 "senders" option, similar to $domain_data and $local_part_data.
6627 6. Moved the writing of debug_print from before to after the "senders" test
6630 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6631 problems for message scanning, either using a data ACL, or using
6632 local_scan() because the Received: header was not generated till after they
6633 were called (in order to set the time as the time of reception completion).
6634 I have revised the way this works. The header is now generated after the
6635 body is received, but before the ACL or local_scan() are called. After they
6636 are run, the timestamp in the header is updated.
6642 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6643 before starting a queue runner without re-exec. This happened only when
6644 deliver_drop_privilege was set or when the Exim user was set to root. The
6645 effect of the bug was that timeouts during subsequent deliveries caused
6646 crashes instead of being properly handled. The handler is now left at its
6647 default (and expected) setting.
6649 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6650 message, again when deliver_drop_privilege is set or Exim is run as root.
6651 The bug described in (1) was not present in this case, but the tidying up
6652 of the other signals was missing. I have made the two cases consistent.
6654 3. The ignore_target_hosts setting on a manualroute router was being ignored
6655 for hosts that were looked up using the /MX notation.
6657 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6660 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6661 operated on the sender address. After changing the $sender_address to <>
6662 for the sender address verify, Exim was re-instated it as the original
6663 (before rewriting) address, but remembering that it had rewritten it, so it
6664 wasn't rewriting it again. This bug also had the effect of breaking the
6665 sender address verification caching when the sender address was rewritten.
6667 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6668 This has been changed so that if the ip literal address matches
6669 ignore_target_hosts, the router declines.
6671 7. Added expansion conditions match_domain, match_address, and match_local_
6672 part (NOT match_host).
6674 8. The placeholder for the Received: header didn't have a length field set.
6676 9. Added code to Exim itself and to exim_lock to test for a specific race
6677 condition that could lead to file corruption when using MBX delivery. The
6678 issue is with the lockfile that is created in /tmp. If this file is removed
6679 after a process has opened it but before that process has acquired a lock,
6680 there is the potential for a second process to recreate the file and also
6681 acquire a lock. This could lead to two Exim processes writing to the file
6682 at the same time. The added code performs the same test as UW imapd; it
6683 checks after acquiring the lock that its file descriptor still refers to
6684 the same named file.
6686 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6687 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6690 11. Added the smtp_active_hostname option. If used, this will typically be made
6691 to depend on the incoming interface address. Because $interface_address is
6692 not set up until the daemon has forked a reception process, error responses
6693 that can happen earlier (such as "too many connections") no longer contain
6696 12. If an expansion in a condition on a "warn" statement fails because a lookup
6697 defers, the "warn" statement is abandoned, and the next ACL statement is
6698 processed. Previously this caused the whole ACL to be aborted.
6700 13. Added the iplsearch lookup type.
6702 14. Added ident_timeout as a log selector.
6704 15. Added tls_certificate_verified as a log selector.
6706 16. Added a global option tls_require_ciphers (compare the smtp transport
6707 option of the same name). This controls incoming TLS connections.
6709 17. I finally figured out how to make tls_require_ciphers do a similar thing
6710 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6711 before starting the TLS session.
6713 18. Tabs are now shown as \t in -bP output.
6715 19. If the log selector return_path_on_delivery was set, Exim crashed when
6716 bouncing a message because it had too many Received: header lines.
6718 20. If two routers both had headers_remove settings, and the first one included
6719 a superfluous trailing colon, the final name in the first list and the
6720 first name in the second list were incorrectly joined into one item (with a
6721 colon in the middle).
6727 1. Added -C and -D options to the exinext utility, mainly to make it easier
6728 to include in the automated testing, but these could be helpful when
6729 multiple configurations are in use.
6731 2. The exinext utility was not formatting the output nicely when there was
6732 an alternate port involved in the retry record key, nor when there was a
6733 message id as well (for retries that were specific to a specific message
6734 and a specific host). It was also confused by IPv6 addresses, because of
6735 the additional colons they contain. I have fixed the IPv4 problem, and
6736 patched it up to do a reasonable job for IPv6.
6738 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6739 delivery, the log line now contains "pipelined" if PIPELINING was used.
6741 4. An SMTP transport process used to panic and die if the bind() call to set
6742 an explicit outgoing interface failed. This has been changed; it is now
6743 treated in the same way as a connect() failure.
6745 5. A reference to $sender_host_name in the part of a conditional expansion
6746 that was being skipped was still causing a DNS lookup. This no longer
6749 6. The def: expansion condition was not recognizing references to header lines
6750 that used bh_ and bheader_.
6752 7. Added the _cache feature to named lists.
6754 8. The code for checking quota_filecount in the appendfile transport was
6755 allowing one more file than it should have been.
6757 9. For compatibility with Sendmail, the command line option
6765 and sets the incoming protocol and host name (for trusted callers). The
6766 host name and its colon can be omitted when only the protocol is to be set.
6767 Note the Exim already has two private options, -pd and -ps, that refer to
6768 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6769 "s", but I don't think that's a major issue.
6771 10. A number of refactoring changes to the code, none of which should affect
6774 (a) The number of logging options was getting close to filling up the
6775 32-bit word that was used as a bit map. I have split them into two classes:
6776 those that are passed in the argument to log_write(), and those that are
6777 only ever tested independently outside of that function. These are now in
6778 separate 32-bit words, so there is plenty of room for expansion again.
6779 There is no change in the user interface or the logging behaviour.
6781 (b) When building, for example, log lines, the code previously used a
6782 macro that called string_cat() twice, in order to add two strings. This is
6783 not really sufficiently general. Furthermore, there was one instance where
6784 it was actually wrong because one of the argument was used twice, and in
6785 one call a function was used. (As it happened, calling the function twice
6786 did not affect the overall behaviour.) The macro has been replaced by a
6787 function that can join an arbitrary number of extra strings onto a growing
6790 (c) The code for expansion conditions now uses a table and a binary chop
6791 instead of a serial search (which was left over from when there were very
6792 few conditions). Also, it now recognizes conditions like "pam" even when
6793 the relevant support is not compiled in: a suitably worded error message is
6794 given if an attempt is made to use such a condition.
6796 11. Added ${time_interval:xxxxx}.
6798 12. A bug was causing one of the ddress fields not to be passed back correctly
6799 from remote delivery subprocesses. The field in question was not being
6800 subsequently used, so this caused to problems in practice.
6802 13. Added new log selectors queue_time and deliver_time.
6804 14. Might have fixed a bug in maildirsizefile handling that threw up
6805 "unexpected character" debug warnings, and recalculated the data
6806 unnecessarily. In any case, I expanded the warning message to give more
6809 15. Added the message "Restricted characters in address" to the statements in
6810 the default ACL that block characters like @ and % in local parts.
