1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
88 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
89 related applications. Previously an "H" was used where available info
90 says that "M" should be, so change to match.
92 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
93 as arguments, so an implementation trying to copy these into a local
94 buffer was taking a taint-enforcement trap. Fix by using dynamically
95 created buffers. Similar fix for radius expansion condition.
97 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
98 reasonable, eg. to count headers. Fix by using dynamically created
99 buffers rather than a local. Do similar fixes for ACL actions "dcc",
100 "log_reject_target", "malware" and "spam"; the arguments are expanded
101 so could be handling tainted values.
103 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
104 broken the (no-op) support for this sendmail command. Restore it
105 to doing nothing, silently, and returning good status.
107 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
108 record path was given (or the default used) without a leading directory
109 path, an error occurred on trying to open it. Use the transport's working
112 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
113 transport. Previously it was the name on the DNS A-record; use instead
114 the head of the CNAME chain leading there (if there is one). This seems
115 to align better with RFC 6125.
117 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
118 smtp_accept_max_per_host allocated resources which were not released
119 when the limit was exceeded. This eventually crashed the daemon. Fix
120 by adding a release action in that path.
122 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
123 expanded; previously using tainted values was rejected. Fix by using
124 dynamically-created buffers.
126 JH/09 Relax restrictions on ACL verify condition needing access to message
127 headers. Previously they were only permitted in data and non-smtp ACLs;
128 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
129 not_blind, header_sender and header_names_ascii verification.
131 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
132 Previously a macro used one argument twice; when called with the
133 argument as an expression having side-effects, incorrect operation
134 resulted. Use an inlineable function.
136 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
137 held open for a verify callout. Previously this wan not accounted for
138 and a corrupt onward SMTP conversation resulted.
140 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
141 Previously they were embedded naively in the constructed address; when
142 needed, strip the quoting and quote the entire local_part.
143 Also make the inbound_srs expansion condition handle quoting.
145 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
146 excluded, not matching the documentation.
148 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
149 was given for the sqlite_dbfile a trap resulted.
151 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
152 "name" argument resulted in a trap. There is no reason to disallow such;
153 this was a coding error.
155 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
156 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
157 spent suspended, ignoring the POSIX definition. Previously we assumed
158 it did and a constant offset from real time could be used as a correction.
159 Change to using the same clock source for the start-of-message and the
160 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
161 exists, just to get a clock slightly more aligned to reality.
163 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
164 RFC says it is optional some validators care. The missing char was not
165 intended but triggered by a line-wrap alignment. Discovery and fix by
166 Guillaume Outters, hacked on by JH.
168 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
169 name being quoted was tainted a trap would be taken. Fix by using
170 dynamically created buffers. The routine could have been called by a
171 rewrite with the "h" flag, by using the "-F" command-line option, or
172 by using a "name=" option on a control=submission ACL modifier.
174 JH/19 SPF: change the Authentication-Results expansion component to give
175 smtp.helo when the sender domain is empty. Previously it gave
178 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
179 not in 127.0.0.0/8 to help in spotting list domains taken over by a
180 domain-parking registrar.
182 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
183 Previously when a whitespace character was specified it was not inserted
184 after removing the newline.
186 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
187 the domain part of the recipient address. This overrides any tls_sni
188 option set, which was previously used.
190 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
193 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
194 is_tainted() had an off-by-one error in the overenthusiastic direction.
195 Find and fix by Gavan. Although NetBSD is not a supported platform for
196 4.94 this bug could affect other platforms.
198 PP/01 Fix default prime selection to be consistent.
199 One path used ike23 still, instead of exim.dev.20160529.3; now both
200 execution flows will use the same DH primes (currently
201 exim.dev.20160529.3).
203 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
204 option in smtp transport, to match the documentation. Previously
205 verification was not being done in some cases where it should have been.
207 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
208 than one server was defined and depending on the platform memory layout
209 details, an internal consistency trap could be hit while walking the list
212 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
213 files. The value is used by the authresults expansion item. Previously
214 if this was used in a router or transport, a crash could result.
216 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
217 have a sufficient incidence of EINTR returns from select that an
218 interaction with other operations done by the main daemon loop exposed
219 a bug in the error-handling. This was benign apart from the log
222 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
223 for consistency with delivered "=>" and failed "**" lines. While we're
224 there, handle PRX and TFO.
226 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
227 applied. This resulted, if any header-line rewrite rules were configured,
228 in a panic-log triggerable by sending a message with a long address in
229 a header. Fix by increasing the arbitrary limit to larger than a single
230 (dewrapped) 5322 header line maximum size.
232 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
233 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
234 RFC 6648 which deprecates X- options in protocols as a general practice.
235 Changeover between the implementations is handled by the mechanisms
238 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
239 a remote which uses callout/hold. Previously the local was queued.
241 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
244 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
245 log line. The asterisk is now consistently appended to the remote IP
246 (and port, if given), and will also be provided on defer and fail log
247 lines. Previously it could be placed on the local IP if that was being
248 logged, and was only provided on delivery lines.
250 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
252 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
253 authenticator client_send option. Previously the next char, after a pair
254 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
255 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
256 documentation. There is still no way to get a leading ^ immediately
257 after a NUL (ie. for the password of a PLAIN method authenticator.
259 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
260 files. For bad sizes read, delete the record and whine to paniclog.
262 JH/38 When logging an AUTH failure, as server, do not include sensitive
263 information. Previously, the credentials would be included if given
264 as part of the AUTH command line and an ACL denied authentication.
266 JH/39 Bug 2691: fix $local_part_data. When the matching list element
267 referred to a file, bad data was returned. This likely also affected
270 JH/40 The gsasl authenticator now supports caching of the salted password
271 generated by the client-side implementation. This required the addition
272 of a new variable: $auth4.
274 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
275 left undeleted; the attempt to re-create it then failed - resulting in
276 the usual "SIGHUP tp have daemon reload configuration" to not work.
277 This affected any platform not supporting "abstract" Unix-domain
278 sockets (i.e. not Linux).
280 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
281 recipients" response to RCPT in a later response, with a 250. The
282 previous coding assumed this would not happen, and under PIPELINING
283 would result in both lost and duplicate recipients for a message.
285 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
286 Previously the weighting was incorrectly applied. Similar fix for socks
287 proxies. Found and fixed by Heiko Schlichting.
289 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
290 not handle sub-lists included using the +namedlist syntax. While
291 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
292 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
295 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
296 Previously the permanent pool was used, so the sore could not be freed.
297 This meant a connection with many messages would use continually-growing
300 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
301 per-pool so as not to waste too much space. Previously a constant size
302 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
303 DOS attacks harder. The cost is wasted memory use in the larger blocks.
305 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
306 same space repeatedly, and a smaller process footprint.
308 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
309 for GnuTLS. Previously the name started "%s" which, while not a bug,
310 looked as if if might be one.
312 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
313 "continued connection") make the $proxy_* variables available. Previously
314 the information was not passed across the exec() call for subsequent
315 transport executions. This also mean that the log lines for the
316 messages can show the proxy information.
318 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
319 receive time. With modern systems the difference is significant.
320 The historical behaviour can be restored by disabling (a new) log_selector
321 "queue_time_exclusive".
323 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
324 content) were specifically tested for. Now, also cover expansions
325 resulting in ACL names and inline ACL content.
327 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
328 making it unusable in complex expressions.
330 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
331 fail with a taint-check on the spoolfile name, and leave the message
334 HS/01 Enforce absolute PID file path name.
336 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
338 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
340 PP/02 Bug 2643: Correct TLS DH constants.
341 A missing NUL termination in our code-generation tool had led to some
342 incorrect Diffie-Hellman constants in the Exim source.
343 Reported by kylon94, code-gen tool fix by Simon Arlott.
345 PP/03 Impose security length checks on various command-line options.
346 Fixes CVE-2020-SPRSS reported by Qualys.
348 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
349 better. Reported by Qualys.
351 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
352 providing a particularly obnoxious sender full name.
355 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
357 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
360 PP/08 Change default for recipients_max from unlimited to 50,000.
362 PP/09 Fix security issue with too many recipients on a message (to remove a
363 known security problem if someone does set recipients_max to unlimited,
364 or if local additions add to the recipient list).
365 Fixes CVE-2020-RCPTL reported by Qualys.
367 PP/10 Fix security issue in SMTP verb option parsing
368 Fixes CVE-2020-EXOPT reported by Qualys.
370 PP/11 Fix security issue in BDAT state confusion.
371 Ensure we reset known-good where we know we need to not be reading BDAT
372 data, as a general case fix, and move the places where we switch to BDAT
373 mode until after various protocol state checks.
374 Fixes CVE-2020-BDATA reported by Qualys.
376 HS/03 Die on "/../" in msglog file names
378 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
379 the Exim runtime user are allowed to create files.
381 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
384 QS/03 When reading the output from interpreted forward files we do not
385 pass the pipe between the parent and the interpreting process to
386 executed child processes (if any).
388 QS/04 Always die if requested from internal logging, even is logging is
391 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
392 the API; compilation noo longer completes with DMARC support included.
393 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
394 been reported on other platforms.
396 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
398 JH/56 Make the majority of info read from config files readonly, for defence-in-
399 depth against exploits. Suggestion by Qualys.
400 Not supported on Solaris 10.
402 JH/57 Fix control=fakreject for a custom message containing tainted data.
403 Previously this resulted in a log complaint, due to a re-expansion present
404 since fakereject was originally introduced.
406 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
407 modifier was given, a loop resulted.
409 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
410 pipelined SMTP command followed the BDAT LAST then it would be
411 incorrectly treated as part of the message body, causing a verification
414 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
415 require looser validation rules than those for 821-level addresses,
416 which only permit IP addresses.
422 JH/01 Avoid costly startup code when not strictly needed. This reduces time
423 for some exim process initialisations. It does mean that the logging
424 of TLS configuration problems is only done for the daemon startup.
426 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
428 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
429 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
432 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
433 DKIM signing is being done. Previously a transport_filter would always
434 disable CHUNKING, falling back to traditional DATA.
436 JH/05 Regard command-line recipients as tainted.
438 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
440 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
441 PAM library frees one of the arguments given to it, despite the
442 documentation. Therefore a plain malloc must be used.
444 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
445 on-stack buffers were used, resulting in a taint trap when DSN information
446 copied from a received message was written into the buffer.
448 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
449 the ordering of its ARC headers. This caused a crash.
451 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
452 a new record was being constructed with information from the peer, a trap
455 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
456 installation would get error messages from DMARC verify, when it hit the
457 nonexistent file indicated by the default. Distros wanting DMARC enabled
458 should both provide the file and set the option.
459 Also enforce no DMARC verification for command-line sourced messages.
461 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
462 could, depending on the platform, hang at the STARTTLS response.
464 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
465 message on a connection. Previously if one message had ARC headers and
466 the following one did not, a crash could result when adding an
467 Authentication-Results: header.
469 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
470 Exim main code and Exim-related utities. The introduction of taint
471 tracking also did many adjustments to string handling. Since then, eximon
472 frequently terminated with an assert failure.
474 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
475 check for 452 responses. This slightly helps the inefficieny of doing
476 a large alias-expansion into a recipient-limited target. The max_rcpt
477 transport option still applies (and at the current default, will override
478 the new feature). The check is done for either cause of synch, and forces
479 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
480 connection. The new facility is not tunable at this time.
482 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
483 library live data was being used, so the results became garbage. Make
484 copies while it is still usable.
486 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
487 on delivery deferred (==) and failed (**) lines (if a delivery was
488 attemtped). Previously it was only on completion (=>) lines.
490 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
491 for the expansion of the server_scram_iter and server_scram_salt options.
493 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
494 are now specifically given a NO_DATA response without hitting the system
495 resolver. The library goes on to do the now-standard TXT lookup.
496 Use of dnsdb lookups is not affected.
498 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
499 only retrieve the errormessage once. Previously two calls to dlerror()
500 were used, and the second one (for mainlog/paniclog) retrieved null
503 JH/20 Taint checking: disallow use of tainted data for
504 - the appendfile transport file and directory options
505 - the pipe transport command
506 - the autoreply transport file, log and once options
507 - file names used by the redirect router (including filter files)
509 - paths used by single-key lookups
510 Previously this was permitted.
512 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
513 adjusted the size of a major service buffer; this failed because the
514 buffer was in use at the time. Change to a compile-time increase in the
515 buffer size, when this authenticator is compiled into exim.
517 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
518 previous fast-mode was untenable in the face of glibs using mmap to
519 support larger malloc requests.
521 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
522 New values supported, if defined on system where compiled:
523 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
524 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
526 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
527 running a limited number of proceses in parallel, a benefit is gained. The
528 amount varies with the platform hardware and load. The use of the option
529 queue_run_in_order means we cannot do this, as ordering becomes
532 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
533 had introduced a string-copy (for ensuring NUL-termination) which was not
534 appropriate for that case, which can include embedded NUL bytes in the
535 block of data. Investigation showed the copy to actually be needless, the
536 data being length-specified.
538 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
539 done during a receiving connection, and both used TLS, global info was
540 used rather than per-connection info for tracking the state of data
541 queued for transmission. This could result in a connection hang.
543 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
544 Previously, when delivering serveral messages down a single connection
545 only the first would provide a SIZE. This was due to the size information
546 not being properly tracked.
548 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
549 TAI (at 37 seconds currently), pretend to be in UTC for time-related
550 expansion and logging. Previously, spurious values such as a future
551 minute could be seen.
553 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
554 it could crash from a null-deref. This could also affect the
555 ${addresses: } operator and ${readsock } item.
557 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
558 message following a mime one, the variable was not reset.
560 JH/30 When an pipelined-connect fails at the first response, assume incorrect
561 cached capability (perhaps the peer reneged?) and immediately retry in
564 JH/31 Fix spurious detection of timeout while writing to transport filter.
566 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
567 an attempt to copy the string was made before checking it.
569 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
570 taint of the lookup key was maintained; we now regard the presence in the
571 filesystem as sufficient validation.
573 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
574 argument is supplied.
576 JH/35 The dsearch lookup now requires that the directory is an absolute path.
577 Previously this was not checked, and nonempty relative paths made an
578 access under Exim's current working directory.
580 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
581 Previously no event was raised.
583 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
584 parameter supplied by the sender MAIL FROM command. Previously it was
585 ignored, and only the check_spool_space option value for the required
588 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
589 the size of the signing public-key. Previously it was instead giving
590 the size of the signature hash.
592 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
593 the default. See the (new) dkim_verify_min_keysizes option.
595 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
596 and an ACL use a lookup for checking either the local_part or domain,
597 stale data could be accessed. Ensure that variable references are
598 dropped between messages.
600 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
601 by the client was not checked as pointing within response data before
602 being used. A malicious client could thus cause an out-of-bounds read and
603 possibly gain authentication. Fix by adding the check.
605 JH/42 Internationalisation: change the default for downconversion in the smtp
606 transport to be "if needed". Previously it was "as previously set" for
607 the message, which usually meant "if needed" for message-submission but
608 "no" for everything else. However, MTAs have been seen using SMTPUTF8
609 even when the envelope addresses did not need it, resulting in forwarding
610 failures to non-supporting MTAs. A downconvert in such cases will be
611 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
612 transport. The change does mean that addresses needing conversion will
613 be converted when previously a delivery failure would occur.
615 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
616 response was received it would be used unchecked in a fail-DSN, violating
617 standards on line-length limits. Truncate if needed.
619 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
620 doesn't work. (Additionally add a new main config option to configure the
627 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
628 side, to decode a TLS 1.3 packet capture.
630 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
631 Previously the default library behaviour applied, sending two, each in
634 JH/03 Debug output for ACL now gives the config file name and line number for
637 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
639 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
641 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
642 buffer overrun for (non-chunking) other transports.
644 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
645 TLS1.3, means that a server rejecting a client certificate is not visible
646 to the client until the first read of encrypted data (typically the
647 response to EHLO). Add detection for that case and treat it as a failed
648 TLS connection attempt, so that the normal retry-in-clear can work (if
649 suitably configured).
651 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
652 and/or domain. Found and fixed by Jason Betts.
654 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
655 configuration). If a CNAME target was not a wellformed name pattern, a
658 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
659 the OS reports them interleaved with other addresses.
661 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
662 used both for input and for a verify callout, both encrypted, SMTP
663 responses being sent by the server could be lost. This resulted in
664 dropped connections and sometimes bounces generated by a peer sending
667 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
668 string. Previously it was possible to cause undefined behaviour in a
669 library routine (usually a crash). Found by "zerons".
671 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
674 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
675 API was removed, so update to use the newer ones.
677 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
678 any timeout set, is taking a long time. Previously we would hang on to a
679 rotated logfile "forever" if the input was arriving with long gaps
680 (a previous attempt to fix addressed lack, for a long time, of initial
683 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
684 shared (NFS) environment. The length of the tempfile name is now
685 4 + 16 ("hdr.$message_exim_id") which might break on file
686 systems which restrict the file name length to lower values.
689 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
690 shared (NFS) environment.
692 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
693 did for all versions <4.90). Notably -M, -m, --invert, -I may be
696 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
697 on some platforms for bit 31.
699 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
700 to changes apparently associated with TLS1.3 handling some of the APIs
701 previously used were either nonfunctional or inappropriate. Strings
702 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
703 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
704 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
705 This affects log line X= elements, the $tls_{in,out}_cipher variables,
706 and the use of specific cipher names in the encrypted= ACL condition.
708 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
710 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
711 verification result was not updated unless hosts_require_ocsp applied.
713 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
714 queue_list_requires_admin set to false, non-admin users were denied the
717 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
718 directory-of-certs mode. Previously they were advertised despite the
721 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
722 A single TCP connection by a client will now hold a TLS connection open
723 for multiple message deliveries, by default. Previously the default was to
726 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
727 default. If built with the facility, DANE will be used. The facility
728 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
730 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
731 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
732 must be defined and you must still, unless you define DISABLE_TLS, manage
733 the the include-dir and library-file requirements that go with that
734 choice. Non-TLS builds are still supported.
736 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
739 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
740 default. If the platform supports and has the facility enabled, it will
741 be requested on all coneections.
743 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
744 controlled by the build-time option SUPPORT_PIPE_CONNECT.
746 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
748 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
749 success-DSN messages. Previously the From: header was always the default
750 one for these; the option was ignored.
752 JH/28 Fix the timeout on smtp response to apply to the whole response.
753 Previously it was reset for every read, so a teergrubing peer sending
754 single bytes within the time limit could extend the connection for a
755 long time. Credit to Qualsys Security Advisory Team for the discovery.
757 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
758 delivery address, which leaked information of the results of local
759 forwarding. Change to the original envelope recipient address, per
762 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
763 requested. Previously not bounce was generated and a log entry of
764 error ignored was made.
766 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
768 JH/32 Introduce a general tainting mechanism for values read from the input
769 channel, and values derived from them. Refuse to expand any tainted
770 values, to catch one form of exploit.
772 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
773 was unused and the unexpanded text used for the test. Found and
774 fixed by Ruben Jenster.
776 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
777 an attempt to use a TLS library read routine dereffed a nul pointer,
780 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
781 them in our smtp response.
783 JH/36 Have the general router option retry_use_local_part default to true when
784 any of the restrictive preconditions are set (to anything). Previously it
785 was only for check_local user. The change removes one item of manual
786 configuration which is required for proper retries when a remote router
787 handles a subset of addresses for a domain.
789 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
790 link count into consideration.
792 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
793 caused the extension of big_buffer, the following lines were ignored.
795 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
796 accordance with RFC 2308. Previously there was no expiry, so a longlived
797 receive process (eg. due to ACL delays) versus a short SOA value could
800 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
802 JH/39 Promote DMARC support to mainline.
804 JH/40 Bug 2452: Add a References: header to DSNs.
806 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
807 parameters. The relevant library call is documented as "Deprecated: This
808 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
809 3.6.0, DH parameters are negotiated following RFC7919."
811 HS/06 Change the default of dnssec_request_domains to "*"
813 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
814 carried on and emitted a BDAT command, even when PIPELINING was not
817 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
818 buffer was used for the filename, resulting in a trap when tainted
819 arguments (eg. $domain) were used.
821 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
822 recommended to avoid a possible server-load attack. The feature can be
823 re-enabled via the openssl_options main cofiguration option.
825 JH/45 local_scan API: documented the current smtp_printf() call. This changed
826 for version 4.90 - adding a "more data" boolean to the arguments.