6812 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6813 Three changes have been made:
6815 (a) There was a serious bug; a negative response to MAIL caused the whole
6816 recipient domain to be cached as invalid, thereby blocking all messages
6817 to all local parts at the same domain, from all senders. This bug has
6818 been fixed. The domain is no longer cached after a negative response to
6819 MAIL if the sender used is not empty.
6821 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6824 (c) A new callout option, "use_sender" has been added for people who want
6825 the modified behaviour.
6831 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6834 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6835 indeed breaks things for older releases.
6837 3. Added additional logging to the case where there is a problem reading data
6838 from a filter that is running in a subprocess using a pipe, in order to
6839 try to track down a specific problem.
6841 4. Testing facility fudge: when running in the test harness and attempting
6842 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6843 getting "No route to host". Convert this to a timeout.
6845 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6848 6. Some OS don't have socklen_t but use size_t instead. This affects the
6849 fifth argument of getsockopt() amongst other things. This is now
6850 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6851 can be set for individual OS. I have set it for SunOS5, OSF1, and
6852 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6853 some earlier ones do not.
6855 7. Change 4.30/15 was not doing the test caselessly.
6857 8. The standard form for an IPv6 address literal was being rejected by address
6858 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6859 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6860 this, as well as the form without the "IPv6" on the front (but only when
6861 address literals are enabled, of course).
6863 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6865 10. Exim crashed if a message with an empty sender address specified by -f
6866 encountered a router with an errors_to setting. This could be provoked only
6867 by a command such as
6871 where an empty string was supplied; "<>" did not hit this bug.
6873 11. Installed PCRE release 4.5.
6875 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6876 remained set. It is now erased.
6878 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6879 times from message ids (which are base 36 rather than the normal 62).
6881 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6882 were being counted as actual protocol errors, and logged if the log
6883 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6884 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6885 and DATA following a set of rejected RCPTs do not count as protocol errors.
6886 In other words, Exim assumes they were pipelined, though this may not
6887 actually be the case. Of course, in all cases the client gets an
6888 appropriate error code.
6890 15. If a lookup fails in an ACL condition, a message about the failure may
6891 be available; it is used if testing the ACL cannot continue, because most
6892 such messages specify what the cause of the deferral is. However, some
6893 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6894 that caused an old message to be retained and used if a later statement
6895 caused a defer, replacing the real cause of the deferral.
6897 16. If an IP address had so many PTR records that the DNS lookup buffer
6898 was not large enough to hold them, Exim could crash while trying to process
6899 the truncated data. It now detects and logs this case.
6901 17. Further to 4.21/58, another change has been made: if (and only if) the
6902 first line of a message (the first header line) ends with CRLF, a bare LF
6903 in a subsequent header line has a space inserted after it, so as not to
6904 terminate the header.
6906 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6907 unnecessarily, used atoi() instead of strtol(), and didn't check the
6908 termination when getting file sizes from file names by regex.
6910 19. Completely re-implemented the support for maildirsize files, in the light
6911 of a number of problems with the previous contributed implementation
6912 (4.30/29). In particular:
6914 . If the quota is zero, the maildirsize file is maintained, but no quota is
6917 . If the maildir directory does not exist, it is created before any attempt
6918 to write a maildirsize file.
6920 . The quota value in the file is just a cache; if the quota is changed in
6921 the transport, the new value overrides.
6923 . A regular expression is available for excluding directories from the
6926 20. The autoreply transport checks the characters in options that define the
6927 message's headers; it allows continued headers, but it was checking with
6928 isspace() after an embedded newline instead of explicitly looking for a
6931 21. If all the "regular" hosts to which an address was routed had passed their
6932 expiry times, and had not reached their retry times, the address was
6933 bounced, even if fallback hosts were defined. Now Exim should go on to try
6936 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6937 equivalent code in the SMTP transport. Some hosts send humungous responses
6938 to HELO/EHLO, more than 1024 it seems.
6940 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6941 gives compiler warnings in some environments. I've now done it "properly",
6944 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6945 (because of problems with the built-in one) was declared to return uschar *
6946 instead of char *, causing compiler failure.
6948 25. Fixed a file descriptor leak when processing alias/forward files.
6950 26. Fixed a minor format string issue in dbfn.c.
6952 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6954 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6955 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6958 29. When a sender address is verified, it is cached, to save repeating the test
6959 when there is more than one recipient in a message. However, when the
6960 verification involves a callout, it is possible for different callout
6961 options to be set for different recipients. It is too complicated to keep
6962 track of this in the cache, so now Exim always runs a verification when a
6963 callout is required, relying on the callout cache for the optimization.
6964 The overhead is duplication of the address routing, but this should not be
6967 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6968 to be verified with callout=postmaster, and the main callout worked but the
6969 postmaster check failed, the verification correctly failed. However, if a
6970 subsequent RCPT command asked for sender verification *without* the
6971 postmaster check, incorrect caching caused this verification also to fail,
6974 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6975 it was not caching the DNS options (qualify_single, search_parents) that
6976 were used when the lookup failed. A subsequent lookup with different
6977 options therefore always gave the same answer, though there were cases
6978 where it should not have. (Example: a "domains = !$mx_any" option on a
6979 dnslookup router: the "domains" option is always processed without any
6980 widening, but the router might have qualify_single set.) Now Exim uses the
6981 cached value only when the same options are set.
6983 32. Added John Jetmore's "exipick" utility to the distribution.
6985 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6986 than a timeout (e.g. a certificate is required, and is not provided), an
6987 Exim server now closes the connection immediately. Previously it waited for
6988 the client to close - but if the client is SSL, it seems that they each
6989 wait for each other, leading to a delay before one of them times out.
6991 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6992 maintained 0.8.x compatibility because I don't think many are using it, and
6993 it is clearly obsolete.
6995 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6998 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6999 client certificate was expired. A simple patch fixes this, though I don't
7000 understand the full logic of why the verify callback is called multiple
7003 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7004 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7005 which causes problems with some clients (such as the Certicom SSL Plus
7006 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7007 disables the coutermeasure allowing Eudora to connect."
7009 38. Exim was not checking that a write() to a log file succeeded. This could
7010 lead to Bad Things if a log got too big, in particular if it hit a file
7011 size limit. Exim now panics and dies if it cannot write to a log file, just
7012 as it does if it cannot open a log file.
7014 39. Modified OS/Makefile-Linux so that it now contains
7016 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7018 The two -D definitions ensure that Exim is compiled with large file
7019 support, which makes it possible to handle log files that are bigger than
7022 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7023 instance) a domain was checked against a named list that involved a lookup,
7024 causing $domain_data to be set, then another domain was checked against the
7025 same list, then the first domain was re-checked, the value of $domain_data
7026 after the final check could be wrong. In particular, if the second check
7027 failed, it could be set empty. This bug probably also applied to
7030 41. The strip_trailing_dot option was not being applied to the address given
7031 with the -f command-line option.
7033 42. The code for reading a message's header from the spool was incrementing
7034 $received_count, but never initializing it. This meant that the value was
7035 incorrect (doubled) while delivering a message in the same process in which
7036 it was received. In the most common configuration of Exim, this never
7037 happens - a fresh exec is done - but it can happen when
7038 deliver_drop_privilege is set.
7040 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7041 it now includes up to 150 characters of the unexpected data in the log
7044 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7045 and building data strings. The size of both of these buffers was 10 000
7046 bytes - far larger than anybody would *ever* want, thought I. Needless to
7047 say, somebody hit the limit. I have increased the maximum line length to
7048 20 000 and the maximum data length of concatenated lines to 100 000. I have
7049 also fixed two bugs, because there was no checking on these buffers. Tsk,
7050 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7051 buffer is too small.
7053 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7054 lsearch lookups. Now it does.
7056 46. When parsing a route_list item in a manualroute router, a fixed-length
7057 buffer was used for the list of hosts. I made this 1024 bytes long,
7058 thinking that nobody would ever have a list of hosts that long. Wrong.
7059 Somebody had a whole pile of complicated expansion conditions, and the
7060 string was silently truncated, leading to an expansion error. It turns out
7061 that it is easier to change to an unlimited length (owing to other changes
7062 that have happened since this code was originally written) than to build
7063 structure for giving a limitation error. The length of the item that
7064 expands into the list of hosts is now unlimited.
7066 47. The lsearch lookup could not handle data where the length of text line was
7067 more than 4095 characters. Such lines were truncated, leading to shortened
7068 data being returned. It should now handle lines of any length.
7070 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7071 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7074 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7075 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7076 no longer happens in comment lines. A list of replacements is now placed
7077 at the head of all of the source files, except those whose only change is
7078 to replace PERL_COMMAND in the very first #! line.
7080 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7081 messages on the queue, with a bottom-up merge sort, using code contributed
7082 by Michael Haardt. This should make operations like -bp somewhat faster on
7083 large queues. It won't affect queue runners, except when queue_run_in_order
7086 51. Installed eximstats 1.31 in the distribution.
7088 52. Added support for SRV lookups to the dnslookup router.
7090 53. If an ACL referred to $message_body or $message_body_end, the value was not
7091 reset for any messages that followed in the same SMTP session.
7093 54. The store-handling optimization for building very long strings was not
7094 differentiating between the different store pools. I don't think this
7095 actually made any difference in practice, but I've tidied it.
7097 55. While running the routers to verify a sender address, $sender_address
7098 was still set to the sender address. This is wrong, because when routing to
7099 send a bounce to the sender, it would be empty. Therefore, I have changed
7100 it so that, while verifying a sender address, $sender_address is set to <>.
7101 (There is no change to what happens when verifying a recipient address.)
7103 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7104 target A or AAAA records (if not already returned) without resetting the
7105 qualify_single or search_parents options of the DNS resolver. These are
7106 inappropriate in this case because the targets of MX and SRV records must
7107 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7108 target that, when qualified, matched something in the local domain. These
7109 two options are now turned off when doing these lookups.
7111 57. It seems that at least some releases of Reiserfs (which does not have the
7112 concept of a fixed number of inodes) returns zero and not -1 for the
7113 number of available inodes. This interacted badly with check_spool_inodes,
7114 which assumed that -1 was the "no such thing" setting. What I have done is
7115 to check that the total number of inodes is greater than zero before doing
7116 the test of how many are available.
7118 58. When a "warn" ACL statement has a log_message modifier, the message is
7119 remembered, and not repeated. This is to avoid a lot of repetition when a
7120 message has many recipients that cause the same warning to be written.
7121 However, Exim was preserving the list of already written lines for an
7122 entire SMTP session, which doesn't seem right. The memory is now reset if a
7123 new message is started.
7125 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7126 debugging output unless log_rewrite was also set.
7128 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7129 of (int)(handle) when we know that handle contains (void *)(-1).
7131 61. The Exim daemon panic-logs an error return when it closes the incoming
7132 connection. However "connection reset by peer" seems to be common, and
7133 isn't really an error worthy of noting specially, so that particular error
7136 62. When Exim is trying to find all the local interfaces, it used to panic and
7137 die if the ioctl to get the interface flags failed. However, it seems that
7138 on at least one OS (Solaris 9) it is possible to have an interface that is
7139 included in the list of interfaces, but for which you get a failure error
7140 for this call. This happens when the interface is not "plumbed" into a
7141 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7142 failure of the "get flags" call assumes that the interface is down.
7144 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7145 makes life easier for people who are doing arithmetic on fields extracted
7146 from dates, where you often get leading zeros that should not be
7147 interpreted as octal.
7149 64. Added qualify_domain to the redirect router, to override the global
7152 65. If a pathologically long header line contained very many addresses (the
7153 report of this problem mentioned 10 000) and each of them was rewritten,
7154 Exim could use up a very large amount of memory. (It kept on making new
7155 copies of the header line as it rewrote, and never released the old ones.)
7156 At the expense of a bit more processing, the header rewriting function has
7157 been changed so that it no longer eats memory in this way.
7159 66. The generation of the Received: header has been moved from the time that a
7160 message starts to be received, to the time that it finishes. The timestamp
7161 in the Received: header should now be very close to that of the <= log
7162 line. There are two side-effects of this change:
7164 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7165 logged header lines no longer include the local Received: line, because
7166 it has not yet been created. The same applies to a copy of the message
7167 that is returned to a non-SMTP sender when a message is rejected.
7169 (b) When a filter file is tested using -bf, no additional Received: header
7170 is added to the test message. After some thought, I decided that this
7173 This change does not affect the value of $received_for. It is still set
7174 after address rewriting, but before local_scan() is called.
7176 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7178 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7179 gave an unhelpful panic error message, and a defer error. I have managed to
7180 change this behaviour so that it now rejects any supplied certificate,
7181 which seems right, as the list of acceptable certificates is empty.
7183 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7184 gave an unhelpful defer error. I have not managed to make this reject any
7185 supplied certificates, but the error message it gives is "no certificate
7186 supplied", which is not helpful.
7188 70. exigrep's output now also includes lines that are not associated with any
7189 message, but which match the given pattern. Implemented by a patch from
7190 Martin Sluka, which also tidied up the Perl a bit.
7192 71. Recipient callout verification, like sender verification, was using <> in
7193 the MAIL FROM command. This isn't really the right thing, since the actual
7194 sender may affect whether the remote host accepts the recipient or not. I
7195 have changed it to use the actual sender in the callout; this means that
7196 the cache record is now keyed on a recipient/sender pair, not just the
7197 recipient address. There doesn't seem to be a real danger of callout loops,
7198 since a callout by the remote host to check the sender would use <>.
7199 [SEE ABOVE: changed after hitting problems.]
7201 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7202 temporary errors. However, in the case of such a code being given after
7203 the end of a data transmission (i.e. after ".") Exim was failing to write
7204 a retry record for the message. (Yes, there was some broken host that was
7205 actually sending 8xx at this point.)