827 Bumped the ABI version number also, this having been missed previously;
828 release versions 4.90 to 4.92.3 inclusive were effectively broken in
829 respect of usage of smtp_printf() by either local_scan code or libraries
830 accessed via the ${dlfunc } expansion item. Both will need coding
831 adjustment for any calls to smtp_printf() to match the new function
832 signature; a FALSE value for the new argument is always safe.
834 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
835 the file-offset (which the Linux syscall does, and exim expects); this
836 resulted in an indefinite loop.
838 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
839 to do ARC verification. The Authentication-Results: header line added
840 by the configuration then had no ARC item.
842 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
843 an incorrect mode was used for reading the result, resulting in it being
850 JH/01 Remove code calling the customisable local_scan function, unless a new
851 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
853 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
854 non-signal-safe functions being used.
856 JH/03 Bug 2269: When presented with a received message having a stupidly large
857 number of DKIM-Signature headers, disable DKIM verification to avoid
858 a resource-consumption attack. The limit is set at twenty.
860 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
861 report of oldest_pass in ${authres } in consequence, and separate out
862 some descriptions of reasons for verification fail.
864 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
865 files in the spool were present and unlocked. A queue-runner could spot
866 them, resulting in a duplicate delivery. Fix that by doing the unlock
867 after the unlink. Investigation by Tim Stewart. Take the opportunity to
868 add more error-checking on spoolfile handling while that code is being
871 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
872 No known attacks, no CVE, this is defensive hardening.
874 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
875 a queue-runner could start a delivery while other operations were ongoing.
876 Cutthrough delivery was a common victim, resulting in duplicate delivery.
877 Found and investigated by Tim Stewart. Fix by using the open message data
878 file handle rather than opening another, and not locally closing it (which
879 releases a lock) for that case, while creating the temporary .eml format
880 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
882 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
883 $sender_verify_failure/$recipient_verify_failure to "random".
885 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
888 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
889 Previously this would segfault.
891 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
894 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
895 like zero, since the resolver should be doing this for us, But we need one
896 as a CNAME but no MX presence gets the CNAME returned; we need to check
897 that doesn't point to an MX to declare it "no MX returned" rather than
898 "error, loop". A new main option is added so the older capability of
899 following some limited number of chain links is maintained.
901 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
903 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
904 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
905 now-deprecated earlier definitions used only specified the range up to TLS
906 1.2 (in the older-version library docs).
908 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
910 JH/15 Rework TLS client-side context management. Stop using a global, and
911 explicitly pass a context around. This enables future use of TLS for
912 connections to service-daemons (eg. malware scanning) while a client smtp
913 connection is using TLS; with cutthrough connections this is quite likely.
915 JH/16 Fix ARC verification to do AS checks in reverse order.
917 JH/17 Support a "tls" option on the ${readsocket } expansion item.
919 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
920 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
921 Previously the "utf8" would be re-prepended for every additional message.
923 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
924 Previously thery were accepted, resulting in issues when attempting to
925 forward messages to a non-supporting MTA.
927 PP/02 Let -n work with printing macros too, not just options.
929 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
930 one parent address was copied, and bogus data was used at delivery-logging
931 time. Either a crash (after delivery) or bogus log data could result.
932 Discovery and analysis by Tim Stewart.
934 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
935 Previously if the string ended mid-character, we did not insert the
936 promised '?' replacement.
938 PP/04 Documentation: current string operators work on bytes, not codepoints.
940 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
941 should pack well giving a smaller memory footprint so better caching and
942 therefore performance. Group the declarations where this can't be done so
943 that the byte-sized flag variables are not interspersed among pointer
944 variables, giving a better chance of good packing by the compiler.
946 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
947 non-null, to avoid issues with sites running BATV. Previously reports were
948 sent with an empty envelope sender so looked like bounces.
950 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
951 The ignore_error flag wasn't being returned from the filter subprocess so
952 was not set for later routers. Investigation and fix by Matthias Kurz.
954 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
955 and a msg:complete for the whole, when a message is manually removed using
956 -Mrm. Developement by Matthias Kurz, hacked on by JH.
958 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
959 a "Gnu special" function, asprintf() in the DB utility binary builds; I
960 hope that is portable enough.
962 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
963 requiring a known-CA anchor certificate; make it now rely entirely on the
964 TLSA as an anchor. Checking the name on the leaf cert against the name
965 on the A-record for the host is still done for TA (but not for EE mode).
967 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
968 included in delivery lines for non-proxied connections, when compiled with
969 SUPPORT_SOCKS and running with proxy logging enabled.
971 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
972 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
973 move the existing event to fire before the normal logging of message
974 failure so that custom logging is bracketed by normal logging.
976 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
977 msg:fail:internal event. Developement by Matthias Kurz.
979 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
980 far too small for todays use of crypto signatures stored there. Go all
981 the way to the max DNS message size of 64kB, even though this might be
982 overmuch for IOT constrained device use.
984 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
985 copy a string over itself. The library routine is documented as not
986 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
988 JH/32 For main options check_spool_space and check_inode_space, where the
989 platform supports 64b integers, support more than the previous 2^31 kB
990 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
991 the previous G, M, k.
993 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
994 $authenticated_fail_id variable on authentication failure. Previously
997 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
998 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
999 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1000 GNUTLS_SEC_PARAM_MEDIUM.
1002 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1003 side. Previously we would continue as if no SNI had been received.
1005 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1006 "<" character, which should be a list-separator specification, we walked
1007 off past the nul-terimation.
1009 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1010 causes) even when the retry time is not yet met. Previously they were
1011 not, meaning that when (say) an account was over-quota and temp-rejecting,
1012 and multiple senders' messages were queued, only one sender would get
1013 notified on each configured delay_warning cycle.
1015 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1017 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1018 cipher-suites, an error can be left on the stack even for a succeeding
1019 accept; this results in impossible error messages when a later operation
1022 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1023 return error codes indicating retry. Under TLS1.3 this becomes required.
1025 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1026 it only wrote the new authenticators, resulting in a lack of tracking of
1027 peer changes of ESMTP extensions until the next cache flush.
1029 JH/41 Fix the loop reading a message header line to check for integer overflow,
1030 and more-often against header_maxsize. Previously a crafted message could
1031 induce a crash of the recive process; now the message is cleanly rejected.
1033 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1034 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1040 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1041 When redis_servers is set to a list of > 1 element, and the Redis servers
1042 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1043 case of MOVED into a DEFER case instead, thus moving the query onto the
1044 next server in the list. For a cluster of N elements, all N servers must
1045 be defined in redis_servers.
1047 GF/02 Catch and remove uninitialized value warning in exiqsumm
1048 Check for existence of @ARGV before looking at $ARGV[0]
1050 JH/01 Replace the store_release() internal interface with store_newblock(),
1051 which internalises the check required to safely use the old one, plus
1052 the allocate and data copy operations duplicated in both (!) of the
1053 extant use locations.
1055 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1056 modifier. This matches the restriction on the commandline.
1058 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1059 Previously only the last row was returned.
1061 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1062 we assumed that tags in the header were well-formed, and parsed the
1063 element content after inspecting only the first char of the tag.
1064 Assumptions at that stage could crash the receive process on malformed
1067 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1068 While running the DKIM ACL we operate on the Permanent memory pool so that
1069 variables created with "set" persist to the DATA ACL. Also (at any time)
1070 DNS lookups that fail create cache records using the Permanent pool. But
1071 expansions release any allocations made on the current pool - so a dnsdb
1072 lookup expansion done in the DKIM ACL releases the memory used for the
1073 DNS negative-cache, and bad things result. Solution is to switch to the
1074 Main pool for expansions.
1075 While we're in that code, add checks on the DNS cache during store_reset,
1076 active in the testsuite.
1077 Problem spotted, and debugging aided, by Wolfgang Breyha.
1079 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1080 When none of the hosts presented to a transport match an already-open
1081 connection, close it and proceed with the list. Previously we would
1082 queue the message. Spotted by Lena with Yahoo, probably involving
1085 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1086 Previously a spurious "250 OK id=" response was appended to the proper
1089 JH/08 The "support for" informational output now, which built with Content
1090 Scanning support, has a line for the malware scanner interfaces compiled
1091 in. Interface can be individually included or not at build time.
1093 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1094 by the template makefile "src/EDITME". The "STREAM" support for an older
1095 ClamAV interface method is removed.
1097 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1098 rows affected is given instead).
1100 JH/11 The runtime Berkeley DB library version is now additionally output by
1101 "exim -d -bV". Previously only the compile-time version was shown.
1103 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1104 SMTP connection. Previously, when one had more recipients than the
1105 first, an abortive onward connection was made. Move to full support for
1106 multiple onward connections in sequence, handling cutthrough connection
1107 for all multi-message initiating connections.
1109 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1110 routers. Previously, a multi-recipient message would fail to match the
1111 onward-connection opened for the first recipient, and cause its closure.
1113 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1114 a timeout on read on a GnuTLS initiating connection, resulting in the
1115 initiating connection being dropped. This mattered most when the callout
1116 was marked defer_ok. Fix to keep the two timeout-detection methods
1119 JH/15 Relax results from ACL control request to enable cutthrough, in
1120 unsupported situations, from error to silently (except under debug)
1121 ignoring. This covers use with PRDR, frozen messages, queue-only and
1124 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1126 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1127 metadata, resulting in a crash in free().
1129 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1130 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1131 Broken also in d185889f4, with init system revamp.
1133 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1134 Previously we abruptly closed the connection after reading a malware-
1135 found indication; now we go on to read the "scan ok" response line,
1138 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1139 ACL. Previously, a crash would result.
1141 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1142 macro text after a replacement (previously it was only once per line) and
1143 by skipping builtin macros when searching for an uppercase lead character.
1145 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1146 for the build is renamed.
1148 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1149 was allocated for every new TLS startup, meaning one per message. Fix
1150 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1152 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1153 reported the original. Fix to report (as far as possible) the ACL
1154 result replacing the original.
1156 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1157 OpenSSL. Certificate information is loaded for every new TLS startup,
1158 and the resources needed to be freed.
1160 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1162 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1165 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1166 DATA response info to the (existing) per-recipient response info for
1167 the "C=" log element. It can have useful tracking info from the
1168 destination system. Patch from Simon Arlott.
1170 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1171 length value. Previously this would segfault.
1173 HS/02 Support Avast multiline protoocol, this allows passing flags to
1174 newer versions of the scanner.
1176 JH/28 Ensure that variables possibly set during message acceptance are marked
1177 dead before release of memory in the daemon loop. This stops complaints
1178 about them when the debug_store option is enabled. Discovered specifically
1179 for sender_rate_period, but applies to a whole set of variables.
1180 Do the same for the queue-runner and queue-list loops, for variables set
1181 from spool message files. Do the same for the SMTP per-message loop, for
1182 certain variables indirectly set in ACL operations.
1184 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1185 as a multi-recipient message from a mailinglist manager). The coding had
1186 an arbitrary cutoff number of characters while checking for more input;
1187 enforced by writing a NUL into the buffer. This corrupted long / fast
1188 input. The problem was exposed more widely when more pipelineing of SMTP
1189 responses was introduced, and one Exim system was feeding another.
1190 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1191 receiving system, and refused recipients seen by the sending system
1192 (propating to people being dropped from mailing lists).
1193 Discovered and pinpointed by David Carter.
1195 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1196 replaced by the ${authresults } expansion.
1198 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1200 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1201 allows proper process termination in container environments.
1203 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1204 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1206 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1207 and "err_perm", deprecated since 4.83 when the RFC-defined words
1208 "temperror" and "permerror" were introduced.
1210 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1211 transport-filters or DKIM-signing. The restriction was lost in the
1212 consolidation of verify-callout and delivery SMTP handling.
1213 Extend the restriction to also cover ARC-signing.
1215 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1216 in defer=pass mode supply a 450 to the initiator. Previously the message
1219 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1220 tls_require_ciphers is used as before.
1222 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1223 "pass_unscanned". Only tmpfails from the scanner are written to
1224 the paniclog, as they may require admin intervention (permission
1225 denied, license issues). Other scanner errors (like decompression
1226 bombs) do not cause a paniclog entry.
1228 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1229 Previously it was possible to log spurious information in receive log
1232 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1233 triggered odd behaviour from Outlook Express clients.
1235 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1238 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1239 since the IETF WG has not yet settled on that versus the original
1240 "bare" representation.
1242 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1243 Previously the millisecond value corrupted the output.
1244 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1245 corrupted the output.
1251 JH/01 Rework error string handling in TLS interface so that the caller in
1252 more cases is responsible for logging. This permits library-sourced
1253 string to be attached to addresses during delivery, and collapses
1254 pairs of long lines into single ones.
1256 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1257 during configuration. Wildcards are allowed and expanded.
1259 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1260 This permits better logging.
1262 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1263 do not close it down and have the child transport start it up again on
1264 the passed-on TCP connection. Instead, proxy the child (and any
1265 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1266 affected: the continued delivery log lines do not have any DNSSEC, TLS
1267 Certificate or OCSP information. TLS cipher information is still logged.
1269 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1270 identical IP addresses on different listening ports. Will also affect
1273 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1274 add noisy ifdef guards to special-case this sillyness.
1275 Patch from Bernd Kuhls.
1277 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1278 than 255 are no longer allowed.
1280 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1281 Disable the session-cache too, which might reduce our load. Since we
1282 currrectly use a new context for every connection, both as server and
1283 client, there is no benefit for these.
1284 GnuTLS appears to not support tickets server-side by default (we don't
1285 call gnutls_session_ticket_enable_server()) but client side is enabled
1286 by default on recent versions (3.1.3 +) unless the PFS priority string
1289 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1290 <https://reproducible-builds.org/specs/source-date-epoch/>.
1292 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1293 the check for any unsuccessful recipients did not notice the limit, and
1294 erroneously found still-pending ones.
1296 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1297 MSG_MORE. Only in-clear (not on TLS connections).
1299 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1300 transport-filter is involved we can buffer the headers in memory for
1301 creating the signature, and read the spool data file once for the
1302 signature and again for transmission.
1304 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1305 4.77 as the kernel support then wasn't solid, having issues in 64bit
1306 mode. Now, it's been long enough. Add support for FreeBSD also.
1308 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1309 case where the routing stage had gathered several addresses to send to
1310 a host before calling the transport for the first, we previously failed
1311 to close down TLS in the old transport process before passing the TCP
1312 connection to the new process. The new one sent a STARTTLS command
1313 which naturally failed, giving a failed delivery and bloating the retry
1314 database. Investigation and fix prototype from Wolfgang Breyha.
1316 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1317 false-negatives in the check that the sender had not preempted a response
1318 or prompt from Exim (running as a server), due to that code's lack of
1319 awareness of the SMTP input buffering.
1321 PP/04 Add commandline_checks_require_admin option.
1322 Exim drops privileges sanely, various checks such as -be aren't a
1323 security problem, as long as you trust local users with access to their
1324 own account. When invoked by services which pass untrusted data to
1325 Exim, this might be an issue. Set this option in main configuration
1326 AND make fixes to the calling application, such as using `--` to stop
1329 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1330 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1331 responses to those, into a single TLS record each way (this usually means
1332 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1335 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1336 affects you only if you're dancing at the edge of the param size limits.
1337 If you are, and this message makes sense to you, then: raise the
1338 configured limit or use OpenSSL 1.1. Nothing we can do for older
1341 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1342 cmdline element to get the documented default one. Previously it was
1345 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1346 are unsuable for non-message-specific reasons - eg. connection timeout,
1347 banner-time rejection.
1349 JH/16 Fix logging of delivery remote port, when specified by router, under
1352 PP/06 Repair manualroute's ability to take options in any order, even if one
1353 is the name of a transport.
1356 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1358 JH/17 Change the list-building routines interface to use the expanding-string
1359 triplet model, for better allocation and copying behaviour.
1361 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1362 Previously it was constructed the first time a possibly-matching string
1363 was met in the configuration file input during startup; now it is done
1366 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1367 compatible one, to avoid the (poorly documented) possibility of a config
1368 file in the working directory redirecting the DB files, possibly correpting
1369 some existing file. CVE-2017-10140 assigned for BDB.
1371 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1372 cache-hot. Previously, although the result was properly cached, the
1373 initial verify call returned a defer.
1375 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1376 the main verify for receipient in uncached-mode.
1378 JH/22 Retire historical build files to an "unsupported" subdir. These are
1379 defined as "ones for which we have no current evidence of testing".
1381 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1382 if present. Previously it was ignored.
1384 JH/24 Start using specified-initialisers in C structure init coding. This is
1385 a C99 feature (it's 2017, so now considered safe).
1387 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1388 if was a fixed-sized field and bitmask ops via macros; it is now more
1391 PP/07 GitHub PR 56: Apply MariaDB build fix.
1392 Patch provided by Jaroslav Škarvada.
1394 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1395 during Coverity cleanups [4.87 JH/47]
1396 Diagnosis and fix provided by Michael Fischer v. Mollard.
1398 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1399 the right size to place the terminating semicolon on its own folded
1400 line, the header hash was calculated to an incorrect value thanks to
1401 the (relaxed) space the fold became.
1403 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1404 and confused the parent.
1406 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1407 which could crash as a result. This could lead to undeliverable messages.
1409 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1412 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1413 matters on fast-turnover and PID-randomising systems, which were getting
1414 out-of-order delivery.
1416 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1417 a possibly-overlapping copy. The symptom was that "Remote host closed
1418 connection in response to HELO" was logged instead of the actual 4xx
1421 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1422 Previously only that bufferd was discarded, resulting in SYMTP command
1425 JH/32 DKIM: when a message has multiple signatures matching an identity given
1426 in dkim_verify_signers, run the dkim acl once for each. Previously only
1427 one run was done. Bug 2189.
1429 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1430 "panic the current process" to "deliberately defer". The panic log is
1431 still written with the problem list name; the mail and reject logs now
1432 get a temp-reject line for the message that was being handled, saying
1433 something like "domains check lookup or other defer". The SMTP 451
1434 message is still "Temporary local problem".
1436 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1437 A crafted sequence of BDAT commands could result in in-use memory beeing
1438 freed. CVE-2017-16943.
1440 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1441 from SMTP input. Previously it was always done; now only done for DATA
1442 and not BDAT commands. CVE-2017-16944.
1444 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1445 to the message (such as an overlong header line). Previously this was
1446 not done and we did not exit BDAT mode. Followon from the previous item
1447 though a different problem.
1453 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1454 than -2003 did; needs libidn2 in addition to libidn.
1456 JH/02 The path option on a pipe transport is now expanded before use.
1458 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1459 Patch provided by "Björn", documentation fix added too.
1461 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1462 missing a wire-to-host endian conversion.
1464 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1465 close after a BDAT command line could be taken as a following command,
1466 giving a synch failure. Fix by only checking for synch immediately
1467 before acknowledging the chunk.
1469 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1470 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1471 macro. Patches provided by Josh Soref.
1473 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1474 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1475 by IANA as a service extension. However, John Klensin suggests that we
1478 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1479 the dkim code may be unix-mode line endings rather than smtp wire-format
1480 CRLF, so prepend a CR to any bare LF.
1482 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1483 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1485 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1486 the first were themselves being wrongly included in the feed into dkim
1487 processing; with most chunk sizes in use this resulted in an incorrect
1488 body hash calculated value.
1490 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1491 DKIM signature block, for verification. Although advised against by
1492 standards it is specifically not ruled illegal.
1494 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1496 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1497 missing a body hash (the bh= tag).
1499 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1500 It seems that HAProxy sends the Proxy Protocol information in clear and
1501 only then does a TLS startup, so do the same.
1503 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1504 TCP connections (such as for Spamd) unless the daemon successfully set
1505 Fast Open mode on its listening sockets. This fixes breakage seen on
1506 too-old kernels or those not configured for Fast Open, at the cost of
1507 requiring both directions being enabled for TFO, and TFO never being used
1508 by non-daemon-related Exim processes.
1510 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1511 endings, at least on the first header line. Try to canonify any that get
1512 past that check, despite the cost.
1514 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1515 now limited to an arbitrary five deep, while parsing addresses with the
1516 strip_excess_angle_brackets option enabled.
1518 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1519 instead leave the unprompted TLS handshake in socket buffer for the
1520 TLS library to consume.
1522 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1524 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1526 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1527 region in one operation, for speed, and this leaves assigned pointers
1528 dangling. Add checks run only under the testsuite which checks all
1529 variables at a store-reset and panics on a dangling pointer; add code
1530 explicitly nulling out all the variables discovered. Fixes one known
1531 bug: a transport crash, where a dangling pointer for $sending_ip_address
1532 originally assigned in a verify callout, is re-used.
1534 PP/06 Drop '.' from @INC in various Perl scripts.
1536 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1538 PP/08 Reduce a number of compilation warnings under clang; building with
1539 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1540 should be warning-free.
1542 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1544 HS/01 Fix portability problems introduced by PP/08 for platforms where
1545 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1547 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1548 chunk. This allows us to accept broken chunked messages. We need a more
1549 general solution here.
1551 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1552 already-broken messages in the queue.
1554 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1556 JH/19 Fix reference counting bug in routing-generated-address tracking.
1562 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1563 supports it and a size is available (ie. the sending peer gave us one).
1565 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1566 years of being deprecated). The replacements are the ACLs
1567 acl_smtp_mime and acl_not_smtp_mime.
1569 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1570 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1571 or even an in-clear connection were permitted. Now, if the host lookup
1572 was dnssec and dane was requested then the host is only used if the
1573 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1574 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1575 if one fails this test.
1576 This means that a poorly-configured remote DNS will make it incommunicado;
1577 but it protects against a DNS-interception attack on it.
1579 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1580 non-noisy when a race steals the message being considered.
1582 JH/05 If main configuration option tls_certificate is unset, generate a
1583 self-signed certificate for inbound TLS connections.
1585 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1586 in rewrites and routers.
1588 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1589 and logged a warning sing 4.83; now they are a configuration file error.
1591 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1592 (lacking @domain). Apply the same qualification processing as RCPT.
1594 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1596 JH/10 Support ${sha256:} applied to a string (as well as the previous
1599 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1600 a cutthrough deliver is pending, as we always want to make a connection.
1601 This also avoids re-routing the message when later placing the cutthrough
1602 connection after a verify cache hit.
1603 Do not update it with the verify result either.
1605 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1606 when routing results in more than one destination address.
1608 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1609 signing (which inhibits the cutthrough capability). Previously only
1610 the presence of an option was tested; now an expansion evaluating as
1611 empty is permissible (obviously it should depend only on data available
1612 when the cutthrough connection is made).
1614 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1615 the relevant preceding SMTP command did not note the pipelining mode.
1617 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1618 Previously they were not counted.
1620 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1621 as one having no matching records. Previously we deferred the message
1622 that needed the lookup.
1624 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1625 distinguished as "(=".
1627 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1628 for missing MX records. Previously it only worked for missing A records.
1630 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1632 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1633 after the data-go-ahead and data-ack. Patch from Jason Betts.
1635 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1636 even for a "none" policy. Patch from Tony Meyer.
1638 JH/22 Fix continued use of a connection for further deliveries. If a port was
1639 specified by a router, it must also match for the delivery to be
1642 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1643 When one of the recipients of a message was unsuitable for the connection
1644 (has no matching addresses), we lost track of needing to mark it
1645 deferred. As a result mail would be lost.
1647 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1649 JH/25 Decoding ACL controls is now done using a binary search; the source code
1650 takes up less space and should be simpler to maintain. Merge the ACL
1651 condition decode tables also, with similar effect.
1653 JH/26 Fix problem with one_time used on a redirect router which returned the
1654 parent address unchanged. A retry would see the parent address marked as
1655 delivered, so not attempt the (identical) child. As a result mail would
1658 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1659 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1660 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1663 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1664 Main config options check_{log,spool}_{inodes,space} are now
1665 100 inodes, 10MB unless set otherwise in the configuration.
1667 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1668 Previously it only applied to the main-section connection policy
1671 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1673 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1674 by me. Added RFC7919 DH primes as an alternative.
1676 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1677 are not in the system include path.
1679 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1680 GnuTLS, when a session startup failed (eg because the client disconnected)
1681 Exim did stdio operations after fclose. This was exposed by a recent
1682 change which nulled out the file handle after the fclose.
1684 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1685 signed directly by the cert-signing cert, rather than an intermediate
1686 OCSP-signing cert. This is the model used by LetsEncrypt.
1688 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1690 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1691 an incoming connection.
1693 HS/02 Bug 1802: Do not half-close the connection after sending a request
1696 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1697 fallback to "prime256v1".
1699 JH/34 SECURITY: Use proper copy of DATA command in error message.
1700 Could leak key material. Remotely exploitable. CVE-2016-9963.
1706 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1707 and 3.4.4 - once the server is enabled to respond to an OCSP request
1708 it does even when not requested, resulting in a stapling non-aware
1709 client dropping the TLS connection.
1711 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1712 support variable-length bit vectors. No functional change.
1714 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1715 The I= interface field on outgoing lines is now after the H= remote
1716 host field, same as incoming lines. There is a separate
1717 outgoing_interface log selector which allows you to disable the
1720 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1721 If not running log_selector +smtp_connection the mainlog would be held
1722 open indefinitely after a "too many connections" event, including to a
1723 deleted file after a log rotate. Leave the per net connection logging
1724 leaving it open for efficiency as that will be quickly detected by the
1725 check on the next write.
1727 HS/01 Bug 1671: Fix post transport crash.
1728 Processing the wait-<transport> messages could crash the delivery
1729 process if the message IDs didn't exist for some reason. When
1730 using 'split_spool_directory=yes' the construction of the spool
1731 file name failed already, exposing the same netto behaviour.
1733 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1734 mime_regex ACL conditions.
1736 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1737 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1738 to HELO, local diagnostic string.
1740 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1741 log line to debug. Even when configured with a tls authenticator many
1742 client connections are expected to not authenticate in this way, so
1743 an authenticate fail is not an error.
1745 HS/02 Add the Exim version string to the process info. This way exiwhat
1746 gives some more detail about the running daemon.
1748 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1749 matter for fast-change records such as DNSBLs.
1751 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1752 retry record, even if constant. There may be multiple transports with
1753 different interface settings and the retry behaviour needs to be kept
1756 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1758 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1760 JH/10 Bug 840: fix log_defer_output option of pipe transport
1762 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1763 hosts_randomize. This matters a lot when combined with mua_wrapper.
1765 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1766 ${quote_pgsql:<string>} operator.
1768 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1770 JH/14 Change header file name for freeradius-client. Relevant if compiling
1771 with Radius support; from the Gentoo tree and checked under Fedora.
1773 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1775 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1776 option transport_filter, meaning no filtering.
1778 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1780 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1781 defaults to "*" (all hosts). The variable is now available when not built
1782 with TLS, default unset, mainly to enable keeping the testsuite sane.
1783 If a server certificate is not supplied (via tls_certificate) an error is
1784 logged, and clients will find TLS connections fail on startup. Presumably
1785 they will retry in-clear.
1786 Packagers of Exim are strongly encouraged to create a server certificate
1787 at installation time.
1789 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1790 with the $config_file variable.
1792 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1793 in transport context, after the attempt, and per-recipient. The latter type
1794 is per host attempted. The event data is the error message, and the errno
1795 information encodes the lookup type (A vs. MX) used for the (first) host,
1796 and the trailing two digits of the smtp 4xx response.
1798 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1799 to write to mainlog (or rejectlog, paniclog) in the window between file
1800 creation and permissions/ownership being changed. Particularly affects
1801 installations where exicyclog is run as root, rather than exim user;
1802 result is that the running daemon panics and dies.
1804 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1806 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1807 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1808 "pri" and "weight". Note that the previous implicit priority given by the
1809 list order is no longer honoured.
1811 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1812 for DKIM processing.
1814 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1815 by defining SUPPORT_SOCKS.
1817 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1818 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1819 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1820 variables are renamed to proxy_{local,external}_{address,port}.
1822 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1823 for a build by defining SUPPORT_I18N
1825 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1826 of the query string, and make ${quote_redis:} do that quoting.
1828 JH/29 Move Events support from Experimental to mainline, enabled by default
1829 and removable for a build by defining DISABLE_EVENT.
1831 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1833 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1834 cached by the daemon.
1836 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1837 by defining LOOKUP_REDIS. The libhiredis library is required.
1839 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1840 keys are given for lookup.
1842 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1843 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1844 only supported when built with TLS support. The PolarSSL SHA routines
1845 are still used when the TLS library is too old for convenient support.
1847 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1848 openssl_options), for security. OpenSSL forces this from version 1.1.0
1849 server-side so match that on older versions.
1851 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1852 allocation for $value could be released as the expansion processing
1853 concluded, but leaving the global pointer active for it.
1855 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1856 and to use the domains and local_parts ACL conditions.
1858 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1859 incorrectly not doubled on cutthrough transmission, hence seen as a
1860 body-termination at the receiving system - resulting in truncated mails.
1861 Commonly the sender saw a TCP-level error, and retransmitted the message
1862 via the normal store-and-forward channel. This could result in duplicates
1863 received - but deduplicating mailstores were liable to retain only the
1864 initial truncated version.
1866 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1868 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1870 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1871 we're in there, support oversigning also; bug 1309.
1873 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1875 HS/04 Add support for keep_environment and add_environment options.
1877 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1878 either intentional arithmetic overflow during PRNG, or testing config-
1881 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1882 delivery resulted in actual delivery. Cancel cutthrough before DATA
1885 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1886 rejects a recipient: pass the reject to the originator.
1888 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1889 Many were false-positives and ignorable, but it's worth fixing the
1892 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1893 for the new environment-manipulation done at startup. Move the routines
1894 from being local to tls.c to being global via the os.c file.
1896 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1897 an extract embedded as result-arg for a map, the first arg for extract
1898 is unavailable so we cannot tell if this is a numbered or keyed
1899 extraction. Accept either.
1905 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1908 JH/02 The smtp transport option "multi_domain" is now expanded.
1910 JH/03 The smtp transport now requests PRDR by default, if the server offers
1913 JH/04 Certificate name checking on server certificates, when exim is a client,
1914 is now done by default. The transport option tls_verify_cert_hostnames
1915 can be used to disable this per-host. The build option
1916 EXPERIMENTAL_CERTNAMES is withdrawn.
1918 JH/05 The value of the tls_verify_certificates smtp transport and main options
1919 default to the word "system" to access the system default CA bundle.
1920 For GnuTLS, only version 3.0.20 or later.
1922 JH/06 Verification of the server certificate for a TLS connection is now tried
1923 (but not required) by default. The verification status is now logged by
1924 default, for both outbound TLS and client-certificate supplying inbound
1927 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1930 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1931 Status Notification (bounce) messages are now MIME format per RFC 3464.
1932 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1933 under the control of the dsn_advertise_hosts option, and routers may
1934 have a dsn_lasthop option.
1936 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1937 default, modifiable by a malware= option. The list separator for
1938 the options can now be changed in the usual way. Bug 68.
1940 JH/10 The smtp_receive_timeout main option is now expanded before use.
1942 JH/11 The incoming_interface log option now also enables logging of the
1943 local interface on delivery outgoing connections.
1945 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1946 if the interface and destination host and port all match.
1948 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1951 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1952 Patch from Andrew Lewis.
1954 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1955 now supports optional time-restrictions, weighting, and priority
1956 modifiers per server. Patch originally by <rommer@active.by>.
1958 JH/16 The spamd_address main option now supports a mixed list of local
1959 and remote servers. Remote servers can be IPv6 addresses, and
1960 specify a port-range.
1962 JH/17 Bug 68: The spamd_address main option now supports an optional
1963 timeout value per server.
1965 JH/18 Bug 1581: Router and transport options headers_add/remove can
1966 now have the list separator specified.
1968 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1971 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1974 JH/21 Support for the A6 type of dns record is withdrawn.
1976 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1977 rather than the verbs used.
1979 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1980 from 255 to 1024 chars.
1982 JH/24 Verification callouts now attempt to use TLS by default.
1984 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1985 are generic router options now. The defaults didn't change.
1987 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1988 Original patch from Alexander Shikoff, worked over by JH.
1990 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1991 files if ZCAT_COMMAND is not executable.
1993 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1995 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1997 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1998 Normally benign, it bites when the pair was led to by a CNAME;
1999 modern usage is to not canonicalize the domain to a CNAME target
2000 (and we were inconsistent anyway for A-only vs AAAA+A).
2002 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2004 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2005 when evaluating $sender_host_dnssec.
2007 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2008 $sender_helo_dnssec variable.
2010 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2012 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2014 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2016 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2017 documented as working, but never had. Support all but $spam_report.
2019 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2020 added for tls authenticator.
2022 HS/03 Add perl_taintmode main config option
2028 TL/01 When running the test suite, the README says that variables such as
2029 no_msglog_check are global and can be placed anywhere in a specific
2030 test's script, however it was observed that placement needed to be near
2031 the beginning for it to behave that way. Changed the runtest perl
2032 script to read through the entire script once to detect and set these
2033 variables, reset to the beginning of the script, and then run through
2034 the script parsing/test process like normal.
2036 TL/02 The BSD's have an arc4random API. One of the functions to induce
2037 adding randomness was arc4random_stir(), but it has been removed in
2038 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2039 function when detected.
2041 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2042 cause callback expansion.
2044 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2045 syntax errors in an expansion can be treated as a string instead of
2046 logging or causing an error, due to the internal use of bool_lax
2047 instead of bool when processing it.
2049 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2050 server certificates when making smtp deliveries.
2052 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2054 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2056 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2058 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2059 Merged patch from Sebastian Wiedenroth.
2061 JH/05 Fix results-pipe from transport process. Several recipients, combined
2062 with certificate use, exposed issues where response data items split
2063 over buffer boundaries were not parsed properly. This eventually
2064 resulted in duplicates being sent. This issue only became common enough
2065 to notice due to the introduction of connection certificate information,
2066 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2068 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2069 size buffer was used, resulting in syntax errors when an expansion
2072 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2073 version 3.3.6 or later.
2075 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2076 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2077 both become "event_action", the variables become $event_name, $event_data
2078 and $event_defer_errno. There is a new variable $verify_mode, usable in
2079 routers, transports and related events. The tls:cert event is now also
2080 raised for inbound connections, if the main configuration event_action
2083 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2084 early OCSP support, but no stapling (appears to be less than 1.0.0).
2086 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2087 server certificate names available under the smtp transport option
2088 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2091 JH/10 Time-related extraction expansions from certificates now use the main
2092 option "timezone" setting for output formatting, and are consistent
2093 between OpenSSL and GnuTLS compilations. Bug 1541.
2095 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2096 encoded parameter in the incoming message. Bug 1558.
2098 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2099 include certificate info, eximon was claiming there were spoolfile
2102 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2104 JH/14 Log delivery-related information more consistently, using the sequence
2105 "H=<name> [<ip>]" wherever possible.
2107 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2108 are problematic for Debian distribution, omit them from the release
2111 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2113 JH/16 Fix string representation of time values on 64bit time_t architectures.
2116 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2117 output list separator was used.
2122 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2123 checkers that were complaining about end of non-void function with no
2126 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2127 This was a regression introduced in 4.83 by another bugfix.
2129 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2131 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2132 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2138 TF/01 Correctly close the server side of TLS when forking for delivery.
2140 When a message was received over SMTP with TLS, Exim failed to clear up
2141 the incoming connection properly after forking off the child process to
2142 deliver the message. In some situations the subsequent outgoing
2143 delivery connection happened to have the same fd number as the incoming
2144 connection previously had. Exim would try to use TLS and fail, logging
2145 a "Bad file descriptor" error.
2147 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2148 utilities have not been installed.
2150 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2151 temporary space as the ACL may create new global variables.
2153 TL/01 LDAP support uses per connection or global context settings, depending
2154 upon the detected version of the libraries at build time.
2156 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2157 to extract and use the src ip:port in logging and expansions as if it
2158 were a direct connection from the outside internet. PPv2 support was
2159 updated based on HAProxy spec change in May 2014.
2161 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2163 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2164 Properly escape header and check for NULL return.
2166 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2169 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2171 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2172 characters in header names, implemented as a verify condition.
2173 Contributed by Michael Fischer v. Mollard.
2175 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2176 results permerror and temperror. Previous values are deprecated but
2177 still accepted. In a future release, err_perm and err_temp will be
2178 completely removed, which will be a backward incompatibility if the
2179 ACL tests for either of these two old results. Patch contributed by
2180 user bes-internal on the mailing list.
2182 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2184 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2185 selectors, in both main and reject logs.
2187 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2190 JH/07 Add malware type "sock" for talking to simple daemon.
2192 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2194 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2195 routers/transports under cutthrough routing.
2197 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2198 numbers. Touch up "bool" conditional to keep the same definition.
2200 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2202 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2204 JH/12 Expand items in router/transport headers_add or headers_remove lists
2205 individually rather than the list as a whole. Bug 1452.
2207 Required for reasonable handling of multiple headers_ options when
2208 they may be empty; requires that headers_remove items with embedded
2209 colons must have them doubled (or the list-separator changed).
2211 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2212 view the policy declared in the DMARC record. Currently, $dmarc_status
2213 is a combined value of both the record presence and the result of the
2216 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2218 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2219 dnslookup router and the smtp transport (applying to the forward
2222 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2223 of ldap servers used for a specific lookup. Patch provided by Heiko
2226 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2227 New variable $lookup_dnssec_authenticated for observability.
2229 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2230 Patch submitted by Lars Timman.
2232 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2234 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2235 Requires trusted mode and valid format message id, aborts otherwise.
2236 Patch contributed by Heiko Schlichting.
2238 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2239 certextract with support for various fields. Bug 1358.
2241 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2242 is requested by default, modifiable by smtp transport option
2245 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2246 operate on certificate variables to give certificate fingerprints
2247 Also new ${sha256:cert_variable}.
2249 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2251 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2252 Christian Aistleitner.
2254 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2256 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2257 file. Patch from Wolfgang Breyha.
2259 JH/25 Expand the coverage of the delivery $host and $host_address to
2260 client authenticators run in verify callout. Bug 1476.
2262 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2263 align with daemon_smtp_ports. Bug 72.
2265 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2266 support and error reporting did not work properly.
2268 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2269 and is readable. Patch from Andrew Colin Kissa.
2271 TL/14 Enhance documentation of ${run expansion and how it parses the
2272 commandline after expansion, particularly in the case when an
2273 unquoted variable expansion results in an empty value.
2275 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2277 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2278 by renaming away from it. Addresses GH issue 15, reported by
2281 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2283 TL/15 SECURITY: prevent double expansion in math comparison functions
2284 (can expand unsanitized data). Not remotely exploitable.
2291 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2293 PP/02 Make -n do something, by making it not do something.
2294 When combined with -bP, the name of an option is not output.
2296 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2299 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2300 $sender_host_name and config options to manage this, and basic check
2303 PP/05 DSCP support for outbound connections and control modifier for inbound.
2305 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2306 (Only plugin which currently uses this is kerberos4, which nobody should
2307 be using, but we should make it available and other future plugins might
2308 conceivably use it, even though it would break NAT; stuff *should* be
2309 using channel bindings instead).
2311 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2312 name; added for Sendmail compatibility; requires admin caller.
2313 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2314 just ignore it); requires trusted caller.
2315 Also parse but ignore: -Ac -Am -X<logfile>
2318 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2320 TL/02 Add +smtp_confirmation as a default logging option.
2322 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2323 Patch by Magnus Holmgren from 2007-02-20.
2325 TL/04 Bugzilla 1281 - Spec typo.
2326 Bugzilla 1283 - Spec typo.
2327 Bugzilla 1290 - Spec grammar fixes.
2329 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2331 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2333 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2334 mailing list by by Dmitry Isaikin.
2336 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2338 JH/02 Support "G" suffix to numbers in ${if comparisons.
2340 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2342 NM/01 Bugzilla 1197 - Spec typo
2343 Bugzilla 1196 - Spec examples corrections
2345 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2347 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2348 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2351 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2352 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2354 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2355 "acl {{name}{arg}...}", and optional args on acl condition
2358 JH/05 Permit multiple router/transport headers_add/remove lines.
2360 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2362 JH/07 Avoid using a waiting database for a single-message-only transport.
2363 Performance patch from Paul Fisher. Bugzilla 1262.
2365 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2368 JH/09 Add $headers_added variable, with content from use of ACL modifier
2369 add_header (but not yet added to the message). Bugzilla 199.