7207 73. An unknown lookup type in a host list could cause Exim to panic-die when
7208 the list was checked. (An example that provoked this was putting <; in the
7209 middle of a list instead of at the start.) If this happened during a DATA
7210 ACL check, a -D file could be left lying around. This kind of configuration
7211 error no longer causes Exim to die; instead it causes a defer error. The
7212 incident is still logged to the main and panic logs.
7214 74. Buglet left over from Exim 3 conversion. The message "too many messages
7215 in one connection" was written to the rejectlog but not the mainlog, except
7216 when address rewriting (yes!) was being logged.
7218 75. Added write_rejectlog option.
7220 76. When a system filter was run not as root (that is, when system_filter_user
7221 was set), the values of the $n variables were not being returned to the
7222 main process; thus, they were not subsequently available in the $sn
7225 77. Added +return_path_on_delivery log selector.
7227 78. A connection timeout was being treated differently from recipients deferred
7228 when testing hosts_max_try with a message that was older than the host's
7229 retry timeout. (The host should not be counted, thus allowing all hosts to
7230 be tried at least once before bouncing.) This may have been the cause of an
7231 occasionally reported bug whereby a message would remain on the queue
7232 longer than the retry timeout, but would be bounced if a delivery was
7233 forced. I say "may" because I never totally pinned down the problem;
7234 setting up timeout/retry tests is difficult. See also the next item.
7236 79. The ultimate address timeout was not being applied to errors that involved
7237 a combination of host plus message (for example, a timeout on a MAIL
7238 command). When an address resolved to a number of possible hosts, and they
7239 were not all tried for each delivery (e.g. because of hosts_max_try), a
7240 message could remain on the queue longer than the retry timeout.
7242 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7245 81. Fixed an obscure SMTP outgoing bug which required at least the following
7246 conditions: (a) there was another message waiting for the same server;
7247 (b) the server returned 5xx to all RCPT commands in the first message so
7248 that the message was not completed; (c) the server dropped the connection
7249 or gave a negative response to the RSET that Exim sends to abort the
7250 transaction. The observed case was a dropped connection after DATA that had
7251 been sent in pipelining mode. That is, the server had advertised PIPELINING
7252 but was not implementing it correctly. The effect of the bug was incorrect
7253 behaviour, such as trying another host, and this could lead to a crash.
7259 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7260 and daemon.c were passed as pointers to ints; they should have been
7261 pointers to socklen_t variables (which are typically unsigned ints).
7263 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7266 3. Fixed a really odd bug that affected only the testing scheme; patching a
7267 certain fixed string in the binary changed the value of another string that
7268 happened to be identical to the end of the original first string.
7270 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7271 name", it returns that address as the IP address. On some operating
7272 systems (e.g. Solaris), it also passes back the IP address string as the
7273 "host name". However, on others (e.g. Linux), it passes back an empty
7274 string. Exim wasn't checking for this, and was changing the host name to an
7275 empty string, assuming it had been canonicalized.
7277 5. Although rare, it is permitted to have more than one PTR record for a given
7278 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7279 all the names associated with an address, because they do in Solaris.
7280 However, it seems that they do not in Linux for data that comes from the
7281 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7282 I found this out when I moved to a new Linux workstation and tried to run
7283 the Exim test suite.
7285 To get round this problem I have changed the code so that it now does its
7286 own call to the DNS to look up PTR records when searching for a host name.
7287 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7288 addresses that are only in /etc/hosts are still found.
7290 This behaviour is, however, controlled by an option called host_lookup_
7291 order, which defaults to "bydns:byaddr". If people want to use the other
7292 order, or indeed, just use one or the other means of lookup, they can
7293 specify it in this variable.
7295 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7296 some operating systems, this comes back from gethostbyaddr() as an empty
7297 string, and this is what Exim used to test for. However, it seems that in
7298 other systems, "." is yielded. Exim now tests for this case too.
7300 7. The values of check_spool_space and check_log_space are now held internally
7301 as a number of kilobytes instead of an absolute number of bytes. If a
7302 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7303 kilobyte. This means that much larger values can be stored.
7305 8. Exim monitor: an attempt to get the action menu when not actually pointing
7306 at a message produces an empty menu entitled "No message selected". This
7307 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7308 no entries in it ("Shell widget menu has zero width and/or height"). So I
7309 have added a single, blank menu entry in this case.
7311 9. Added ${quote_local_part.
7313 10. MIME decoding is now applied to the contents of Subject: header lines when
7316 11. Now that a reference to $sender_host_address automatically causes a reverse
7317 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7318 host lookup before query-style lookups in lists that might use this
7319 variable. This has therefore been abolished, and the "net-" prefix is no
7320 longer necessary for query-style lookups.
7322 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7323 have been a typo for LFLAGS, so it has been changed.
7325 13. The install script calls Exim with "-C /dev/null" in order to find the
7326 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7327 to be output. However, since Exim outputs its version number before the
7328 error, it didn't break the script. It just looked ugly. I fixed this by
7329 always allowing "-C /dev/null" if the caller is root.
7331 14. Ignore overlarge ACL variable number when reading spool file - insurance
7332 against a later release with more variables having written the file.
7334 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7335 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7336 this, as well as the form without the "IPv6" on the front.
7338 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7339 OS/Makefile-Darwin file.
7341 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7342 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7344 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7345 variable. However, some OS use ints for inodes. I've added cast to long int
7346 to get rid of the compiler warning.
7348 19. I had forgotten to lock out "/../" in configuration file names when
7349 ALT_CONFIG_PREFIX was set.
7351 20. Routers used for verification do not need to specify transports. However,
7352 if such a router generated a host list, and callout was configured, Exim
7353 crashed, because it could not find a port number from the (non-existent)
7354 transport. It now assumes port 25 in this circumstance.
7356 21. Added the -t option to exigrep.
7358 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7359 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7360 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7361 them is set, LOOKUP_LSEARCH is forced.
7363 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7365 24. Added sender and host information to the "rejected by local_scan()" log
7366 line; previously there was no indication of these.
7368 25. Added .include_if_exists.
7370 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7371 Linux. It turns out that not all file systems support this. Apparently some
7372 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7373 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7374 not supported on the file descriptor, is now ignored when Exim is trying to
7375 sync a directory. This applies only to Linux.
7377 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7379 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7380 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7383 29. Added support for maildirsize files from supplied patch (modified a bit).
7385 30. The use of :fail: followed by an empty string could lead Exim to respond to
7386 sender verification failures with (e.g.):
7388 550 Verification failed for <xxx>
7389 550 Sender verify failed
7391 where the first response line was missing the '-' that indicates it is not
7392 the final line of the response.
7394 31. The loop for finding the name of the user that called Exim had a hardwired
7395 limit of 10; it now uses the value of finduser_retries, which is used for
7396 all other user lookups.