2371 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2372 Pulled from Bugzilla 817 by Wolfgang Breyha.
2374 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2376 (nb: this is the same fix as in Exim 4.80.1)
2378 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2381 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2384 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2386 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2387 advertise SMTP AUTH mechanism to us, instead of a generic
2388 protocol violation error. Also, make Exim more robust to bad
2389 data from the Dovecot auth socket.
2391 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2393 When a queue runner is handling a message, Exim first routes the
2394 recipient addresses, during which it prunes them based on the retry
2395 hints database. After that it attempts to deliver the message to
2396 any remaining recipients. It then updates the hints database using
2399 So if a recipient address works intermittently, it can get repeatedly
2400 deferred at routing time. The retry hints record remains fresh so the
2401 address never reaches the final cutoff time.
2403 This is a fairly common occurrence when a user is bumping up against
2404 their storage quota. Exim had some logic in its local delivery code
2405 to deal with this. However it did not apply to per-recipient defers
2406 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2408 This change adds a proper retry rule check during routing so that the
2409 final cutoff time is checked against the message's age. We only do
2410 this check if there is an address retry record and there is not a
2411 domain retry record; this implies that previous attempts to handle
2412 the address had the retry_use_local_parts option turned on. We use
2413 this as an approximation for the destination being like a local
2414 delivery, as in LMTP.
2416 I suspect this new check makes the old local delivery cutoff check
2417 redundant, but I have not verified this so I left the code in place.
2419 TF/02 Correct gecos expansion when From: is a prefix of the username.
2421 Test 0254 submits a message to Exim with the header
2425 When I ran the test suite under the user fanf2, Exim expanded
2426 the header to contain my full name, whereas it should have added
2427 a Resent-Sender: header. It erroneously treats any prefix of the
2428 username as equal to the username.
2430 This change corrects that bug.
2432 GF/01 DCC debug and logging tidyup
2433 Error conditions log to paniclog rather than rejectlog.
2434 Debug lines prefixed by "DCC: " to remove any ambiguity.
2436 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2438 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2439 Bug spotted by Jeremy Harris; was flawed since initial commit.
2440 Would have resulted in OCSP responses post-SNI triggering an Exim
2441 NULL dereference and crash.
2443 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2445 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2446 Bug detection, analysis and fix by Samuel Thibault.
2447 Bugzilla 1331, Debian bug #698092.
2449 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2451 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2452 Server implementation by Todd Lyons, client by JH.
2453 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2454 config variable "prdr_enable" controls whether the server
2455 advertises the facility. If the client requests PRDR a new
2456 acl_data_smtp_prdr ACL is called once for each recipient, after
2457 the body content is received and before the acl_smtp_data ACL.
2458 The client is controlled by both of: a hosts_try_prdr option
2459 on the smtp transport, and the server advertisement.
2460 Default client logging of deliveries and rejections involving
2461 PRDR are flagged with the string "PRDR".
2463 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2464 fclose(). Diagnosis by Todd Lyons.
2466 PP/17 Update configure.default to handle IPv6 localhost better.
2467 Patch by Alain Williams (plus minor tweaks).
2470 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2471 This is now consistent with GnuTLS, and is now documented: the
2472 previous undocumented portable approach to treating the option as
2473 unset was to force an expansion failure. That still works, and
2474 an empty string is now equivalent.
2476 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2477 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2478 not performing validation itself.
2480 PP/20 Added force_command boolean option to pipe transport.
2481 Patch from Nick Koston, of cPanel Inc.
2483 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2486 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2488 PP/21 Fix eximon continuous updating with timestamped log-files.
2489 Broken in a format-string cleanup in 4.80, missed when I repaired the
2490 other false fix of the same issue.
2491 Report and fix from Heiko Schlichting.
2494 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2495 Report from Prashanth Katuri.
2497 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2498 It's SecureTransport, so affects any MacOS clients which use the
2499 system-integrated TLS libraries, including email clients.
2501 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2502 using a MIME ACL for non-SMTP local injection.
2503 Report and assistance in diagnosis by Warren Baker.
2505 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2507 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2509 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2510 last id that failed so it may be referenced in subsequent ACL's.
2512 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2515 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2516 ldap library initialization, allowing self-signed CA's to be
2517 used. Also properly sets require_cert option later in code by
2518 using NULL (global ldap config) instead of ldap handle (per
2519 session). Bug diagnosis and testing by alxgomz.
2521 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2522 the src/util/ subdirectory.
2524 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2525 renamed to Transport Post Delivery Action by Jeremy Harris, as
2528 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2529 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2530 redis_servers = needs to be configured which will be used by the redis
2531 lookup. Patch from Warren Baker, of The Packet Hub.
2533 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2535 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2536 hostname or reverse DNS when processing a host list. Used suggestions
2537 from multiple comments on this bug.
2539 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2541 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2542 Missed a few lines, added it to make the runtest require no keyboard
2545 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2546 contains upper case chars. Make router use caseful_local_part.
2548 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2549 support when GnuTLS has been built with p11-kit.
2555 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2557 This, or similar/improved, will also be change PP/11 of 4.82.
2563 PP/01 Handle short writes when writing local log-files.
2564 In practice, only affects FreeBSD (8 onwards).
2565 Bugzilla 1053, with thanks to Dmitry Isaikin.
2567 NM/01 Bugzilla 949 - Documentation tweak
2569 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2572 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2574 PP/02 Implemented gsasl authenticator.
2576 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2578 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2579 `pkg-config foo` for cflags/libs.
2581 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2582 with rest of GSASL and with heimdal_gssapi.
2584 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2585 `pkg-config foo` for cflags/libs for the TLS implementation.
2587 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2588 properties get this fed in as external SSF. A number of robustness
2589 and debugging improvements to the cyrus_sasl authenticator.
2591 PP/08 cyrus_sasl server now expands the server_realm option.
2593 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2594 Patch by Jeremy Harris.
2596 PP/10 Added dbmjz lookup type.
2598 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2600 PP/12 MAIL args handles TAB as well as SP, for better interop with
2601 non-compliant senders.
2602 Analysis and variant patch by Todd Lyons.
2604 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2605 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2606 Patch from Dirk Mueller <dmueller@suse.com>
2608 PP/13 tls_peerdn now print-escaped for spool files.
2609 Observed some $tls_peerdn in wild which contained \n, which resulted
2610 in spool file corruption.
2612 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2613 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2614 or write after TLS renegotiation, which otherwise led to messages
2617 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2618 as a tracking header (ie: a signed header comes before the signature).
2619 Patch from Wolfgang Breyha.
2621 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2622 comma-sep list; embedded commas doubled.
2624 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2626 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2628 Report and patch from Dmitry Banschikov.
2630 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2631 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2632 failures. We appear to now support TLS1.1+ with Exim.
2634 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2635 lets Exim select keys and certificates based upon TLS SNI from client.
2636 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2637 before an outbound SMTP session. New log_selector, +tls_sni.
2639 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2640 NULL dereference. Report and patch from Alun Jones.
2642 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2643 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2644 Not seeing resolver debug output on NetBSD, but suspect this is a
2645 resolver implementation change.
2647 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2648 Left warnings. Added "eximon gdb" invocation mode.
2650 PP/21 Defaulting "accept_8bitmime" to true, not false.
2652 PP/22 Added -bw for inetd wait mode support.
2654 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2655 locate the relevant includes and libraries. Made this the default.
2657 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2658 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2660 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2661 This may cause build issues on older platforms.
2663 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2664 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2665 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2666 Added SNI support via GnuTLS too.
2667 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2669 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2671 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2672 Applied second patch from Janne, implementing suggestion to default
2673 multiple-strings-in-record handling to match SPF spec.
2675 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2677 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2678 read-only, out of scope).
2679 Patch from Wolfgang Breyha, report from Stuart Northfield.
2681 PP/29 Fix three issues highlighted by clang analyser static analysis.
2682 Only crash-plausible issue would require the Cambridge-specific
2683 iplookup router and a misconfiguration.
2684 Report from Marcin Mirosław.
2686 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2688 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2689 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2690 As part of this, removing so much warning spew let me fix some minor
2691 real issues in debug logging.
2693 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2694 assignment on my part. Fixed.
2696 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2697 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2698 Janne Snabb (who went above and beyond: thank you).
2700 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2701 string otherwise requires a connection and a bunch more work and it's
2702 relatively easy to get wrong. Should also expose TLS library linkage
2705 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2706 64-bit ${eval} (JH/03).
2708 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2709 GNU libc to support some of the 64-bit stuff, should not lead to
2710 conflicts. Defined before os.h is pulled in, so if a given platform
2711 needs to override this, it can.
2713 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2714 protection layer was required, which is not implemented.
2715 Bugzilla 1254, patch from Wolfgang Breyha.
2717 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2718 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2719 tls_dhparam take prime identifiers. Also unbreak combination of
2720 OpenSSL+DH_params+TLSSNI.
2722 PP/39 Disable SSLv2 by default in OpenSSL support.
2728 PP/01 Solaris build fix for Oracle's LDAP libraries.
2729 Bugzilla 1109, patch from Stephen Usher.
2731 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2733 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2736 TF/02 Fix a couple more cases where we did not log the error message
2737 when unlink() failed. See also change 4.74-TF/03.
2739 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2740 lock up or crash if it happened to be inside a call to libc when it
2741 got a SIGUSR1 from exiwhat.
2743 The SIGUSR1 handler appends the current process status to the process
2744 log which is later printed by exiwhat. It used to use the general
2745 purpose logging code to do this, but several functions it calls are
2746 not safe for signals.
2748 The new output code in the SIGUSR1 handler is specific to the process
2749 log, and simple enough that it's easy to inspect for signal safety.
2750 Removing some special cases also simplifies the general logging code.
2751 Removing the spurious timestamps from the process log simplifies
2754 TF/04 Improved ratelimit ACL condition.
2756 The /noupdate option has been deprecated in favour of /readonly which
2757 has clearer semantics. The /leaky, /strict, and /readonly update modes
2758 are mutually exclusive. The update mode is no longer included in the
2759 database key; it just determines when the database is updated. (This
2760 means that when you upgrade Exim will forget old rate measurements.)
2762 Exim now checks that the per_* options are used with an update mode that
2763 makes sense for the current ACL. For example, when Exim is processing a
2764 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2765 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2766 must specify per_mail/readonly. If you omit the update mode it defaults to
2767 /leaky where that makes sense (as before) or /readonly where required.
2769 The /noupdate option is now undocumented but still supported for
2770 backwards compatibility. It is equivalent to /readonly except that in
2771 ACLs where /readonly is required you may specify /leaky/noupdate or
2772 /strict/noupdate which are treated the same as /readonly.
2774 A useful new feature is the /count= option. This is a generalization
2775 of the per_byte option, so that you can measure the throughput of other
2776 aggregate values. For example, the per_byte option is now equivalent
2777 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2779 The per_rcpt option has been generalized using the /count= mechanism
2780 (though it's more complicated than the per_byte equivalence). When it is
2781 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2782 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2783 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2784 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2785 Note that using per_rcpt with a non-readonly update mode in more than
2786 one ACL will cause the recipients to be double-counted. (The per_mail
2787 and per_byte options don't have this problem.)
2789 The handling of very low rates has changed slightly. If the computed rate
2790 is less than the event's count (usually one) then this event is the first
2791 after a long gap. In this case the rate is set to the same as this event's
2792 count, so that the first message of a spam run is counted properly.
2794 The major new feature is a mechanism for counting the rate of unique
2795 events. The new per_addr option counts the number of different
2796 recipients that someone has sent messages to in the last time period. It
2797 behaves like per_rcpt if all the recipient addresses are different, but
2798 duplicate recipient addresses do not increase the measured rate. Like
2799 the /count= option this is a general mechanism, so the per_addr option
2800 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2801 example, measure the rate that a client uses different sender addresses
2802 with the options per_mail/unique=$sender_address. There are further
2803 details in the main documentation.
2805 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2807 TF/06 Removed a few PCRE remnants.
2809 TF/07 Automatically extract Exim's version number from tags in the git
2810 repository when doing development or release builds.
2812 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2813 Bugzilla 879. Patch from Paul Fisher.
2815 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2816 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2819 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2821 PP/05 Variable $av_failed, true if the AV scanner deferred.
2822 Bugzilla 1078. Patch from John Horne.
2824 PP/06 Stop make process more reliably on build failure.
2825 Bugzilla 1087. Patch from Heiko Schlittermann.
2827 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2828 Bugzilla 1089. Patch from Heiko Schlittermann.
2830 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2831 Bugzilla 1131. Patch from Holger Weiß.
2833 PP/09 Handle IPv6 addresses with SPF.
2834 Bugzilla 860. Patch from Wolfgang Breyha.
2836 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2838 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2841 PP/11 match_* no longer expand right-hand-side by default.
2842 New compile-time build option, EXPAND_LISTMATCH_RHS.
2843 New expansion conditions, "inlist", "inlisti".
2845 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2847 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2849 PP/14 fix log_write() format string regression from TF/03.
2850 Bugzilla 1152. Patch from Dmitry Isaikin.
2856 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2858 PP/02 Harmonised TLS library version reporting; only show if debugging.
2859 Layout now matches that introduced for other libraries in 4.74 PP/03.
2861 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2863 PP/04 New "dns_use_edns0" global option.
2865 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2868 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2869 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2871 TK/01 Updated PolarSSL code to 0.14.2.
2872 Bugzilla 1097. Patch from Andreas Metzler.
2874 PP/07 Catch divide-by-zero in ${eval:...}.
2875 Fixes bugzilla 1102.
2877 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2880 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2881 format-string attack -- SECURITY: remote arbitrary code execution.
2883 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2884 time unintentionally subject to list matching rules, letting the header
2885 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2886 arbitrary string expansion). This allowed for information disclosure.
2888 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2889 INT_MIN/-1 -- value coerced to INT_MAX.
2895 NM/01 Workaround for PCRE version dependency in version reporting
2898 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2899 This fixes portability to compilers other than gcc, notably
2900 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2902 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2903 makefiles for portability to HP-UX and POSIX correctness.
2905 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2906 Also via indented variable definition in the Makefile.
2907 (Debugging by Oliver Heesakkers).
2909 PP/02 Restore caching of spamd results with expanded spamd_address.
2910 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2912 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2913 Improves build reliability. Fix from: Frank Elsner
2915 NM/02 Fix wide character breakage in the rfc2047 coding
2916 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2918 NM/03 Allow underscore in dnslist lookups
2919 Fixes bug 1026. Patch from Graeme Fowler
2921 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2922 Code patches from Adam Ciarcinski of NetBSD.
2924 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2927 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2928 is logged, to avoid truncation. Patch from John Horne.
2930 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2931 Patch from Jakob Hirsch.
2933 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2934 SQL string expansion failure details.
2935 Patch from Andrey Oktyabrski.
2937 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2938 Patch from Simon Arlott.
2940 PP/09 New lookups functionality failed to compile on old gcc which rejects
2941 extern declarations in function scope.
2942 Patch from Oliver Fleischmann
2944 PP/10 Use sig_atomic_t for flags set from signal handlers.
2945 Check getgroups() return and improve debugging.
2946 Fixed developed for diagnosis in bug 927 (which turned out to be
2949 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2950 Patch from Mark Zealey.
2952 PP/12 Bugzilla 1056: Improved spamd server selection.
2953 Patch from Mark Zealey.
2955 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2956 Based on patch from Heiko Schlittermann.
2958 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2959 Patch from Uwe Doering, sign-off by Michael Haardt.
2961 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2962 variable declaration deep within a block. Bug and patch from
2965 PP/15 lookups-Makefile IRIX compatibility coercion.
2967 PP/16 Make DISABLE_DKIM build knob functional.
2969 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2970 Patch by Simon Arlott
2972 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2973 variable argument macros. Our copy now differs from upstream.
2979 TF/01 Failure to get a lock on a hints database can have serious
2980 consequences so log it to the panic log.
2982 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2983 controlled using the smtp_confirmation log selector.
2985 TF/03 Include the error message when we fail to unlink a spool file.
2987 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2988 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2989 for maintaining out-of-tree patches for some time.
2991 PP/01 Bugzilla 139: Documentation and portability issues.
2992 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2993 Handle per-OS dynamic-module compilation flags.
2995 PP/02 Let /dev/null have normal permissions.
2996 The 4.73 fixes were a little too stringent and complained about the
2997 permissions on /dev/null. Exempt it from some checks.
2998 Reported by Andreas M. Kirchwitz.
3000 PP/03 Report version information for many libraries, including
3001 Exim version information for dynamically loaded libraries. Created
3002 version.h, now support a version extension string for distributors
3003 who patch heavily. Dynamic module ABI change.
3005 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3006 privilege escalation vulnerability whereby the Exim run-time user
3007 can cause root to append content of the attacker's choosing to
3010 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3013 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3014 If dropping privileges for untrusted macros, we disabled normal logging
3015 on the basis that it would fail; for the Exim run-time user, this is not
3016 the case, and it resulted in successful deliveries going unlogged.
3017 Fixed. Reported by Andreas Metzler.
3023 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3024 only prepend for the Resent-* case. Fixes regression introduced in
3025 Exim 4.70 by NM/22 for Bugzilla 607.
3027 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3028 increasing numbers of administrators be bitten by this.
3030 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3032 PP/03 Bugzilla 994: added openssl_options main configuration option.
3034 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3036 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3038 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3040 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3041 without a peer certificate, leading to a segfault because of an
3042 assumption that peers always have certificates. Be a little more
3043 paranoid. Problem reported by Martin Tscholak.
3045 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3046 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3047 NB: ClamAV planning to remove STREAM in "middle of 2010".
3048 CL also introduces -bmalware, various -d+acl logging additions and
3049 more caution in buffer sizes.
3051 PP/09 Implemented reverse_ip expansion operator.
3053 PP/10 Bugzilla 937: provide a "debug" ACL control.
3055 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3057 PP/12 Bugzilla 973: Implement --version.
3059 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3061 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3063 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3065 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3066 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3067 ignore trailing whitespace.
3069 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3071 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3074 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3075 Notification from Dr Andrew Aitchison.
3077 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3078 ExtendedDetectionInfo response format.
3079 Notification from John Horne.
3081 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3084 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3085 XSL and documented dependency on system catalogs, with examples of how
3088 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3091 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3092 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3093 configuration file which is writeable by the Exim user or group.
3095 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3096 of configuration files to cover files specified with the -C option if
3097 they are going to be used with root privileges, not just the default
3100 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3101 option (effectively making it always true).
3103 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3104 files to be used while preserving root privileges.
3106 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3107 that rogue child processes cannot use them.
3109 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3110 run-time user, instead of root.
3112 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3113 Exim run-time user without dropping privileges.
3115 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3116 result string, instead of calling string_vformat() twice with the same
3119 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3120 for other users. Others should always drop root privileges if they use
3121 -C on the command line, even for a whitelisted configure file.
3123 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3125 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3131 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3132 $data_path, and $header_path variables; fixed documentation bugs and
3135 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3136 exipick to access non-standard spools, including the "frozen" queue
3139 NM/01 Bugzilla 965: Support mysql stored procedures.
3140 Patch from Alain Williams
3142 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3144 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3145 Patch from Andreas Metzler
3147 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3148 Patch from Kirill Miazine
3150 NM/05 Bugzilla 671: Added umask to procmail example.
3152 JJ/03 installed exipick 20100323.0, fixing doc bug
3154 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3155 directory. Notification and patch from Dan Rosenberg.
3157 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3159 TK/02 Improve log output when DKIM signing operation fails.
3161 MH/01 Treat the transport option dkim_domain as a colon separated
3162 list, not as a single string, and sign the message with each element,
3163 omitting multiple occurences of the same signer.
3165 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3166 Bugzilla 985, 986. Patch by Simon Arlott
3168 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3169 Patch by Simon Arlott
3171 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3172 MBX locking. Notification from Dan Rosenberg.
3178 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3180 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3182 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3184 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3186 NM/04 Bugzilla 915: Flag broken perl installation during build.
3192 TK/01 Added patch by Johannes Berg that expands the main option
3193 "spamd_address" if it starts with a dollar sign.
3195 TK/02 Write list of recipients to X-Envelope-Sender header when building
3196 the mbox-format spool file for content scanning (suggested by Jakob
3199 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3200 (http://www.dcc-servers.net/) support via dccifd. Activated by
3201 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3203 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3204 by Mark Daniel Reidel <mr@df.eu>.
3206 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3207 When building exim an external PCRE library is now needed -
3208 PCRE is a system library on the majority of modern systems.
3209 See entry on PCRE_LIBS in EDITME file.
3211 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3212 conversation. Added nologin parameter to request.
3213 Patch contributed by Kirill Miazine.
3215 TF/01 Do not log submission mode rewrites if they do not change the address.
3217 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3219 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3220 log files in place. Contributed by Roberto Lima.
3222 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3224 TF/03 Bugzilla 615: When checking the local_parts router precondition
3225 after a local_part_suffix or local_part_prefix option, Exim now
3226 does not use the address's named list lookup cache, since this
3227 contains cached lookups for the whole local part.
3229 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3230 Robert Millan. Documentation is in experimental-spec.txt.
3232 TF/04 Bugzilla 668: Fix parallel build (make -j).
3234 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3236 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3237 Patch provided by Jan Srzednicki.
3239 TF/05 Leading white space used to be stripped from $spam_report which
3240 wrecked the formatting. Now it is preserved.
3242 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3243 that they are available at delivery time.
3245 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3247 TF/08 TLS error reporting now respects the incoming_interface and
3248 incoming_port log selectors.
3250 TF/09 Produce a more useful error message if an SMTP transport's hosts
3251 setting expands to an empty string.
3253 NM/06 Bugzilla 744: EXPN did not work under TLS.
3254 Patch provided by Phil Pennock.
3256 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3257 Patch provided by Richard Godbee.
3259 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3260 acl_smtp_notquit, added index entry.
3262 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3263 Patch provided by Eugene Bujak.
3265 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3266 accept(). Patch provided by Maxim Dounin.
3268 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3269 Patch provided by Phil Pennock.
3271 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3273 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3274 Patch provided by Brad "anomie" Jorsch.
3276 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3277 Patch provided by Dean Brooks.
3279 TK/05 Add native DKIM support (does not depend on external libraries).
3281 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3282 Patch provided by Graeme Fowler.
3284 NM/16 Bugzilla 851: Documentation example syntax fix.
3286 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3288 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3291 NM/19 Bugzilla 745: TLS version reporting.
3292 Patch provided by Phil Pennock.
3294 NM/20 Bugzilla 167: bool: condition support.
3295 Patch provided by Phil Pennock.
3297 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3298 clients. Patch provided by Phil Pennock.
3300 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3301 Patch provided by Brad "anomie" Jorsch.
3303 NM/23 Bugzilla 687: Fix misparses in eximstats.
3304 Patch provided by Heiko Schlittermann.
3306 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3307 Patch provided by Heiko Schlittermann.
3309 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3310 plus update to original patch.
3312 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3314 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3315 Patch provided by David Brownlee.
3317 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3319 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3321 NM/30 Bugzilla 888: TLS documentation bugfixes.
3323 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3325 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3326 Unlike the original bugzilla I have changed all shell scripts in src tree.
3328 NM/33 Bugzilla 898: Transport filter timeout fix.
3329 Patch by Todd Rinaldo.
3331 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3332 Patch by Serge Demonchaux.
3334 NM/35 Bugzilla 39: Base64 decode bug fixes.
3335 Patch by Jakob Hirsch.
3337 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3339 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3341 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3343 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3349 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3350 ALT-N's libdkim that I have put here:
3351 http://duncanthrax.net/exim-experimental/
3353 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3354 were called 'true' and it seems that C99 defines that as a reserved
3355 keyword to be used with 'bool' variable types. That means you could
3356 not include C99-style headers which use bools without triggering
3357 build errors in sieve.c.
3359 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3360 as mailq or other aliases. Changed the --help handling significantly
3361 to do whats expected. exim_usage() emits usage/help information.
3363 SC/01 Added the -bylocaldomain option to eximstats.
3365 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3367 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3369 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3375 PH/01 Another patch from the Sieve maintainer.
3377 PH/02 When an IPv6 address is converted to a string for single-key lookup
3378 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3379 dots are used instead of colons so that keys in lsearch files need not
3380 contain colons. This was done some time before quoting was made available
3381 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3382 (notated using the quote facility) so as to distinguish them from IPv4
3383 keys. This meant that lookups for IP addresses in host lists did not work
3384 for iplsearch lookups.
3386 This has been fixed by arranging for IPv6 addresses to be expressed with
3387 colons if the lookup type is iplsearch. This is not incompatible, because
3388 previously such lookups could never work.
3390 The situation is now rather anomalous, since one *can* have colons in
3391 ordinary lsearch keys. However, making the change in all cases is
3392 incompatible and would probably break a number of configurations.
3394 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3397 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3398 conversion specification without a maximum field width, thereby enabling
3399 a rogue spamd server to cause a buffer overflow. While nobody in their
3400 right mind would setup Exim to query an untrusted spamd server, an
3401 attacker that gains access to a server running spamd could potentially
3402 exploit this vulnerability to run arbitrary code as the Exim user.
3404 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3405 $primary_hostname instead of what libspf2 thinks the hosts name is.
3407 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3408 a directory entry by the name of the lookup key. Previously, if a
3409 symlink pointed to a non-existing file or a file in a directory that
3410 Exim lacked permissions to read, a lookup for a key matching that
3411 symlink would fail. Now it is enough that a matching directory entry
3412 exists, symlink or not. (Bugzilla 503.)
3414 PH/03 The body_linecount and body_zerocount variables are now exported in the
3417 PH/04 Added the $dnslist_matched variable.
3419 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3420 This means they are set thereafter only if the connection becomes
3423 PH/06 Added the client_condition to authenticators so that some can be skipped
3424 by clients under certain conditions.
3426 PH/07 The error message for a badly-placed control=no_multiline_responses left
3427 "_responses" off the end of the name.
3429 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3431 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3432 (without spaces) instead of just copying the configuration text.
3434 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3436 PH/11 Added $max_received_linelength.
3438 PH/12 Added +ignore_defer and +include_defer to host lists.
3440 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3441 way in which PCRE > 7.0 is built.
3443 PH/14 Implemented queue_only_load_latch.
3445 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3446 MAIL command. The effect was to mangle the value on 64-bit systems.
3448 PH/16 Another patch from the Sieve maintainer.
3450 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3452 PH/18 If a system quota error occurred while trying to create the file for
3453 a maildir delivery, the message "Mailbox is full" was not appended to the
3454 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3455 only to a quota excession during the actual writing of the file.
3457 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3458 characters?) which causes problems in log lines. The DN values are now
3459 passed through string_printing() before being added to log lines.
3461 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3462 and InterBase are left for another time.)
3464 PH/21 Added message_body_newlines option.
3466 PH/22 Guard against possible overflow in moan_check_errorcopy().
3468 PH/23 POSIX allows open() to be a macro; guard against that.
3470 PH/24 If the recipient of an error message contained an @ in the local part
3471 (suitably quoted, of course), incorrect values were put in $domain and
3472 $local_part during the evaluation of errors_copy.
3478 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3479 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3482 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3483 issue a MAIL command.
3485 PH/02 In an ACL statement such as
3487 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3489 if a client was not listed at all, or was listed with a value other than
3490 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3491 the condition was not true (as it should be), so access was not denied.
3492 The bug was that the ! inversion was incorrectly passed on to the second
3493 item. This has been fixed.
3495 PH/03 Added additional dnslists conditions == and =& which are different from
3496 = and & when the dns lookup returns more than one IP address.
3498 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3499 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3501 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3502 FSYNC, which compiles an option called disable_fsync that allows for
3503 bypassing fsync(). The documentation is heavily laced with warnings.
3505 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3507 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3508 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3509 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3510 including adding "make clean"; (3) Added -fPIC when compiling the test
3511 dynamically loaded module, to get rid of a warning.
3513 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3514 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3515 = 0s. The bug is otherwise harmless.
3517 PH/07 There was a bug in the dovecot authenticator such that the value of
3518 $auth1 could be overwritten, and so not correctly preserved, after a
3519 successful authentication. This usually meant that the value preserved by
3520 the server_setid option was incorrect.
3522 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3524 PH/09 Installed PCRE release 7.0.
3526 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3527 run for batched SMTP input. It is now run at the start of every message
3528 in the batch. While fixing this I discovered that the process information
3529 (output by running exiwhat) was not always getting set for -bs and -bS
3530 input. This is fixed, and it now also says "batched" for BSMTP.
3532 PH/11 Added control=no_pipelining.
3534 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3535 patch, slightly modified), and move the expansion of helo_data till after
3536 the connection is made in the smtp transport (so it can use these
3539 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3541 PH/14 Added log_selector = +pid.
3543 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3545 PH/16 Add ${if forany and ${if forall.
3547 PH/17 Added dsn_from option to vary the From: line in DSNs.
3549 PH/18 Flush SMTP output before performing a callout, unless control =
3550 no_callout_flush is set.
3552 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3553 was true (the default) a successful delivery failed to delete the retry
3554 item, thus causing premature timeout of the address. The bug is now
3557 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3559 PH/21 Long custom messages for fakedefer and fakereject are now split up
3560 into multiline responses in the same way that messages for "deny" and
3561 other ACL rejections are.
3563 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3564 with slight modification.
3566 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3567 draft, changing the syntax and factoring some duplicate code".
3569 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3570 for deliveries of the second and subsequent messages over the same SMTP
3573 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3574 ${reduce, with only minor "tidies".
3576 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3578 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3579 expansion side effects.
3581 PH/27 When a message times out after an over-quota error from an Exim-imposed
3582 quota, the bounce message says "mailbox is full". This message was not
3583 being given when it was a system quota that was exceeded. It now should
3586 MH/03 Made $recipients available in local_scan(). local_scan() already has
3587 better access to the recipient list through recipients_list[], but
3588 $recipients can be useful in postmaster-provided expansion strings.
3590 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3591 in the case of a MAIL command with additional options following the
3592 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3593 were accidentally chopped off.
3595 PH/29 SMTP synchronization checks are implemented when a command is read -
3596 there is a check that no more input is waiting when there shouldn't be
3597 any. However, for some commands, a delay in an ACL can mean that it is
3598 some time before the response is written. In this time, more input might
3599 arrive, invalidly. So now there are extra checks after an ACL has run for
3600 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3601 pipelining has not been advertised.
3603 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3605 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3606 returned in the SMTP response when smtp_return_error_details was false.
3607 This has been fixed.
3609 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3610 instead of the C I/O that was originally supplied, because problems were
3611 reported on Solaris.
3613 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3614 Exim which did not show up earlier: it was assuming that a call to
3615 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3616 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3617 was a macro that became an assignment, so it seemed to work. This has
3618 changed to a proper function call with a void return, hence the compile
3619 error. Exim's code has been fixed.
3621 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3624 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3626 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3628 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3629 because it turns out to be too big or there is a timeout) there is no
3630 "Completed" line in the log. When some messages of this type were
3631 selected by exigrep, they were listed as "not completed". Others were
3632 picked up by some special patterns. I have improved the selection
3633 criteria to be more general.
3635 PH/38 The host_find_failed option in the manualroute router can now be set
3636 to "ignore", to completely ignore a host whose IP address cannot be
3637 found. If all hosts are ignored, the behaviour is controlled by the new
3638 host_all_ignored option.
3640 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3641 homing or because of multiple MX records with /mx) generated more than
3642 one IP address, and the following item turned out to be the local host,
3643 all the secondary addresses of the first item were incorrectly removed
3644 from the list, along with the local host and any following hosts (which
3645 is what is supposed to happen).
3647 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3648 whoever called Exim into the -H file. In the case of the daemon it was
3649 behaving confusingly. When first started, it used values for whoever
3650 started the daemon, but after a SIGHUP it used the Exim user (because it
3651 calls itself on a restart). I have changed the code so that it now always
3654 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3655 message are rejected with the same error (e.g. no authentication or bad
3656 sender address), and a DATA command is nevertheless sent (as can happen
3657 with PIPELINING or a stupid MUA), the error message that was given to the
3658 RCPT commands is included in the rejection of the DATA command. This is
3659 intended to be helpful for MUAs that show only the final error to their
3662 PH/42 Another patch from the Sieve maintainer.
3664 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3665 Eximstats - Fixed some broken HTML links and added missing column headers
3667 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3668 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3670 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3672 PH/43 Yet another patch from the Sieve maintainer.
3674 PH/44 I found a way to check for a TCP/IP connection going away before sending
3675 the response to the final '.' that terminates a message, but only in the
3676 case where the client has not sent further data following the '.'
3677 (unfortunately, this is allowed). However, in many cases there won't be
3678 any further data because there won't be any more messages to send. A call
3679 to select() can be used: if it shows that the input is "ready", there is
3680 either input waiting, or the socket has been closed. An attempt to read
3681 the next input character can distinguish the two cases. Previously, Exim
3682 would have sent an OK response which the client would never have see.
3683 This could lead to message repetition. This fix should cure that, at
3684 least in a lot of common cases.
3686 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3687 advertised in response to EHLO.
3693 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3694 fixed by 4.65/MH/01 (is this a record?) are fixed:
3696 (i) An empty string was always treated as zero by the numeric comparison
3697 operators. This behaviour has been restored.
3699 (ii) It is documented that the numeric comparison operators always treat
3700 their arguments as decimal numbers. This was broken in that numbers
3701 starting with 0 were being interpreted as octal.
3703 While fixing these problems I realized that there was another issue that
3704 hadn't been noticed. Values of message_size_limit (both the global option
3705 and the transport option) were treated as octal if they started with 0.
3706 The documentation was vague. These values are now always treated as
3707 decimal, and I will make that clear in the documentation.
3713 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3714 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3717 MH/01 Don't check that the operands of numeric comparison operators are
3718 integers when their expansion is in "skipping" mode (fixes bug
3719 introduced by 4.64-PH/07).
3721 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3722 child addresses, Exim now panics and dies. Previously, because the count
3723 is held in a short int, deliveries were likely to be lost. As such a
3724 large number of recipients for a single message is ridiculous
3725 (performance will be very, very poor), I have chosen to impose a limit
3726 rather than extend the field.
3732 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3733 leftover -K file (the existence of which was triggered by #402).
3734 While we were at it, introduced process PID as part of the -K
3735 filename. This should rule out race conditions when creating
3738 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3739 processing considerably. Previous code took too long for large mails,
3740 triggering a timeout which in turn triggers #401.
3742 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3743 in the DK code in transports.c. sendfile() is not really portable,
3744 hence the _LINUX specificness.
3746 TF/01 In the add_headers option to the mail command in an Exim filter,
3747 there was a bug that Exim would claim a syntax error in any
3748 header after the first one which had an odd number of characters
3751 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3752 callout verification, Exim cached a "reject" for the entire domain. This
3753 is correct for most verifications, but it is not correct for a recipient
3754 verification with use_sender or use_postmaster set, because in that case
3755 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3756 case of MAIL FROM:<> rejection from other early rejections (e.g.
3757 rejection of HELO). When verifying a recipient using a non-null MAIL
3758 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3759 Whatever the result of the callout, the value of the domain cache is
3760 left unchanged (for any other kind of callout, getting as far as trying
3761 RCPT means that the domain itself is ok).
3763 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3766 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3767 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3768 session, the daemon ignores ECONNECT errors and logs others; it now
3769 ignores EPIPE as well.
3771 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3772 (quoted-printable decoding).
3774 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3775 later the small subsequent patch to fix an introduced bug.
3777 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3779 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3781 PH/08 An error is now given if message_size_limit is specified negative.
3783 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3784 to be given (somewhat) arbitrary names.
3786 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3789 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3790 miscellaneous code fixes
3792 PH/10 Added the log_reject_target ACL modifier to specify where to log
3795 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3796 hostname. This is wrong, because it relates to the incoming message (and
3797 probably the interface on which it is arriving) and not to the outgoing
3798 callout (which could be using a different interface). This has been
3799 changed to use the value of the helo_data option from the smtp transport
3800 instead - this is what is used when a message is actually being sent. If
3801 there is no remote transport (possible with a router that sets up host
3802 addresses), $smtp_active_hostname is used.
3804 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3805 tweaks were necessary in order to get it to work (see also 21 below):
3806 (a) The code assumed that strncpy() returns a negative number on buffer
3807 overflow, which isn't the case. Replaced with Exim's string_format()
3809 (b) There were several signed/unsigned issues. I just did the minimum
3810 hacking in of casts. There is scope for a larger refactoring.
3811 (c) The code used strcasecmp() which is not a standard C function.
3812 Replaced with Exim's strcmpic() function.
3813 (d) The code set only $1; it now sets $auth1 as well.
3814 (e) A simple test gave the error "authentication client didn't specify
3815 service in request". It would seem that Dovecot has changed its
3816 interface. Fortunately there's a specification; I followed it and
3817 changed what the client sends and it appears to be working now.
3819 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3822 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3823 address A is aliased to B and C, where B exists and C does not. Without
3824 -v the output is "A verified" because verification stops after a
3825 successful redirection if more than one address is generated. However,
3826 with -v the child addresses are also verified. Exim was outputting "A
3827 failed to verify" and then showing the successful verification for C,
3828 with its parentage. It now outputs "B failed to verify", showing B's
3829 parentage before showing the successful verification of C.
3831 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3832 look up a TXT record in a specific list after matching in a combined
3835 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3836 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3837 they consult the DNS. I had assumed they would set it the way they
3838 wanted; and indeed my experiments on Linux seem to show that in some
3839 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3840 To be on the safe side, however, I have now made the interface to
3841 host_find_byname() similar to host_find_bydns(), with an argument
3842 containing the DNS resolver options. The host_find_byname() function now
3843 sets these options at its start, just as host_find_bydns() does. The smtp
3844 transport options dns_qualify_single and dns_search_parents are passed to
3845 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3846 of host_find_byname() use the default settings of RES_DEFNAMES
3847 (qualify_single) but not RES_DNSRCH (search_parents).
3849 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3850 spool_read_header() do less string testing, by means of a preliminary
3851 switch on the second character of optional "-foo" lines. (This is
3852 overdue, caused by the large number of possibilities that now exist.
3853 Originally there were few.) While I was there, I also converted the
3854 str(n)cmp tests so they don't re-test the leading "-" and the first
3855 character, in the hope this might squeeze out yet more improvement.
3857 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3858 flag allowing group syntax was set by the header_syntax check but not
3859 turned off, possible causing trouble later; (2) The flag was not being
3860 set at all for the header_verify test, causing "group"-style headers to
3861 be rejected. I have now set it in this case, and also caused header_
3862 verify to ignore an empty address taken from a group. While doing this, I
3863 came across some other cases where the code for allowing group syntax
3864 while scanning a header line wasn't quite right (mostly, not resetting
3865 the flag correctly in the right place). These bugs could have caused
3866 trouble for malformed header lines. I hope it is now all correct.
3868 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3869 with the "reply" argument non-NULL. The code, however (which originally
3870 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3871 but it didn't always do it. This confused somebody who was copying the
3872 code for some other use. I have removed all the tests.
3874 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3875 feature that was used to support insecure browsers during the U.S. crypto
3876 embargo. It requires special client support, and Exim is probably the
3877 only MTA that supported it -- and would never use it because real RSA is
3878 always available. This code has been removed, because it had the bad
3879 effect of slowing Exim down by computing (never used) parameters for the
3880 RSA_EXPORT functionality.
3882 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3883 authenticator to fail if there's a tab character in the incoming data
3884 (there should never be unless someone is messing about, as it's supposed
3885 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3886 if the connection is using TLS or if the remote IP is the same as the
3887 local IP, and the "valid-client-cert option" if a client certificate has
3890 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3891 authenticators. This can be used for authorization after authentication
3892 succeeds. (In the case of plaintext, it servers for both authentication
3895 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3896 if any retry times were supplied.
3898 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3899 connection, where there is no client IP address to check. In this
3900 situation, the verify now always succeeds.
3902 PH/25 Applied John Jetmore's -Mset patch.
3904 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3906 PH/27 In a string expansion for a processed (not raw) header when multiple
3907 headers of the same name were present, leading whitespace was being
3908 removed from all of them, but trailing whitespace was being removed only
3909 from the last one. Now trailing whitespace is removed from each header
3910 before concatenation. Completely empty headers in a concatenation (as
3911 before) are ignored.
3913 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3914 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3916 PH/29 [Removed. This was a change that I later backed out, and forgot to
3917 correct the ChangeLog entry (that I had efficiently created) before
3918 committing the later change.]