7398 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7401 33. Exim was neglecting to zero errno before one call of strtol() when
7402 expanding a string and expecting an integer value. On some systems this
7403 resulted in spurious "integer overflow" errors. Also, it was casting the
7404 result into an int without checking.
7406 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7407 did not work. The code looks as if it has *never* worked, though it appears
7408 to have been documented since at least release 1.62. I have made it work.
7410 35. The "timeout_DNS" error in retry rules, also documented since at least
7411 1.62, also never worked. As it isn't clear exactly what this means, and
7412 clearly it isn't a major issue, I have abolished the feature by treating it
7413 as "timeout", and writing a warning to the main and panic logs.
7415 36. The display of retry rules for -brt wasn't always showing the error code
7418 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7419 timeout_connect_A, timeout_connect_MX.
7421 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7422 to the empty sender.
7424 39. The daemon was not analysing the content of -oX till after it had closed
7425 stderr and disconnected from the controlling terminal. This meant that any
7426 syntax errors were only noted on the panic log, and the return code from
7427 the command was 0. By re-arranging the code a little, I've made the
7428 decoding happen first, so such errors now appear on stderr, and the return
7429 code is 1. However, the actual setting up of the sockets still happens in
7430 the disconnected process, so errors there are still only recorded on the
7433 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7434 connections (as happens on some IP stacks) was logged at start up time as
7435 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7436 it from "IPv6 and IPv4", which means that two separate sockets are being
7439 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7440 says whether AF_INET or AF_INET6 was passed as an argument.
7442 42. Exiwhat output was messed up when time zones were included in log
7445 43. Exiwhat now gives more information about the daemon's listening ports,
7446 and whether -tls-on-connect was used.
7448 44. The "port" option of the smtp transport is now expanded.
7450 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7451 silently ignored. Now an error message is written to the main and panic
7454 46. There's a new ACL modifier called "logwrite" which writes to a log file
7455 as soon as it is encountered.
7457 47. Added $local_user_uid and $local_user_gid at routing time.
7459 48. Exim crashed when trying to verify a sender address that was being
7462 49. Exim was recognizing only a space character after ".include". It now also
7463 recognizes a tab character.
7465 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7466 extracting the relevant information from the specification. The man page no
7467 longer contains scrambled data for the -d option, and I've added a section
7468 at the front about calling Exim under different names.
7470 51. Added "extra_headers" argument to the "mail" command in filter files.
7472 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7475 53. Installed eximstats 1.29.
7477 54. Added transport_filter_timeout as a generic transport option.
7479 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7480 Cc: header lines. This was required by RFC 822, but it not required by RFC
7483 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7484 incoming messages that did not have them. Now it does so only if the
7485 message originates locally, that is, if there is no associated remote host
7486 address. When Resent- header lines are present, this applies to the Resent-
7487 lines rather than the non-Resent- lines.
7489 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7490 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7492 58. Messages for configuration errors now include the name of the main
7493 configuration file - useful now that there may be more than one file in a
7494 list (.included file names were always shown).
7496 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7497 for those rare installations that do not start the daemon as root or run it
7498 setuid root. I've cut out the call to initgroups() if the daemon is not
7501 60. The Exim user and group can now be bound into the binary as text strings
7502 that are looked up at the start of Exim's processing.
7504 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7506 62. Added $mailstore_basename variable.
7508 63. Installed patch to sieve.c from Michael Haardt.
7510 64. When Exim failed to open the panic log after failing to open the main log,
7511 the original message it was trying to log was written to stderr and debug
7512 output, but if they were not available (the usual case in production), it
7513 was lost. Now it is written to syslog before the two lines that record the
7514 failures to open the logs.
7516 65. Users' Exim filters run in subprocesses under the user's uid. It is
7517 possible for a "deliver" command or an alias in a "personal" command to
7518 provoke an address rewrite. If logging of address rewriting is configured,
7519 this fails because the process is not running as root or exim. There may be
7520 a better way of dealing with this, but for the moment (because 4.30 needs
7521 to be released), I have disabled address rewrite logging when running a
7522 filter in a non-root, non-exim process.
7528 1. The buildconfig auxiliary program wasn't quoting the value set for
7529 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7530 not defined. This bug was masked in 4.22 by the effect that was fixed in
7533 2. Some messages that were rejected after a message id was allocated were
7534 shown as "incomplete" by exigrep. It no longer does this for messages that
7535 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7537 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7538 have allow_domain_literals set, the ID did not get logged in the <= line.
7539 Domain literals are now always recognized in Message-ID: header lines.
7541 4. The first argument for a ${extract expansion item is the key name or field
7542 number. Leading and trailing spaces in this item were not being ignored,
7543 causing some misleading effects.
7545 5. When deliver_drop_privilege was set, single queue runner processes started
7546 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7547 same command in the process it spins off) were not dropping privilege.
7549 6. When the daemon running as "exim" started a queue runner, it always
7550 re-executed Exim in the spun-off process. This is a waste of effort when
7551 deliver_drop_privilege is set. The new process now just calls the
7552 queue-runner function directly.
7558 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7561 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7562 ignored. Though the use of -oP was forcing the writing of a pid file, it
7563 was always written to the default place.
7565 3. If the message "no IP address found for host xxxx" is generated during
7566 incoming verification, it is now followed by identification of the incoming
7567 connection (so you can more easily find what provoked it).
7569 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7571 5. Added some features to "harden" Exim a bit more against certain attacks:
7573 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7574 be put in Local/Makefile. This is like the never_users runtime option,
7575 but it cannot be overridden. The default setting is "root".
7577 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7578 prefix string with which any file named in a -C command line option
7581 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7582 is retained for -C and -D only if the caller of Exim is root. Without
7583 it, the exim user may also use -C and -D and retain privilege.
7585 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7586 command line option is disabled.
7588 6. Macro names set by the -D option must start with an upper case letter, just
7589 like macro names defined in the configuration file.
7591 7. Added "dereference=" facility to LDAP.
7593 8. Two instances of the typo "uknown" in the source files are fixed.
7595 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7596 the Configure-Makefile script screwed up while processing it.
7598 10. Incorporated PCRE 4.4.
7600 11. The SMTP synchronization check was not operating right at the start of an
7601 SMTP session. For example, it could not catch a HELO sent before the client
7602 waited for the greeting. There is now a check for outstanding input at the
7603 point when the greeting is written. Because of the duplex, asynchronous
7604 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7605 way, but not yet received, when the check is performed.
7607 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7608 on TCP/IP sockets, because this apparently causes some broken clients to
7611 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7612 unchanged) from the Cygwin maintainer.
7614 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7615 was supported instead of testing for mbx. Effectively a typo.
7617 15. The spa authenticator server code was not checking that the input it
7618 received was valid base64.
7620 16. The debug output line for the "set" modifier in ACLs was not showing the
7621 name of the variable that was being set.
7623 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7625 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7626 lookup on its own. Something else was needed to trigger the lookup. For
7627 example, a match in host_lookup or the need for a host name in a host list.