3920 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3921 address errors (4xx response to RCPT) over the same connection as other
3922 messages routed to the same hosts. Such deliveries are always "forced",
3923 so retry times are not inspected. This resulted in far too many retries
3924 for the affected addresses. The effect occurred only when there were more
3925 hosts than the hosts_max_try setting in the smtp transport when it had
3926 the 4xx errors. Those hosts that it had tried were not added to the list
3927 of hosts for which the message was waiting, so if all were tried, there
3928 was no problem. Two fixes have been applied:
3930 (i) If there are any address or message errors in an SMTP delivery, none
3931 of the hosts (tried or untried) are now added to the list of hosts
3932 for which the message is waiting, so the message should not be a
3933 candidate for sending over the same connection that was used for a
3934 successful delivery of some other message. This seems entirely
3935 reasonable: after all the message is NOT "waiting for some host".
3936 This is so "obvious" that I'm not sure why it wasn't done
3937 previously. Hope I haven't missed anything, but it can't do any
3938 harm, as the worst effect is to miss an optimization.
3940 (ii) If, despite (i), such a delivery is accidentally attempted, the
3941 routing retry time is respected, so at least it doesn't keep
3942 hammering the server.
3944 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3945 in ${readsocket because some servers need this prod.
3947 PH/32 Added some extra debug output when updating a wait-xxx database.
3949 PH/33 The hint "could be header name not terminated by colon", which has been
3950 given for certain expansion errors for a long time, was not being given
3951 for the ${if def:h_colon_omitted{... case.
3953 PH/34 The spec says: "With one important exception, whenever a domain list is
3954 being scanned, $domain contains the subject domain." There was at least
3955 one case where this was not true.
3957 PH/35 The error "getsockname() failed: connection reset by peer" was being
3958 written to the panic log as well as the main log, but it isn't really
3959 panic-worthy as it just means the connection died rather early on. I have
3960 removed the panic log writing for the ECONNRESET error when getsockname()
3963 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3964 runs only) independently of the message's sender address. This meant
3965 that, if the 4xx error was in fact related to the sender, a different
3966 message to the same recipient with a different sender could confuse
3967 things. In particular, this can happen when sending to a greylisting
3968 server, but other circumstances could also provoke similar problems.
3969 I have changed the default so that the retry time for these errors is now
3970 based a combination of the sender and recipient addresses. This change
3971 can be overridden by setting address_retry_include_sender=false in the
3974 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3975 remote server are returned as part of bounce messages. This was not
3976 happening for LMTP over a pipe (the lmtp transport), but now it is the
3977 same for both kinds of LMTP.
3979 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3980 in header lines that were in fact CNAMEs. This is no longer the case.
3982 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3983 and queue runs started by the daemon processed all messages. This has
3984 been fixed so that -R and -S can now usefully be given with -q<time>.
3986 PH/40 Import PCRE release 6.7 (fixes some bugs).
3988 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3990 PH/42 Give an error if -q is specified more than once.
3992 PH/43 Renamed the variables $interface_address and $interface_port as
3993 $received_ip_address and $received_port, to make it clear that these
3994 values apply to message reception, and not to the outgoing interface when
3995 a message is delivered. (The old names remain recognized, of course.)
3997 PH/44 There was no timeout on the connect() call when using a Unix domain
3998 socket in the ${readsocket expansion. There now is.
4000 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4001 be meaningful with "accept".
4003 SC/01 Eximstats V1.43
4004 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4006 SC/02 Eximstats V1.44
4007 Use a glob alias rather than an array ref in the generated
4008 parser. This improves both readability and performance.
4010 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4011 Collect SpamAssassin and rejection statistics.
4012 Don't display local sender or destination tables unless
4013 there is data to show.
4014 Added average volumes into the top table text output.
4016 SC/04 Eximstats V1.46
4017 Collect data on the number of addresses (recipients)
4018 as well as the number of messages.
4020 SC/05 Eximstats V1.47
4021 Added 'Message too big' to the list of mail rejection
4022 reasons (thanks to Marco Gaiarin).
4024 SC/06 Eximstats V1.48
4025 Mainlog lines which have GMT offsets and are too short to
4026 have a flag are now skipped.
4028 SC/07 Eximstats V1.49 (Alain Williams)
4029 Added the -emptyok flag.
4031 SC/08 Eximstats V1.50
4032 Fixes for obtaining the IP address from reject messages.
4034 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4035 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4036 whitespace changes from 4.64-PH/27
4038 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4041 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4042 are found, allow negative numbers in numeric criteria)
4044 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4046 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4047 to match changes made in 4.64-PH/43
4049 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4051 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4052 "domain" is actually an IP address, and force a failure. This locks out
4053 those revolvers/nameservers that support "A-for-A" lookups, in
4054 contravention of the specifications.
4056 PH/48 When a host name was looked up from an IP address, and the subsequent
4057 forward lookup of the name timed out, the host name was left in
4058 $sender_host_name, contrary to the specification.
4060 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4061 restricted to single-key lookups, Exim was not diagnosing an error if
4062 * or *@ was used with a query-style lookup.
4064 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4066 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4067 long ago, but noone interested enough thought of it. Let's just say that
4068 the "1.1" means that there are some new functions that weren't there at
4069 some point in the past.
4071 PH/51 Error processing for expansion failure of helo_data from an smtp
4072 transport during callout processing was broken.
4074 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4075 tested/used via the -bh/-bhc/-bs options.
4077 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4078 bug, fixed in subsequent PCRE releases).
4080 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4081 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4083 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4089 SC/01 Use a glob alias rather than an array ref in eximstats generated
4090 parser. This improves both readability and performance.
4092 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4093 Don't display local sender or destination tables in eximstats unless
4094 there is data to show.
4095 Added average volumes into the eximstats top table text output.
4097 SC/03 Collect data on the number of addresses (recipients) as well
4098 as the number of messages in eximstats.
4100 TF/01 Correct an error in the documentation for the redirect router. Exim
4101 does (usually) call initgroups() when daemonizing.
4103 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4104 with consistent privilege compared to when running as a daemon.
4106 TF/03 Note in the spec that $authenticated_id is not set for local
4107 submissions from trusted users.
4109 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4110 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4112 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4113 by adding some example configuration directives to the default
4114 configuration file. A little bit of work is required to uncomment the
4115 directives and define how usernames and passwords are checked, but
4116 there is now a framework to start from.
4118 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4119 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4120 without this. I don't know how relevant this is to other LDAP libraries.
4122 PH/02 Add the verb name to the "unknown ACL verb" error.
4124 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4126 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4128 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4129 directory not expanded when it should be if an expanded home directory
4130 was set for the address (which is overridden by the transport).
4132 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4135 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4136 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4137 because it is too late at that time, and has no effect.
4139 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4140 security issue with \' (bugzilla #107). I could not use the
4141 PQescapeStringConn() function, because it needs a PGconn value as one of
4144 PH/08 When testing addresses using -bt, indicate those final addresses that
4145 are duplicates that would not cause an additional delivery. At least one
4146 person was confused, thinking that -bt output corresponded to deliveries.
4147 (Suppressing duplicates isn't a good idea as you lose the information
4148 about possibly different redirections that led to the duplicates.)
4150 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4151 systems where poll() doesn't work, in particular OS X.
4153 PH/10 Added more information to debugging output for retry time not reached.
4155 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4156 operations in malware.c.
4158 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4161 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4162 syslog_duplication set false, log lines that would normally be written
4163 both the the main log and to the reject log were not written to syslog at
4166 PH/14 In the default configuration, change the use of "message" in ACL warn
4167 statements to "add_header".
4169 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4170 not followed by a command (e.g. "seen endif").
4172 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4173 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4176 PH/17 Added extra conditions to the default value of delay_warning_condition
4180 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4181 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4182 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4185 The Auto-Submitted: and various List- headers are standardised, whereas I
4186 don't think Precedence: ever was.
4188 PH/18 Refactored debugging code in route_finduser() to show more information,
4189 in particular, the error code if getpwnam() issues one.
4191 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4192 This is apparently needed in addition to the PH/07 change above to avoid
4193 any possible encoding problems.
4195 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4196 but not after initializing Perl.
4198 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4199 output them only if debugging. By default they are written stderr,
4200 apparently, which is not desirable.
4202 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4205 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4208 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4210 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4211 authenticated or an ident call has been made. Suppress the default
4212 values for $authenticated_id and $authenticated_sender (but permit -oMai
4213 and -oMas) when testing with -bh.
4215 PH/24 Re-jigged the order of the tests in the default configuration so that the
4216 tests for valid domains and recipients precede the DNS black list and CSA
4217 tests, on the grounds that those ones are more expensive.
4219 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4220 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4221 This bug exists in every version of Exim that I still have, right back to
4224 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4225 However, an attempt to turn on case-sensitivity in a regex key by
4226 including (?-i) didn't work because the subject string was already
4227 lowercased, and the effects were non-intuitive. It turns out that a
4228 one-line patch can be used to allow (?-i) to work as expected.
4234 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4235 other effects) broke the use of negated acl sub-conditions.
4237 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4240 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4241 "Deny" causes Exim to reject the incoming connection with a 554 error.
4242 Unfortunately, if there is a major crisis, such as a disk failure,
4243 tcp-wrappers gives "deny", whereas what one would like would be some
4244 kind of temporary error. A kludge has been added to help with this.
4245 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4246 554 error is used if errno is still zero or contains ENOENT (which occurs
4247 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4250 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4252 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4253 errors. Otherwise a message that provokes a temporary error (when other
4254 messages do not) can cause a whole host to time out.
4256 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4257 addresses were routed directly to files or pipes from a redirect router.
4258 File deliveries just didn't batch; pipe deliveries might have suffered
4261 PH/06 A failure to get a lock for a hints database would erroneously always say
4262 "Failed to get write lock", even when it was really a read lock.
4264 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4265 of 0600. This has been changed to use the value of the lockfile_mode
4266 option (which defaults to 0600).
4268 PH/08 Applied small patch from the Sieve maintainer.
4270 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4271 folder from quota calculations, a direct delivery into this folder messed
4272 up the contents of the maildirsize file. This was because the regex was
4273 used only to exclude .Trash (or whatever) when the size of the mailbox
4274 was calculated. There was no check that a delivery was happening into an
4275 excluded directory. This bug has been fixed by ignoring all quota
4276 processing for deliveries into excluded directories.
4278 PH/10 Added the maildirfolder_create_regex option to appendfile.
4284 PH/01 The code for finding all the local interface addresses on a FreeBSD
4285 system running IPv6 was broken. This may well have applied to all BSD
4286 systems, as well as to others that have similar system calls. The broken
4287 code found IPv4 interfaces correctly, but gave incorrect values for the
4288 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4289 that it would not match correctly against @[] and not recognize the IPv6
4292 PH/02 The ipliteral router was not recognizing addresses of the form user@
4293 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4295 PH/03 Added disable_ipv6.
4297 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4298 decoded form, because it is most often used to construct To: headers
4299 lines in autoreplies, and the decoded form may well be syntactically
4300 invalid. However, $reply_address has leading white space removed, and all
4301 newlines turned into spaces so that the autoreply transport does not
4304 PH/05 If group was specified without a user on a router, and no group or user
4305 was specified on a transport, the group from the router was ignored.
4307 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4308 for visible compile-time settings that can be used to change these
4309 numbers, for those that want even more. Backwards compatibility with old
4310 spool files has been maintained. However, going back to a previous Exim
4311 release will lost any variables that are in spool files.
4313 PH/07 Two small changes when running in the test harness: increase delay when
4314 passing a TCP/IP connection to a new process, in case the original
4315 process has to generate a bounce, and remove special handling of
4316 127.0.0.2 (sic), which is no longer necessary.
4318 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4319 be the same on different OS.
4321 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4324 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4325 whether --show-vars was specified or not
4327 JJ/02 exipick: Added support for new ACL variable spool format introduced
4330 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4331 syntactically invalid From: or Reply-to: line, and a filter used this to
4332 generate an autoreply, and therefore failed to obtain an address for the
4333 autoreply, Exim could try to deliver to a non-existent relative file
4334 name, causing unrelated and misleading errors. What now happens is that
4335 it logs this as a hard delivery error, but does not attempt to create a
4338 PH/11 The exinext utility has a -C option for testing purposes, but although
4339 the given file was scanned by exinext itself; it wasn't being passed on
4340 when Exim was called.
4342 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4343 an end-of-file indication when reading a command response.
4345 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4346 compiled. In many other places in Exim, IPv6 addresses are always
4347 recognized, so I have changed this. It also means that IPv4 domain
4348 literals of the form [IPV4:n.n.n.n] are now always recognized.
4350 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4351 used if the router is not running as root, for example, when verifying at
4352 ACL time, or when using -bh. The debugging output from this situation was
4353 non-existent - all you got was a failure to exec. I have made two
4356 (a) Failures to set uid/gid, the current directory, or a process leader
4357 in a subprocess such as that created by queryprogram now generate
4358 suitable debugging output when -d is set.
4360 (b) The queryprogram router detects when it is not running as root,
4361 outputs suitable debugging information if -d is set, and then runs
4362 the subprocess without attempting to change uid/gid.
4364 PH/15 Minor change to Makefile for building test_host (undocumented testing
4367 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4368 additional section of a DNS packet that returns MX or SRV records.
4369 Instead, it always explicitly searches for A/AAAA records. This avoids
4370 major problems that occur when a DNS server includes only records of one
4371 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4372 fixed another bug: if SRV records were looked up and the corresponding
4373 address records were *not* found in the additional section, the port
4374 values from the SRV records were lost.
4376 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4377 using the correct key (the original address) when searching the retry
4378 rules in order to find which one to use for generating the retry hint.
4380 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4381 adding the default one. Similarly, if it contains a Reply-To: header, the
4382 errors_reply_to option, if set, is not used.
4384 PH/19 When calculating a retry time, Exim used to measure the "time since
4385 failure" by looking at the "first failed" field in the retry record. Now
4386 it does not use this if it is later than than the arrival time of the
4387 message. Instead it uses the arrival time. This makes for better
4388 behaviour in cases where some deliveries succeed, thus re-setting the
4389 "first failed" field. An example is a quota failure for a huge message
4390 when small messages continue to be delivered. Without this change, the
4391 "time since failure" will always be short, possible causing more frequent
4392 delivery attempts for the huge message than are intended.
4393 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4395 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4396 $1, $2, $3) because the numerical variables can be reset during some
4397 expansion items (e.g. "match"), thereby losing the authentication data.
4399 PH/21 Make -bV show the size of off_t variables so that the test suite can
4400 decide whether to run tests for quotas > 2G.
4402 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4403 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4404 filecount value is greater than 2G or if a quota value is greater than 2G
4405 on a system where the size of off_t is not greater than 4, a panic error
4408 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4409 never match. The debug and -bh output now contains an explicit error
4410 message indicating a malformed IPv4 address or mask.
4412 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4413 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4414 PH/23 above applies.
4416 PH/25 Do not write to syslog when running in the test harness. The only
4417 occasion when this arises is a failure to open the main or panic logs
4418 (for which there is an explicit test).
4420 PH/26 Added the /no_tell option to "control=freeze".
4422 PH/27 If a host name lookup failed very early in a connection, for example, if
4423 the IP address matched host_lookup and the reverse lookup yielded a name
4424 that did not have a forward lookup, an error message of the form "no IP
4425 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4426 could be logged. Now it outputs the IP address instead of "NULL".
4428 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4429 allows the sender and the authenticated sender to be set when
4430 submitting a message from within Exim. Since child_open_exim() is
4431 documented for local_scan(), the new function should be too.
4433 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4434 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4435 results in an empty string is now treated as unset.
4437 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4439 PH/31 Added sender_verify_fail logging option.
4441 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4442 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4443 tidied the source and removed it altogether.
4445 PH/33 When a queue run was abandoned because the load average was too high, a
4446 log line was always written; now it is written only if the queue_run log
4447 selector is set. In addition, the log line for abandonment now contains
4448 information about the queue run such as the pid. This is always present
4449 in "start" and "stop" lines but was omitted from the "abandon" line.
4451 PH/34 Omit spaces between a header name and the colon in the error message that
4452 is given when verify = headers_syntax fails (if there are lots of them,
4453 the message gets confusing).
4455 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4456 names, as there are now some PTR records that contain slashes. This check
4457 is only to protect against broken name servers that fall over on strange
4458 characters, so the fact that it applies to all lookups doesn't matter.
4460 PH/36 Now that the new test suite is complete, we can remove some of the
4461 special code in Exim that was needed for the old test suite. For example,
4462 sorting DNS records because real resolvers return them in an arbitrary
4463 order. The new test suite's fake resolver always returns records in the
4466 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4467 bounces) except when queue_only is set, to avoid logging races between
4468 the different processes.
4470 PH/38 Panic-die if .include specifies a non-absolute path.
4472 PH/39 A tweak to the "H" retry rule from its user.
4474 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4475 a label. They prevented compilation on older perls.
4477 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4478 a warning to be raised on newish perls.
4480 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4481 on queue. Changes to match documented behaviour of showing count of
4482 messages matching specified criteria.
4484 PH/40 Changed the default ident timeout from 30s to 5s.
4486 PH/41 Added support for the use of login_cap features, on those BSD systems
4487 that have them, for controlling the resources used by pipe deliveries.
4489 PH/42 The content-scanning code uses fopen() to create files in which to put
4490 message data. Previously it was not paying any attention to the mode of
4491 the files. Exim runs with umask(0) because the rest of the code creates
4492 files with open(), and sets the required mode explicitly. Thus, these
4493 files were ending up world-writeable. This was not a big issue, because,
4494 being within the spool directory, they were not world-accessible. I have
4495 created a function called modefopen, which takes an additional mode
4496 argument. It sets umask(777), creates the file, chmods it to the required
4497 mode, then resets the umask. All the relevant calls to fopen() in the
4498 content scanning code have been changed to use this function.
4500 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4501 to 24 hours. This avoids potential overflow problems when processing G
4502 and H retry rules. I suspect nobody ever tinkers with this value.
4504 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4506 PH/45 When the plaintext authenticator is running as a client, the server's
4507 challenges are checked to ensure they are valid base64 strings. By
4508 default, the authentication attempt is cancelled if an invalid string is
4509 received. Setting client_ignore_invalid_base64 true ignores these errors.
4510 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4511 they are received. Thus, the responses can be made to depend on the
4512 challenges. If an invalid string is ignored, an empty string is placed in
4515 PH/46 Messages that are created by the autoreply transport now contains a
4516 References: header, in accordance with RFCs 2822 and 3834.
4518 PH/47 Added authenticated_sender_force to the smtp transport.
4520 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4522 PH/49 Installed latest patch from the Sieve maintainer.
4524 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4525 was also set, the appendfile transport was unnecessarily scanning a
4526 directory of message files (e.g. for maildir delivery) to find the count
4527 of files (along with the size), even though it did not need this
4528 information. It now does the scan only if it needs to find either the
4529 size of the count of files.
4531 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4533 PH/52 Two bugs concerned with error handling when the smtp transport is
4536 (i) Exim was not creating retry information for temporary errors given
4537 for individual recipients after the DATA command when the smtp transport
4538 was used in LMTP mode. This meant that they could be retried too
4539 frequently, and not timed out correctly.
4541 (ii) Exim was setting the flag that allows error details to be returned
4542 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4543 recipients that were returned after the DATA command.
4545 PH/53 This is related to PH/52, but is more general: for any failing address,
4546 when detailed error information was permitted to be returned to the
4547 sender, but the error was temporary, then after the final timeout, only
4548 "retry timeout exceeded" was returned. Now it returns the full error as
4549 well as "retry timeout exceeded".
4551 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4552 do this, and (what is worse) MTAs that accept it.
4554 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4555 will now be deprecated.
4557 PH/56 New os.c-cygwin from the Cygwin maintainer.
4559 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4560 formats (previously only available in exim formats via -bpr, -bpru,
4561 and -bpra. Now also available in native and exiqgrep formats)
4563 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4564 with very large, slow to parse queues
4566 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4568 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4570 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4571 responses to authentication challenges, though it was showing the
4572 challenges; (ii) I've removed the CR characters from the debug output for
4575 PH/58 Allow for the insertion of a newline as well as a space when a string
4576 is turned into more than one encoded-word during RFC 2047 encoding. The
4577 Sieve code now uses this.
4579 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4580 data_4xx, lost_connection, tls_required.
4582 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4583 message was being sent as an SMTP response.
4585 PH/61 Add -l and -k options to exicyclog.
4587 PH/62 When verifying, if an address was redirected to one new address, so that
4588 verification continued, and the new address failed or deferred after
4589 having set something in $address_data, the value of $address_data was not
4590 passed back to the ACL. This was different to the case when no
4591 redirection occurred. The value is now passed back in both cases.
4593 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4594 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4595 use login_cap.h, so on its own it isn't the distinguishing feature. The
4596 new name refers directly to the setclassresources() function.
4598 PH/65 Added configuration files for NetBSD3.
4600 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4602 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4603 is preferred over IPv4.
4605 PH/68 The bounce_return_message and bounce_return_body options were not being
4606 honoured for bounces generated during the reception of non-SMTP messages.
4607 In particular, this applied to messages rejected by the ACL. This bug has
4608 been fixed. However, if bounce_return_message is true and bounce_return_
4609 body is false, the headers that are returned for a non-SMTP message
4610 include only those that have been read before the error was detected.
4611 (In the case of an ACL rejection, they have all been read.)
4613 PH/69 The HTML version of the specification is now built in a directory called
4614 spec_html instead of spec.html, because the latter looks like a path with
4615 a MIME-type, and this confuses some software.
4617 PH/70 Catch two compiler warnings in sieve.c.
4619 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4620 function verify_get_ident() calls ip_connect() to connect a socket, but
4621 if the "connect()" function timed out, ip_connect() used to close the
4622 socket. However, verify_get_ident() also closes the socket later, and in
4623 between Exim writes to the log, which may get opened at this point. When
4624 the socket was closed in ip_connect(), the log could get the same file
4625 descriptor number as the socket. This naturally causes chaos. The fix is
4626 not to close the socket in ip_connect(); the socket should be closed by
4627 the function that creates it. There was only one place in the code where
4628 this was missing, in the iplookup router, which I don't think anybody now
4629 uses, but I've fixed it anyway.
4631 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4632 well as to direct DNS lookups. Otherwise the handling of names in host
4633 lists is inconsistent and therefore confusing.
4639 PH/01 Two changes to the default runtime configuration:
4641 (1) Move the checks for relay_from_hosts and authenticated clients from
4642 after to before the (commented out) DNS black list checks.
4644 (2) Add control=submission to the relay_from_hosts and authenticated
4645 clients checks, on the grounds that messages accepted by these
4646 statements are most likely to be submissions.
4648 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4650 (1) Generate an error if the third argument for the ${prvs expansion is
4653 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4656 (3) Reset the variables that are obtained from the first argument of
4657 ${prvscheck and used in the second argument before leaving the code,
4658 because their memory is reclaimed, so using them afterwards may do
4661 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4662 one (it's much easier than Tom thought :-).
4664 (5) Because of (4), we can now allow for the use of $prvscheck_result
4665 inside the third argument.
4667 PH/03 For some reason, the default setting of PATH when running a command from
4668 a pipe transport was just "/usr/bin". I have changed it to
4671 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4672 anything to be listed in the output from -bV.
4674 PH/05 When a filter generated an autoreply, the entire To: header line was
4675 quoted in the delivery log line, like this:
4677 => >A.N.Other <ano@some.domain> <original@ddress> ...
4679 This has been changed so that it extracts the operative address. There
4680 may be more than one such address. If so, they are comma-separated, like
4683 => >ano@some.domain,ona@other.domain <original@ddress> ...
4685 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4686 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4687 not being looked up in the rDNS to get a host name, Exim was showing the
4688 IP address twice in Received: lines, even though the IP addresses were
4689 identical. For example:
4691 Received: from [1.2.3.4] (helo=[1.2.3.4])
4693 However, if the real host name was known, it was omitting the HELO data
4694 if it matched the actual IP address. This has been tidied up so that it
4695 doesn't show the same IP address twice.
4697 PH/07 When both +timestamp and +memory debugging was on, the value given by
4698 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4699 called by the debug printing, thereby overwriting the timestamp buffer.
4700 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4702 PH/08 When the original message was included in an autoreply transport, it
4703 always said "this is a copy of the message, including all the headers",
4704 even if body_only or headers_only was set. It now gives an appropriate
4707 PH/09 Applied a patch from the Sieve maintainer which:
4709 o fixes some comments
4710 o adds the (disabled) notify extension core
4711 o adds some debug output for the result of if/elsif tests
4712 o points to the current vacation draft in the documentation
4713 and documents the missing references header update
4717 o fixes a bug in processing the envelope test (when testing
4718 multiple envelope elements, the last element determined the
4721 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4722 Electronic Mail") by including:
4724 Auto-submitted: auto-generated
4726 in the messages that it generates (bounce messages and others, such as
4727 warnings). In the case of bounce messages for non-SMTP messages, there was
4728 also a typo: it was using "Auto_submitted" (underscore instead of
4729 hyphen). Since every message generated by Exim is necessarily in response
4730 to another message, thes have all been changed to:
4732 Auto-Submitted: auto-replied
4734 in accordance with these statements in the RFC:
4736 The auto-replied keyword:
4738 - SHOULD be used on messages sent in direct response to another
4739 message by an automatic process,
4741 - MUST NOT be used on manually-generated messages,
4743 - MAY be used on Delivery Status Notifications (DSNs) and Message
4744 Disposition Notifications (MDNs),
4746 - MUST NOT be used on messages generated by automatic or periodic
4747 processes, except for messages which are automatic responses to
4750 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4751 to the default Received: header definition.
4753 PH/12 Added log selector acl_warn_skipped (default on).
4755 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4756 variables because (a) they are in the wrong storage pool and (b) even if
4757 they were copied, it wouldn't work properly because of the caching.
4759 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4760 checking when decoding. Apparently there are clients that generate
4761 overlong encoded strings. Why am I not surprised?
4763 PH/15 If the first argument of "${if match_address" was not empty, but did not
4764 contain an "@" character, Exim crashed. Now it writes a panic log message
4765 and treats the condition as false.
4767 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4769 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4770 extension "envelope-auth". The code is finished and in agreement with
4771 other implementations, but there is no documentation so far and in fact,
4772 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4773 not changing the active code.
4775 Print executed "if" and "elsif" statements when debugging is used. This
4776 helps a great deal to understand what a filter does.
4778 Document more things not specified clearly in RFC3028. I had all this
4779 sorted out, when out of a sudden new issues came to my mind. Oops."
4781 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4784 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4785 canonical form (as documented). However, after a host name lookup from
4786 the IP address, check_host() was doing a simple string comparison with
4787 addresses acquired from the DNS when checking that the found name did
4788 have the original IP as one of its addresses. Since any found IPv6
4789 addresses are likely to be in abbreviated form, the comparison could
4790 fail. Luckily, there already exists a function for doing the comparison
4791 by converting both addresses to binary, so now that is used instead of
4792 the text comparison.
4794 PH/20 There was another similar case to PH/19, when a complete host name was
4795 given in a host list; looking up its IP address could give an abbreviated
4796 form, whereas the current host's name might or might not be abbreviated.
4797 The same fix has been applied.
4803 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4804 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4807 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4808 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4810 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4812 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4813 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4814 Also, for "Auto-Submitted", treat anything other than "no" as
4815 non-personal, in accordance with RFC 3834. (Previously it treated
4816 anything starting "auto-" as non-personal.)
4818 TF/01 The control=submission/name=... option had a problem with syntax
4819 errors if the name included a slash character. The /name= option
4820 now slurps the rest of the string, so it can include any characters
4821 but it must come last in the list of options (after /sender_retain
4824 PH/05 Some modifications to the interface to the fake nameserver for the new
4832 TK/01 Added the "success_on_redirect" address verification option. See
4833 NewStuff for rationale and an example.
4835 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4837 PH/02 Patch to exigrep to allow it to work on syslog lines.
4839 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4840 fread() to skip over the body file's header line, because in Cygwin the
4841 header line is locked and is inaccessible.
4843 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4844 co-exist for some time) to make it clear that it is the Exim ID that is
4845 referenced, not the Message-ID: header line.
4847 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4848 string_format() function, because snprintf() does not exist on all
4851 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4852 ${stat: expansion item.
4854 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4855 protocol synchronization error", to keep the pedants happy.
4857 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4858 well as for IRIX systems, when gcc is being used. See the host.c source
4861 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4863 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4866 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4867 transport and to the smtp transport in LMTP mode.
4869 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4871 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4872 run a filter in a subprocess. This could lead to confusion in subsequent
4873 lookups in the parent process. There should also be a search_tidyup() at
4874 the end of the subprocess.
4876 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4877 only if the host matched helo_try_verify_hosts, which caused the
4878 verification to occur when the EHLO/HELO command was issued. The ACL just
4879 tested the remembered result. Now, if a previous verification attempt has
4880 not happened, "verify = helo" does it there and then.
4882 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4884 TK/03 Fix log output including CR from clamd.
4886 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4887 exist provoked a memory error which could cause a segfault.
4889 PH/15 Installed PCRE 6.2
4891 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4893 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4894 of the problem. Specifically, suggested +O2 rather than +O1 for the
4897 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4899 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4900 record, the port was not correctly logged when the outgoing_port log
4901 selector was set (it logged the transort's default port).
4903 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4904 fallback_hosts, and "hosts" in the smtp transport.
4906 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4907 host errors such as "Connection refused".
4909 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4910 authentication with radiusclient 0.4.9:
4912 - Error returned from rc_read_config was caught wrongly
4913 - Username/password not passed on to radius server due to wrong length.
4915 The presumption is that some radiusclient API changes for 4.51/PH/17
4916 were not taken care of correctly. The code is still untested by me (my
4917 Linux distribution still has 0.3.2 of radiusclient), but it was
4918 contributed by a Radius user.
4920 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4921 expanding the "port" option of the smtp transport.
4923 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4924 while reading a MIME header. Thanks to Tom Hughes for a patch.
4926 PH/24 Include config.h inside local_scan.h so that configuration settings are
4929 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4930 that in an ACL for RCPT (for example), you can examine exactly what was
4933 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4934 commands, but it was not correctly comparing the address with the actual
4935 client host address. Thus, it would show the EHLO address in Received:
4936 header lines when this was not necessary.
4938 PH/27 Added the % operator to ${eval:}.
4940 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4941 it should be ignoring failures (because with -C, for example, it has lost
4942 privilege). It wasn't ignoring creation failures other than "already
4945 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4948 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4949 because an input line was too long, either on its own, or by virtue of
4950 too many continuations, the temporary file was not being removed, and the
4951 return code was incorrect.
4953 PH/31 Missing "BOOL" in function definition in filtertest.c.
4955 PH/32 Applied Sieve patches from the maintainer.
4957 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4959 PH/33 Added "verify = not_blind".
4961 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4962 Local/Makefile (with some defaults set). These are used in built scripts
4963 such as exicyclog, but they have never been used in the exim_install
4964 script (though there are many overriding facilities there). I have
4965 arranged that the exim_install script now takes note of these two
4968 PH/35 Installed configuration files for Dragonfly.
4970 PH/36 When a locally submitted message by a trusted user did not contain a
4971 From: header, and the sender address was obtained from -f or from an SMTP
4972 MAIL command, and the trusted user did not use -F to supply a sender
4973 name, $originator_name was incorrectly used when constructing a From:
4974 header. Furthermore, $originator_name was used for submission mode
4975 messages from external hosts without From: headers in a similar way,
4976 which is clearly wrong.
4978 PH/37 Added control=suppress_local_fixups.
4980 PH/38 When log_selector = +received_sender was set, and the addition of the
4981 sender made the log line's construction buffer exactly full, or one byte
4982 less than full, an overflow happened when the terminating "\n" was
4985 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4986 when the result of a list match is failure because a DNS lookup failed.
4988 PH/40 RM_COMMAND is now used in the building process.
4990 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4991 the "build-* directories that it finds.
4993 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4994 address in a domain literal was a prefix of an interface address.
4996 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4997 when verifying a sender address, unless rewrite_headers is false.
4999 PH/44 Wrote a long comment about why errors_to addresses are verified as
5000 recipients, not senders.
5002 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5003 the ratelimit ACL was added.
5005 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5007 PH/46 Added extra information about PostgreSQL errors to the error string.
5009 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5010 avoiding the need to install special zones in a real server. This is
5011 backwards compatible; if it can't find the fake resolver, it drops back.
5012 Thus, both old and new test suites can be run.
5014 TF/02 Added util/ratelimit.pl
5016 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5017 clock is set back in time.
5019 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5020 Candler <B.Candler@pobox.com>.
5022 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5023 OK for addresses that are the result of redirections.
5025 PH/48 A number of further additions for the benefit of the new test suite,
5026 including a fake gethostbyname() that interfaces to the fake DNS resolver
5029 TF/06 The fix for widen_domains has also been applied to qualify_single and
5030 search_parents which are the other dnslookup options that can cause
5033 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5036 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5038 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5039 service" if it encountered an MX record with an empty target hostname.
5040 The message is now "an MX or SRV record indicated no SMTP service".
5042 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5043 if the DNS of the sending site is misconfigured. This is quite a
5044 common situation. This change restores the behaviour of treating a
5045 helo verification defer as a failure.
5047 PH/51 If self=fail was set on a router, the bounce message did not include the
5048 actual error message.
5054 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5056 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5057 command itself ended in error, the underlying message about the transport
5058 filter timeout was being overwritten with the pipe command error. Now the
5059 underlying error message should be appended to the second error message.
5061 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5063 PH/02 Reduce the amount of output that "make" produces by default. Full output
5064 can still be requested.
5066 PH/03 The warning log line about a condition test deferring for a "warn" verb
5067 was being output only once per connection, rather than after each
5068 occurrence (because it was using the same function as for successful
5069 "warn" verbs). This seems wrong, so I have changed it.
5071 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5072 it should not have, which might have caused a crash in the right
5073 circumstances, but probably never did.
5075 PH/04 Installed a modified version of Tony Finch's patch to make submission
5076 mode fix the return path as well as the Sender: header line, and to
5077 add a /name= option so that you can make the user's friendly name appear
5080 TF/03 Added the control = fakedefer ACL modifier.
5082 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5083 Mark Lowes for thorough testing.
5085 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5087 TK/03 Merged latest SRS patch from Miles Wilton.
5089 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5090 with the definition in sysexits.h (which is #included earlier).
5091 Fortunately, Exim does not actually use EX_OK. The code used to try to
5092 preserve the sysexits.h value, by assuming that macro definitions were
5093 scanned for macro replacements. I have been disabused of this notion,
5094 so now the code just undefines EX_OK before #including unistd.h.
5096 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5097 in the smtp transport. When a block could not be written in a single
5098 write() function, the timeout was being re-applied to each part-write.
5099 This seems wrong - if the receiver was accepting one byte at a time it
5100 would take for ever. The timeout is now adjusted when this happens. It
5101 doesn't have to be particularly precise.
5103 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5104 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5106 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5107 without a domain if the check to <postmaster@domain> fails.
5109 SC/01 Eximstats: added -xls and the ability to specify output files
5110 (patch written by Frank Heydlauf).
5112 SC/02 Eximstats: use FileHandles for outputting results.
5114 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5116 SC/04 Eximstats: fixed display of large numbers with -nvr option
5118 SC/05 Eximstats: fixed merging of reports with empty tables.
5120 SC/06 Eximstats: added the -include_original_destination flag
5122 SC/07 Eximstats: removed tabs and trailing whitespace.
5124 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5126 TK/06 MBOX spool code: Add real "From " MBOX separator line
5127 so the .eml file is really in mbox format (even though
5128 most programs do not really care). Patch from Alex Miller.
5130 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5131 The latter is generated from $received_to and is only set if the
5132 message has one envelope recipient. SA can use these headers,
5133 obviously out-of-the-box. Patch from Alex Miller.
5135 PH/08 The ${def test on a variable was returning false if the variable's
5136 value was "0", contrary to what the specification has always said!
5137 The result should be true unless the variable is empty.
5139 PH/09 The syntax error of a character other than { following "${if
5140 def:variable_name" (after optional whitespace) was not being diagnosed.
5141 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5142 accidental colon was present, for example, could give incorrect results.
5144 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5145 result is used (not including appendfile, where other changes are about
5148 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5149 This involved changing a lot of size variables from int to off_t. It
5150 should work with maildirs and everything.
5152 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5153 spamd dying while we are connected to it.
5155 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5158 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5159 function for BDB 4.3.
5161 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5163 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5164 the sources for PCRE have been re-arranged and more files are now
5167 PH/15 The code I had for printing potentially long long variables in PH/11
5168 above was not the best (it lost precision). The length of off_t variables
5169 is now inspected at build time, and an appropriate printing format (%ld
5170 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5171 to be "long long int" or "long int". This is needed for the internal
5172 formatting function string_vformat().
5174 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5175 the configuration file to be ":syslog", then the script "guesses" where
5176 the logs files are, rather than using the compiled in default. In our
5177 case the guess is not the same as the compiled default, so the script
5178 suddenly stopped working when I started to use syslog. The patch checks
5179 to see if log_file_path is "". If so, it attempts to read it from exim
5180 with no configuration file to get the compiled in version, before it
5181 falls back to the previous guessing code."
5183 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5184 implementing BATV in an Exim configuration. See NewStuff for the gory
5187 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5188 Makefile that are specific to HP-UX.
5190 PH/18 If the "use_postmaster" option was set for a recipient callout together
5191 with the "random" option, the postmaster address was used as the MAIL
5192 FROM address for the random test, but not for the subsequent recipient
5193 test. It is now used for both.
5195 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5196 patch removes a few documentation additions to RFC 3028, because the
5197 latest draft now contains them. It adds the new en;ascii-case comparator
5198 and a new error check for 8bit text in MIME parts. Comparator and
5199 require names are now matched exactly. I enabled the subaddress
5200 extension, but it is not well tested yet (read: it works for me)."
5202 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5203 rework some of the code of TK/09 above to avoid the hardwired use of
5204 "%lld" and "long long". Replaced the call to snprintf() with a call to
5207 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5208 records point to non-existent hosts", "retry timeout exceeded", and
5209 "retry time not reached for any host after a long failure period".
5211 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5212 experimental DomainKeys support:
5214 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5215 (2) On an error such as an illegally used "control", the wrong name for
5216 the control was given.
5218 These problems did NOT occur unless DomainKeys support was compiled.
5220 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5222 PH/24 Added ${if match_ip condition.
5224 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5225 databases so that it will be absolutely obvious if a crash occurs in the
5226 DB library. This is a regular occurrence (often caused by mis-matched
5229 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5230 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5231 picked up on a user's system that detects such things. There doesn't seem
5232 to be a gcc warning option for this - only an attribute that has to be
5233 put on the function's prototype. It seems that in Fedora Core 4 they have
5234 set this on a number of new functions. No doubt there will be more in due
5237 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5238 specify a transport. However, if an address that was verified by such a
5239 router was the subject of a callout, Exim crashed because it tried to
5240 read the rcpt_include_affixes from the non-existent transport. Now it
5241 just assumes that the setting of that option is false. This bug was
5242 introduced by 4.51/PH/31.
5244 PH/28 Changed -d+all to exclude +memory, because that information is very
5245 rarely of interest, but it makes the output a lot bigger. People tend to
5246 do -d+all out of habit.
5248 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5249 code in os-type was giving problems when libc.so lives in lib64, like on
5252 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5253 aren't the modern standard, and it seems that some systems' include files
5254 don't always have them. Exim was already checking for some of the newer
5255 ones like T_AAAA, and defining it itself. I've added checks for all the
5256 record types that Exim uses.
5258 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5259 not automatically generating a new one, as it is supposed to. This
5260 prevented TLS from working. If the file did exist, but contained invalid
5261 data, a new version was generated, as expected. It was only the case of a
5262 non-existent file that was broken.
5264 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5265 with a change in libdomainkeys > 0.64.
5267 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5268 from DNS. If the selector record carries the flag, it now has
5269 precedence over the domain-wide flag.
5271 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5273 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5274 the use of a port name that isn't defined in /etc/services) occurred, the
5275 message was deferred as in a normal delivery, and thus remained on the
5276 spool, instead of being failed because of the mua_wrapper setting. This
5277 is now fixed, and I tidied up some of the mua_wrapper messages at the
5280 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5281 the messages in a hash of arrays rather than using individual hashes.