7628 Now, if $sender_host_name is referenced and the host name has not yet been
7629 looked up, a lookup is performed. If the lookup fails, the variable remains
7630 empty, and $host_lookup_failed is set to "1".
7632 19. Added "eqi" as a case-independent comparison operator.
7634 20. The saslauthd authentication condition could segfault if neither service
7635 nor realm was specified.
7637 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7638 error message that was logged was misleading, and incoming SMTP
7639 connections were dropped. The message is now more accurate, and temporary
7640 errors are given to SMTP connections.
7642 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7643 (except RSET) with a 421 error, until QUIT is received. However, it was
7644 failing to send a response to QUIT.
7646 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7647 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7648 ACL is now run after the helo_try_verify_hosts code.
7650 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7651 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7652 case-independent, so other case variants are also recognized). Apparently
7653 some systems use these upper case variants.
7655 25. If more than two messages were waiting for the same host, and a transport
7656 filter was specified for the transport, Exim sent two messages over the
7657 same TCP/IP connection, and then failed with "socket operation on non-
7658 socket" when it tried to send the third.
7660 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7662 27. The extern definition of crypt16() in expand.c was not being excluded when
7663 the OS had its own crypt16() function.
7665 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7666 as a preferred synonym for return_size_limit, both as an option and as an
7669 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7671 30. Changed the default configuration ACL to relax the local part checking rule
7672 for addresses that are not in any local domains. For these addresses,
7673 slashes and pipe symbols are allowed within local parts, but the sequence
7674 /../ is explicitly forbidden.
7676 31. SPA server authentication was not clearing the challenge buffer before
7679 32. log_message in a "warn" ACL statement was writing to the reject log as
7680 well as to the main log, which contradicts the documentation and doesn't
7681 seem right (because no rejection is happening). So I have stopped it.
7683 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7684 However, I am unable to do any testing of this.
7686 34. Fixed an infelicity in the appendfile transport. When checking directories
7687 for a mailbox, to see if any needed to be created, it was accidentally
7688 using path names with one or more superfluous leading slashes; tracing
7689 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7691 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7692 contents are added to the log line that is written for every discarded
7693 recipient. (Previously a log_message setting was ignored.)
7695 36. The ${quote: operator now quotes the string if it is empty.
7697 37. The install script runs exim in order to find its version number. If for
7698 some reason other than non-existence or emptiness, which it checks, it
7699 could not run './exim', it was installing it with an empty version number,
7700 i.e. as "exim-". This error state is now caught, and the installation is
7703 38. An argument was missing from the function that creates an error message
7704 when Exim fails to connect to the socket for saslauthd authentication.
7705 This could cause Exim to crash, or give a corrupted message.
7707 39. Added isip, isip4, and isip6 to ${if conditions.
7709 40. The ACL variables $acl_xx are now saved with the message, and can be
7710 accessed later in routers, transports, and filters.
7712 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7713 strings in the file are not string-expanded.
7715 42. If a MAIL command specified a SIZE value that was too large to fit into an
7716 int variable, the check against message_size_limit failed. Such values are
7717 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7718 day this will have to be increased, but I don't think I want to be around
7719 when emails are that large.
7726 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7727 iconv() is not standard in FreeBSD.
7729 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7730 IPv6 enabled. The observed symptom was a segmentation fault on return from
7731 the function os_common_find_running_interfaces() in src/os.c.
7733 3. In the check_special_case() function in daemon.c I had used "errno" as an
7734 argument name, which causes warnings on some systems. This was basically a
7735 typo, since it was named "eno" in the comments!
7737 4. The code that waits for the clock to tick (at a resolution of some fraction
7738 of a second) so as to ensure message-id uniqueness was always waiting for
7739 at least one whole tick, when it could have waited for less. [This is
7740 almost certainly not relevant at current processor speeds, where it is
7741 unlikely to ever wait at all. But we try to future-proof.]
7743 5. The function that sleeps for a time interval that includes fractions of a
7744 second contained a race. It did not block SIGALRM between setting the
7745 timer, and suspending (a couple of lines later). If the interval was short
7746 and the sigsuspend() was delayed until after it had expired, the suspension
7747 never ended. On busy systems this could lead to processes getting stuck for
7750 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7751 process, before it forks any delivery processes. The open lookup caching
7752 mechanism meant that the open file or database connection was passed into
7753 the delivery process. The problem was that delivery processes always tidy
7754 up cached lookup data. This could cause a problem for the next delivery
7755 process started by the queue runner, because the external queue runner
7756 process does not know about the closure. So the next delivery process
7757 still has data in the lookup cache. In the case of a file lookup, there was
7758 no problem because closing a file descriptor in a subprocess doesn't affect
7759 the parent. However, if the lookup was caching a connection to a database,
7760 the connection was closed, and the second delivery process was likely to
7761 see errors such as "PGSQL: query failed: server closed the connection
7762 unexpectedly". The problem has been fixed by closing all cached lookups
7763 in a queue runner before running a delivery process.
7765 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7766 seem to have the "const" qualifier which it has on other OS. I've
7769 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7770 *of the same type* (client or server) with the same public name that an
7771 error should be diagnosed.
7773 9. When Exim looked up a host name for an IP address, but failed to find the
7774 original IP address when looking up the host name (a safety check), it
7775 output the message "<ip address> does not match any IP for NULL", which was
7776 confusing, to say the least. The bug was that the host name should have
7777 appeared instead of "NULL".
7779 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7780 user that is built into the binary, and the -C or -D options is used, root
7781 privilege is dropped before the configuration file is read. In addition,
7782 logging is switched to stderr instead of the normal log files. If the
7783 configuration then re-defines the Exim user, the unprivileged environment
7784 is probably not what is expected, so Exim logs a panic warning message (but
7787 However, if deliver_drop_privilege is set, the unprivileged state may well
7788 be exactly what is intended, so the warning has been cut out in that case,
7789 and Exim is allowed to try to write to its normal log files.
7795 1. smtp_return_error_details was not giving details for temporary sender
7796 or receiver verification errors.
7798 2. Diagnose a configuration error if two authenticators have the same public
7801 3. Exim used not to create the message log file for a message until the first
7802 delivery attempt. This could be confusing when incoming messages were held
7803 for policy or load reasons. The message log file is now created at the time
7804 the message is received, and an initial "Received" line is written to it.
7806 4. The automatically generated man page for command line options had a minor
7807 bug that caused no ill effects; however, a more serious problem was that
7808 the procedure for building the man page automatically didn't always
7809 operate. Consequently, release 4.20 contains an out-of-date version. This
7810 shouldn't happen again.
7812 5. When building Exim with embedded Perl support, the script that builds the
7813 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7814 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7816 6. The freeze_tell option was not being used for messages that were frozen on
7817 arrival, either by an ACL or by local_scan().
7819 7. Added the smtp_incomplete_transaction log selector.
7821 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7822 was accepting AUTH without a new EHLO.