5282 This is a bit cleaner and results in dramatic memory savings, albeit
5283 at a slight CPU cost.
5285 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5286 as requested by Marc Sherman.
5288 SC/10 Eximstats: added histograms for user specified patterns as requested
5291 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5293 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5294 fopen() in the content-scanning modules that did not already have it.
5300 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5301 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5303 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5305 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5307 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5308 to test Sieve filters that use "vacation".
5310 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5311 that changes the way the GnuTLS parameters are stored in the cache file.
5312 The new format can be generated externally. For backward compatibility,
5313 if the data in the cache doesn't make sense, Exim assumes it has read an
5314 old-format file, and it generates new data and writes a new file. This
5315 means that you can't go back to an older release without removing the
5318 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5319 work if there are any delivery delays because "one_time" forces the
5320 parent to be marked "delivered", so its unseen clone is never tried
5321 again. For this reason, Exim now forbids the simultaneous setting of
5324 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5325 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5326 entry if you want to know the details. The fix, however, neglected to
5327 consider the case when local delivery batching is involved. The test for
5328 "previously delivered" was not happening when checking to see if an
5329 address could be batched with a previous (undelivered) one; under
5330 certain circumstances this could lead to multiple deliveries to the same
5333 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5334 in its include files, and this causes problems building Exim.
5336 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5337 header_syntax) but Exim was just ignoring anything given after a slash.
5338 In particular, this caused confusion with an attempt to use "verify =
5339 reverse_host_lookup/defer_ok". An error is now given when options are
5340 supplied for verify items that do not have them. (Maybe reverse_host_
5341 lookup should have a defer_ok option, but that's a different point.)
5343 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5344 defined by RFC 821) to 2048, because there were problems with some AUTH
5345 commands, and RFC 1869 says the size should be increased for extended
5346 SMTP commands that take arguments.
5348 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5351 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5352 "unknown" error; now it says that the functionality isn't in the binary.
5354 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5355 an address' error message when a string expansion fails (syntax or
5356 whatever). Otherwise the password may appear in the log. Following change
5357 PH/42 below, there is no longer a chance of it appearing in a bounce
5360 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5362 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5363 of its addresses was ever tried. (Bugzilla bug #2.)
5365 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5366 the result incorrectly in the debug output. (It correctly added a newline
5367 to what was transported.)
5369 TF/01 Added $received_time.
5371 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5372 commented out examples of how to interface to a virus scanner and to
5373 SpamAssassin. Also added commented examples of av_scanner and
5374 spamd_address settings.
5376 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5377 and controls are allowed in which ACLs. There were a couple of minor
5378 errors. Some of the entries in the conditions table (which is a table of
5379 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5380 negation of where the condition IS allowed.
5382 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5384 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5385 header file does not have a version number, so I've had to invent a new
5386 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5387 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5388 radiusclient), but it was contributed by a Radius user.
5390 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5391 files or directories, for OpenSSL.
5393 PH/19 When an Exim process that is running as root has to create an Exim log
5394 file, it does so in a subprocess that runs as exim:exim so as to get the
5395 ownership right at creation (otherwise, other Exim processes might see
5396 the file with the wrong ownership). There was no test for failure of this
5397 fork() call, which would lead to the process getting stuck as it waited
5398 for a non-existent subprocess. Forks do occasionally fail when resources
5399 run out. I reviewed all the other calls to fork(); they all seem to check
5402 PH/20 When checking for unexpected SMTP input at connect time (before writing
5403 the banner), Exim was not dealing correctly with a non-positive return
5404 from the read() function. If the client had disconnected by this time,
5405 the result was a log entry for a synchronization error with an empty
5406 string after "input=" when read() returned zero. If read() returned -1
5407 (an event I could not check), uninitialized data bytes were printed.
5408 There were reports of junk text (parts of files, etc) appearing after
5411 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5413 PH/22 Added support for macro redefinition, and (re)definition in between
5414 driver and ACL definitions.
5416 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5417 forgetting to use the resulting value; it was using the unexpanded value.
5419 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5420 hadn't been configured. The fix is from Juergen Kreileder, who
5421 understands it better than I do:
5423 "Here's what I see happening with three configured cyrus_sasl
5424 authenticators configured (plain, login, cram-md5):
5426 On startup auth_cyrus_sasl_init() gets called for each of these.
5427 This means three calls to sasl_listmech() without a specified mech_list.
5428 => SASL tests which mechs of all available mechs actually work
5429 => three warnings about OTP not working
5430 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5432 With the patch, sasl_listmech() also gets called three times. But now
5433 SASL's mech_list option is set to the server_mech specified in the the
5434 authenticator. Or in other words, the answer from sasl_listmech()
5435 gets limited to just the mech you're testing for (which is different
5437 => the return list contains just 'plain' or 'login', 'cram-md5' or
5438 nothing depending on the value of ob->server_mech.
5440 I've just tested the patch: Authentication still works fine,
5441 unavailable mechs specified in the exim configuration are still
5442 caught, and the auth.log warnings about OTP are gone."
5444 PH/25 When debugging is enabled, the contents of the command line are added
5445 to the debugging output, even when log_selector=+arguments is not
5448 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5449 answer is "GNU", and only if the return is "GNU/something" is the answer
5452 PH/27 $acl_verify_message is now set immediately after the failure of a
5453 verification in an ACL, and so is available in subsequent modifiers. In
5454 particular, the message can be preserved by coding like this:
5456 warn !verify = sender
5457 set acl_m0 = $acl_verify_message
5459 Previously, $acl_verify_message was set only while expanding "message"
5460 and "log_message" when a very denied access.
5462 PH/28 Modified OS/os.c-Linux with
5464 -#ifndef OS_LOAD_AVERAGE
5465 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5467 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5468 nomenclature these days.)
5470 PH/29 Installed patch from the Sieve maintainer that adds the options
5471 sieve_useraddress and sieve_subaddress to the redirect router.
5473 PH/30 In these circumstances:
5474 . Two addresses routed to the same list of hosts;
5475 . First host does not offer TLS;
5476 . First host accepts first address;
5477 . First host gives temporary error to second address;
5478 . Second host offers TLS and a TLS session is established;
5479 . Second host accepts second address.
5480 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5481 and peerdn, if requested) that were in fact used only for the second
5484 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5485 attention to any local part prefix or suffix that was matched by the
5486 router that accepted the address. It now behaves in the same way as it
5487 does for delivery: the affixes are removed from the local part unless
5488 rcpt_include_affixes is set on the transport.
5490 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5491 timeout during the DATA phase of an incoming message.
5493 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5494 applied a patch sanctioned by the Sieve maintainer.
5496 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5497 the uid or gid is negative. A case of a negative gid caused this to be
5498 noticed. The fix allows for either to be negative.
5500 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5501 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5504 PH/36 The change PH/12 above was broken. Fixed it.
5506 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5507 the grounds that routing the same address twice would always produce the
5508 same answer. This might have been true once, but it is certainly no
5509 longer true now. Routing a child address may depend on the previous
5510 routing that produced that child. Some complicated redirection strategies
5511 went wrong when messages had multiple recipients, and made Exim's
5512 behaviour dependent on the order in which the addresses were given.
5514 I have moved the duplicate checking until after the routing is complete.
5515 Exim scans the addresses that are assigned to local and remote
5516 transports, and removes any duplicates. This means that more work will be
5517 done, as duplicates will always all be routed, but duplicates are
5518 presumably rare, so I don't expect this is of any significance.
5520 For deliveries to pipes, files, and autoreplies, the duplicate checking
5521 still happens during the routing process, since they are not going to be
5524 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5525 It corrects a timeout issue with spamd. This is Ian's comment: "The
5526 background is that sometimes spamd either never reads data from a
5527 connection it has accepted, or it never writes response data. The exiscan
5528 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5529 blindly assumes that writes won't block so it may never time out."
5531 PH/39 Allow G after quota size as well as K and M.
5533 PH/40 The value set for $authenticated_id in an authenticator may not contain
5534 binary zeroes or newlines because the value is written to log lines and
5535 to spool files. There was no check on this. Now the value is run through
5536 the string_printing() function so that such characters are converted to
5537 printable escape sequences.
5539 PH/41 $message_linecount is a new variable that contains the total number of
5540 lines in the message. Compare $body_linecount, which is the count for the
5543 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5544 bounce and delay warning messages, except in certain special cases, which
5547 (a) An SMTP error message from a remote host;
5548 (b) A message specified in a :fail: redirection;
5549 (c) A message specified in a "fail" command in a system filter;
5550 (d) A message specified in a FAIL return from the queryprogram router;
5551 (e) A message specified by the cannot_route_message router option.
5553 In these cases only, Exim does include the error details in bounce and
5554 warning messages. There are also a few cases where bland messages such
5555 as "unrouteable address" or "local delivery error" are given.
5557 PH/43 $value is now also set for the "else" part of a ${run expansion.
5559 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5560 being worked on, but at least Exim now implements the latest version to
5563 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5564 process to complete was treated as a delivery failure, a timeout while
5565 writing the message to the pipe was logged, but erroneously treated as a
5566 successful delivery. Such timeouts include transport filter timeouts. For
5567 consistency with the overall process timeout, these timeouts are now
5568 treated as errors, giving rise to delivery failures by default. However,
5569 there is now a new Boolean option for the pipe transport called
5570 timeout_defer, which, if set TRUE, converts the failures into defers for
5571 both kinds of timeout. A transport filter timeout is now identified in
5574 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5575 systems where "make" and "gmake" are different, calling "gmake" at top
5576 level broke things. I've arranged for the value of $(MAKE) to be passed
5577 from the Makefile to this script so that it can call the same version of
5581 A note about Exim versions 4.44 and 4.50
5582 ----------------------------------------
5584 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5585 changes of various kinds. As a consequence, a big documentation update was
5586 needed. This delayed the release for rather longer than seemed good, especially
5587 in the light of a couple of (minor) security issues. Therefore, the changes
5588 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5589 release. So 4.44 and 4.50 are in effect two different branches that both start
5592 I have left the 4.50 change log unchanged; it contains all the changes since
5593 4.43. The change log for 4.44 is below; many of its items are identical to
5594 those for 4.50. This seems to be the most sensible way to preserve the
5595 historical information.
5601 1. Minor wording change to the doc/README.SIEVE file.
5603 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5604 computation of the current number of files was incorrect.
5606 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5607 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5610 4. Give more explanation in the error message when the command for a transport
5611 filter fails to execute.
5613 5. There are several places where Exim runs a non-Exim command in a
5614 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5615 was being done only for the command run by the queryprogram router. It is
5616 now done for all such subprocesses. The other cases are: ${run, transport
5617 filters, and the commands run by the lmtp and pipe transports.
5619 6. Added CONFIGURE_GROUP build-time option.
5621 7. Some older OS have a limit of 256 on the maximum number of file
5622 descriptors. Exim was using setrlimit() to set 1000 as a large value
5623 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5624 systems. I've change it so that if it can't get 1000, it tries for 256.
5626 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5627 was an oversight, and furthermore, ever since the addition of extra
5628 controls (e.g. 4.43/32), the checks on when to allow different forms of
5629 "control" were broken. There should now be diagnostics for all cases when a
5630 control that does not make sense is encountered.
5632 9. Added the /retain_sender option to "control=submission".
5634 10. $recipients is now available in the predata ACL (oversight).
5636 11. Tidy the search cache before the fork to do a delivery from a message
5637 received from the command line. Otherwise the child will trigger a lookup
5638 failure and thereby defer the delivery if it tries to use (for example) a
5639 cached ldap connection that the parent has called unbind on.
5641 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5642 of $address_data from the recipient verification was clobbered by the
5643 sender verification.
5645 13. The value of address_data from a sender verification is now available in
5646 $sender_address_data in subsequent conditions in the ACL statement.
5648 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5650 15. Added a new option "connect=<time>" to callout options, to set a different
5653 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5654 was its contents. (It was OK if the option was not defined at all.)
5656 17. A "Completed" log line is now written for messages that are removed from
5657 the spool by the -Mrm option.
5659 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5660 information about exactly what failed.
5662 19. Added -dd to debug only the daemon process.
5664 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5665 handles timeouts, both on the server side and network timeouts. Renamed the
5666 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5668 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5669 It is now set to "smtps".
5671 22. $host_address is now set to the target address during the checking of
5672 ignore_target_hosts.
5674 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5675 being passed; this would have caused $sender_host_name to have been used if
5676 matching the list had actually called for a host name (not very likely,
5677 since this list is usually IP addresses). A host name is now passed as
5680 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5681 code that specifies a non-restarting handler (typically sigaction() in
5682 modern systems) in an attempt to fix a rare and obscure crash bug.
5684 25. Narrowed the window for a race in the daemon that could cause it to ignore
5685 SIGCHLD signals. This is not a major problem, because they are used only to
5686 wake it up if nothing else does.
5688 26. A malformed maildirsize file could cause Exim to calculate negative values
5689 for the mailbox size or file count. Odd effects could occur as a result.
5690 The maildirsize information is now recalculated if the size or filecount
5693 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5694 support for a long time. Removed HAVE_SYS_VFS_H.
5696 28. Installed the latest version of exipick from John Jetmore.
5698 29. In an address list, if the pattern was not a regular expression, an empty
5699 subject address (from a bounce message) matched only if the pattern was an
5700 empty string. Non-empty patterns were not even tested. This was the wrong
5701 because it is perfectly reasonable to use an empty address as part of a
5702 database query. An empty address is now tested by patterns that are
5703 lookups. However, all the other forms of pattern expect the subject to
5704 contain a local part and a domain, and therefore, for them, an empty
5705 address still always fails if the pattern is not itself empty.
5707 30. Exim went into a mad DNS loop when attempting to do a callout where the
5708 host was specified on an smtp transport, and looking it up yielded more
5709 than one IP address.
5711 31. Re-factored the code for checking spool and log partition space into a
5712 function that finds that data and another that does the check. The former
5713 is then used to implement four new variables: $spool_space, $log_space,
5714 $spool_inodes, and $log_inodes.
5716 32. The RFC2047 encoding function was originally intended for short strings
5717 such as real names; it was not keeping to the 75-character limit for
5718 encoded words that the RFC imposes. It now respects the limit, and
5719 generates multiple encoded words if necessary. To be on the safe side, I
5720 have increased the buffer size for the ${rfc2047: expansion operator from
5723 33. It is now permitted to omit both strings after an "if" condition; if the
5724 condition is true, the result is "true". As before, when the second string
5725 is omitted, a false condition yields an empty string. This makes it less
5726 cumbersome to write custom ACL and router conditions.
5728 34. Failure to deliver a bounce message always caused it to be frozen, even if
5729 there was an errors_to setting on the router. The errors_to setting is now
5732 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5733 canonical form (fully expanded) before being placed in
5734 $sender_host_address.
5736 36. The table in the code that translates DNS record types into text (T_A to
5737 "A" for instance) was missing entries for NS and CNAME. It is just possible
5738 that this could have caused confusion if both these types were looked up
5739 for the same domain, because the text type is used as part of Exim's
5740 per-process caching. But the chance of anyone hitting this buglet seems
5743 37. The dnsdb lookup has been extended in a number of ways.
5745 (1) There is a new type, "zns", which walks up the domain tree until it
5746 finds some nameserver records. It should be used with care.
5748 (2) There is a new type, "mxh", which is like "mx" except that it returns
5749 just the host names, not the priorities.
5751 (3) It is now possible to give a list of domains (or IP addresses) to be
5752 looked up. The behaviour when one of the lookups defers can be
5753 controlled by a keyword.
5755 (4) It is now possible to specify the separator character for use when
5756 multiple records are returned.
5758 38. The dnslists ACL condition has been extended: it is now possible to supply
5759 a list of IP addresses and/or domains to be looked up in a particular DNS
5762 39. Added log_selector=+queue_time_overall.
5764 40. When running the queue in the test harness, wait just a tad after forking a
5765 delivery process, to get repeatability of debugging output.
5767 41. Include certificate and key file names in error message when GnuTLS fails
5768 to set them up, because the GnuTLS error message doesn't include the name
5769 of the failing file when there is a problem reading it.
5771 42. Allow both -bf and -bF in the same test run.
5773 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5775 44. The "Exiscan patch" is now merged into the mainline Exim source.
5777 45. Sometimes the final signoff response after QUIT could fail to get
5778 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5779 < 0 before doing a fflush(). This bug looks as though it goes back to the
5780 introduction of TLS in release 3.20, but "sometimes" must have been rare
5781 because the tests only now provoked it.
5783 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5784 (this can affect the format of dates).
5786 47. exim_tidydb, when checking for the continued existence of a message for
5787 which it has found a message-specific retry record, was not finding
5788 messages that were in split spool directories. Consequently, it was
5789 deleting retry records that should have stayed in existence.
5791 48. Steve fixed some bugs in eximstats.
5793 49. The SPA authentication driver was not abandoning authentication and moving
5794 on to the next authenticator when an expansion was forced to fail,
5795 contradicting the general specification for all authenticators. Instead it
5796 was generating a temporary error. It now behaves as specified.
5798 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5799 (the order specifies the preference for clients). The order is now AES256,
5800 AES128, 3DES, ARCFOUR128.
5802 51. Small patch to Sieve code - explicitly set From: when generating an
5805 52. Exim crashed if a remote delivery caused a very long error message to be
5806 recorded - for instance if somebody sent an entire SpamAssassin report back
5807 as a large number of 550 error lines. This bug was coincidentally fixed by
5808 increasing the size of one of Exim's internal buffers (big_buffer) that
5809 happened as part of the Exiscan merge. However, to be on the safe side, I
5810 have made the code more robust (and fixed the comments that describe what
5813 53. Now that there can be additional text after "Completed" in log lines (if
5814 the queue_time_overall log selector is set), a one-byte patch to exigrep
5815 was needed to allow it to recognize "Completed" as not the last thing in
5818 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5819 patch that reportedly fixes this has been added. I am not expert enough to
5820 create a test for it. This is what the patch creator wrote:
5822 "I found a little strange behaviour of ldap code when working with
5823 Windows 2003 AD Domain, where users was placed in more than one
5824 Organization Units. When I tried to give exim partial DN, the exit code
5825 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5826 But simultaneously result of request was absolutely normal ldap result,
5827 so I produce this patch..."
5829 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5830 REFERENCE, so I have modified the code to exclude the patch when that macro
5833 55. Some experimental protocols are using DNS PTR records for new purposes. The
5834 keys for these records are domain names, not reversed IP addresses. The
5835 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5836 leaves it alone. Component reversal etc. now happens only for IP addresses.
5839 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5841 57. Double the size of the debug message buffer (to 2048) so that more of very
5842 long debug lines gets shown.
5844 58. The exicyclog utility now does better if the number of log files to keep
5845 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5847 59. Two changes related to the smtp_active_hostname option:
5849 (1) $smtp_active_hostname is now available as a variable.
5850 (2) The default for smtp_banner uses $smtp_active_hostname instead
5851 of $primary_hostname.
5853 60. The host_aton() function is supposed to be passed a string that is known
5854 to be a valid IP address. However, in the case of IPv6 addresses, it was
5855 not checking this. This is a hostage to fortune. Exim now panics and dies
5856 if the condition is not met. A case was found where this could be provoked
5857 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5858 components; fortuitously, this particular loophole had already been fixed
5859 by change 4.50/55 above.
5861 If there are any other similar loopholes, the new check in host_aton()
5862 itself should stop them being exploited. The report I received stated that
5863 data on the command line could provoke the exploit when Exim was running as
5864 exim, but did not say which command line option was involved. All I could
5865 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5866 running as the user.
5869 61. There was a buffer overflow vulnerability in the SPA authentication code
5870 (which came originally from the Samba project). I have added a test to the
5871 spa_base64_to_bits() function which I hope fixes it.
5874 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5875 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5877 63. The daemon start-up calls getloadavg() while still root for those OS that
5878 need the first call to be done as root, but it missed one case: when
5879 deliver_queue_load_max is set with deliver_drop_privilege. This is
5880 necessary for the benefit of the queue runner, because there is no re-exec
5881 when deliver_drop_privilege is set.
5883 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5884 This has been fixed.
5886 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5887 was in use, was not putting the data itself into the right store pool;
5888 consequently, it could be overwritten for a subsequent message in the same
5889 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5892 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5894 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5895