7824 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7825 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7828 10. Logging of TCP/IP connections (when configured) now happens in the main
7829 daemon process instead of the child process, so that the TCP/IP connection
7830 count is more accurate (but it can never be perfect).
7832 11. The use of "drop" in a nested ACL was not being handled correctly in the
7833 outer ACL. Now, if condition failure induced by the nested "drop" causes
7834 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7835 or "require"), the connection is dropped.
7837 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7838 that yield "discard" can now be used with an "accept" or a "discard" verb,
7839 but an error is generated for any others (because I can't see a useful way
7840 to define what should happen).
7842 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7843 are now processed in the same way as lines in the Exim configuration file.
7844 In particular, continuation lines are supported.
7846 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7848 15. Added -ti meaning -t -i.
7850 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7851 domains, and warn by logging if others are found.
7853 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7854 returned from GIFCONF when Exim is trying to find the list of interfaces on
7855 a host. The code in os.c has been modified to copy each ifreq to an aligned
7856 structure in all cases.
7858 Also, in some cases, the returned ifreq's were being copied to a 'struct
7859 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7860 means the last couple of bytes of an IPv6 address could be chopped if the
7861 ifreq contained only a normal sockaddr (14 bytes storage).
7863 18. Named domain lists were not supported in the hosts_treat_as_local option.
7864 An entry such as +xxxx was not recognized, and was treated as a literal
7867 19. Ensure that header lines added by a DATA ACL are included in the reject log
7868 if the ACL subsequently rejects the message.
7870 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7871 MD5 (which is deprecated).
7873 21. When testing a filter file using -bf, Exim was writing a message when it
7874 took the sender from a "From " line in the message, but it was not doing so
7875 when it took $return_path from a Return-Path: header line. It now does.
7877 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7878 with a valid header line field name (a series of printing characters
7879 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7881 23. Changed "disc" in the source to "disk" to conform to the documentation and
7882 the book and for uniformity.
7884 24. Ignore Sendmail's -Ooption=value command line item.
7886 25. When execve() failed while trying to run a command in a pipe transport,
7887 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7888 could be confused with a return value of 69 from the command itself. This
7889 has been changed to 127, the value the shell returns if it is asked to run
7890 a non-existent command. The wording for the related log line suggests a
7891 non-existent command as the problem.
7893 26. If received_header_text expands to an empty string, do not add a Received:
7894 header line to the message. (Well, it adds a token one on the spool, but
7895 marks it "old" so that it doesn't get used or transmitted.)
7897 27. Installed eximstats 1.28 (addition of -nt option).
7899 28. There was no check for failure on the call to getsockname() in the daemon
7900 code. This can fail if there is a shortage of resources on the system, with
7901 ENOMEM, for example. A temporary error is now given on failure.
7903 29. Contrary to the C standard, it seems that in some environments, the
7904 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7905 program. Exim now does this explicitly; it affects the formatting of
7906 timestamps using strftime().
7908 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7909 complaints. I've now initialized all the variables, to avoid this.
7911 32. Header lines added by a system filter were not being "seen" during
7912 transport-time rewrites.
7914 33. The info_callback() function passed to OpenSSL is set up with type void
7915 (*)(SSL *, int, int), as described somewhere. However, when calling the
7916 function (actually a macro) that sets it up, the type void(*)() is
7917 expected. I've put in a cast to prevent warnings from picky compilers.
7919 34. If a DNS black list lookup found a CNAME record, but there were no A
7920 records associated with the domain it pointed at, Exim crashed.
7922 35. If a DNS black list lookup returned more than one A record, Exim ignored
7923 all but the first. It now scans all returned addresses if a particular IP
7924 value is being sought. In this situation, the contents of the
7925 $dnslist_value variable are a list of all the addresses, separated by a
7928 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7929 to accept a host name and all its aliases if the forward lookup for any of
7930 them yielded the IP address of the incoming connection. Now it accepts only
7931 those names whose forward lookup yields the correct IP address. Any other
7932 names are discarded. This closes a loophole whereby a rogue DNS
7933 administrator could create reverse DNS records to break through a
7934 wildcarded host restriction in an ACL.
7936 37. If a user filter or a system filter that ran in a subprocess used any of
7937 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7938 the wrong values were passed to the pipe command ($thisaddress had the
7939 value of $0, $0 had the value of $1, etc). This bug was introduced by
7940 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7942 38. Improved the line breaking for long SMTP error messages from ACLs.
7943 Previously, if there was no break point between 40 and 75 characters, Exim
7944 left the rest of the message alone. Two changes have been made: (a) I've
7945 reduced the minimum length to 35 characters; (b) if it can't find a break
7946 point between 35 and 75 characters, it looks ahead and uses the first one
7947 that it finds. This may give the occasional overlong line, but at least the
7948 remaining text gets split now.
7950 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7951 file descriptors might be low, and that setting 1000 would always raise it.
7952 It turns out that in some environments, the limit is already over 1000 and
7953 that lowering it causes trouble. So now Exim takes care not to decrease it.
7955 40. When delivering a message, the value of $return_path is set to $sender_
7956 address at the start of routing (routers may change the value). By an
7957 oversight, this default was not being set up when an address was tested by
7958 -bt or -bv, which affected the outcome if any router or filter referred to
7961 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7962 log only when "message" or "log_message" are set. However, if one of the
7963 conditions was an address verification, or a call to a nested ACL, the
7964 messages generated by the underlying test were being passed through. This
7965 no longer happens. The underlying message is available in $acl_verify_
7966 message for both "message" and "log_message" expansions, so it can be
7967 passed through if needed.
7969 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7970 new expansion $bh_ to give the encoded byte string without charset
7971 translation. Translation happens only if iconv() is available; HAVE_ICONV
7972 indicates this at build time. HEADERS_CHARSET gives the charset to
7973 translate to; headers_charset can change it in the configuration, and
7974 "headers charset" can change it in an individual filter file.
7976 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7977 that creates RFC 2047 encoded "words" labels them as that charset instead
7978 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7979 expansion operator; (ii) when Exim creates a From: line for a local
7980 message; (iii) when a header line is rewritten to include a "phrase" part.
7982 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7983 buggy, causing it to skip the first lines of messages whose message ID
7984 ended in 'D'. This would not have bitten before Exim release 4.14, because
7985 message IDs were unlikely to end in 'D' before then. The effect was to have
7986 incorrect size information for certain domains.
7988 45. #include "config.h" was missing at the start of the crypt16.c module. This
7989 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7992 46. If there was a timeout during a "random" callout check, Exim treated it as
7993 a failure of the random address, and carried on sending RSET and the real
7994 address. If the delay was just some slowness somewhere, the response to the
7995 original RCPT would be taken as a response to RSET and so on, causing
7996 mayhem of various kinds.
7998 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7999 when I implemented it. It didn't allow for the fact that some option values
8000 may legitimately be negative (e.g. size_addition), and it didn't even do
8001 the right test for positive values.
8003 48. Domain names in DNS records are case-independent. Exim always looks them up
8004 in lower case. Some resolvers return domain names in exactly the case they
8005 appear in the zone file, that is, they may contain uppercase letters. Not
8006 all resolvers do this - some return always lower case. Exim was treating a
8007 change of case by a resolver as a change of domain, similar to a widening
8008 of a domain abbreviation. This triggered its re-routing code and so it was
8009 trying to route what was effectively the same domain again. This normally
8010 caused routing to fail (because the router wouldn't handle the domain
8011 twice). Now Exim checks for this case specially, and just changes the
8012 casing of the domain that it ultimately uses when it transmits the message
8015 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8018 50. If a filter generated a file delivery with a non-absolute name (possible if
8019 no home directory exists for the router), the forbid_file option was not
8022 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8023 the existing equality matching.
8025 52. Exim was using ints instead of ino_t variables in some places where it was
8026 dealing with inode numbers.
8028 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8029 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8030 called TMPDIR, and if it finds it is different, it changes its value.
8032 54. The smtp_printf() function is now made available to local_scan() so
8033 additional output lines can be written before returning. There is also an
8034 smtp_fflush() function to enable the detection of a dropped connection.
8035 The variables smtp_input and smtp_batched_input are exported to
8038 55. Changed the default runtime configuration: the message "Unknown user"
8039 has been removed from the ACL, and instead placed on the localuser router,
8040 using the cannot_route_message feature. This means that any verification
8041 failures that generate their own messages won't get overridden. Similarly,
8042 the "Unrouteable address" message that was in the ACL for unverifiable
8043 relay addresses has also been removed.
8045 56. Added hosts_avoid_esmtp to the smtp transport.
8047 57. The exicyclog script was not checking for the esoteric option
8048 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8049 will work only if exicyclog is run under the appropriate euid.
8051 58. Following a discussion on the list, the rules by which Exim recognises line
8052 endings on incoming messages have been changed. The -dropcr and drop_cr
8053 options are now no-ops, retained only for backwards compatibility. The
8054 following line terminators are recognized: LF CRLF CR. However, special
8055 processing applies to CR:
8057 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8058 nor a local message in the state where . is a terminator.
8060 (ii) If a bare CR is encountered in a header line, an extra space is added
8061 after the line terminator so as not to end the header. The reasoning
8062 behind this is that bare CRs in header lines are most likely either
8063 to be mistakes, or people trying to play silly games.
8065 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8066 was being incorrectly given as 18 bytes larger than it should have been.
8067 This is a VOB (very old bug).
8069 60. This may never have affected anything current, but just in case it has:
8070 When the local host is found other than at the start of a list of hosts,
8071 the local host, those with the same MX, and any that follow, are discarded.
8072 When the list in question was part of a longer list of hosts, the following
8073 hosts (not currently being processed) were also being discarded. This no
8074 longer happens. I'm not sure if this situation could ever has previously
8077 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8080 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8081 header. This is something that is recommended in a new Internet Draft, and
8082 is something that is documented as being done by Sendmail. There are two
8083 possible values. For messages generated by the autoreply transport, Exim
8086 Auto-Submitted: auto-replied
8088 whereas for all other generated messages (e.g. bounces) it adds
8090 Auto-Submitted: auto-generated
8092 63. The "personal" condition in filters now includes a test for the
8093 Auto-Submitted: header. If it contains the string "auto-" the message it
8094 not considered personal.
8096 64. Added rcpt_include_affixes as a generic transport option.
8098 65. Added queue_only_override (default true).
8100 66. Added the syslog_duplication option.
8102 67. If what should have been the first header line of a message consisted of
8103 a space followed by a colon, Exim was mis-interpreting it as a header line.
8104 It isn't of course - it is syntactically invalid and should therefore be
8105 treated as the start of the message body. The misbehaviour could have
8106 caused a number of strange effects, including loss of data in subsequent
8107 header lines, and spool format errors.
8109 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8110 client host had authenticated. This control can now be exercised by an ACL
8111 for more flexibility.
8113 69. By default, callouts do not happen when testing with -bh. There is now a
8114 variant, -bhc, which does actually run the callout code, including
8115 consulting and updating the callout cache.
8117 70. Added support for saslauthd authentication, courtesy of Alexander
8120 71. If statvfs() failed on the spool or log directories while checking their
8121 size for availability, Exim confusingly gave the error "space shortage".
8122 Furthermore, in debugging mode it crashed with a floating point exception.
8123 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8124 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8125 serious problem, Exim now writes to the main and panic logs when this
8126 happens, with details of the failure. It then refuses to accept the
8127 incoming message, giving the message "spool directory problem" or "log
8128 directory problem" with a 421 code for SMTP messages.
8130 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8131 0, 1, and 2 exist, because some OS complain for execs without them (see
8132 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8133 descriptors. However, the code omitted to check that the open succeeded,
8134 causing mysterious errors if for some reason the permissions on /dev/null
8135 got screwed. Now Exim writes a message to the main and panic logs, and
8136 bombs out if it can't open /dev/null.
8138 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8139 interact so that it is all more flexible. It is supposed to remain
8140 backwards compatible. Also added extra_local_interfaces.
8142 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8143 to bomb out with an assertion failure - to the client this appears as a
8144 connection drop. This problem occurs in the part of the code that was taken
8145 from the Samba project. Fortunately, the assertion is in a very simple
8146 function, so I have fixed this by reproducing the function inline in the
8147 one place where it is called, and arranging for authentication to fail
8148 instead of killing the process with assert().
8150 75. The SPA client code was not working when the server requested OEM rather
8151 than Unicode encoding.
8153 76. Added code to make require_files with a specific uid setting more usable in
8154 the case where statting the file as root fails - usually a non-root-mounted
8155 NFS file system. When this happens and the failure is EACCES, Exim now
8156 forks a subprocess and does the per-uid checking as the relevant uid.
8158 77. Added process_log_path.
8160 78. If log_file_path was not explicitly set, a setting of check_log_space or
8161 check_log_inodes was ignored.
8163 79. If a space check for the spool or log partitions fails, the incident is now
8164 logged. Of course, in the latter case the data may get lost...
8166 80. Added the %p formatting code to string_format() so that it can be used to
8167 print addresses in debug_print(). Adjusted all the address printing in the
8168 debugging in store.c to use %p rather than %d.
8170 81. There was a concern that a line of code in smtp_in.c could overflow a
8171 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8172 initially expressed, the concern was not well-founded, because trailing
8173 spaces are removed early. However, if the trailing spaces were followed by
8174 a NULL, they did not get removed, so the overflow was possible. Two fixes
8177 (a) I re-wrote the offending code in a cleaner fashion.
8178 (b) If an incoming SMTP command contains a NULL character, it is rejected
8181 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8182 runs initgroups(), so that if the Exim user is in any additional groups,
8183 they will be used during message reception.
8189 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob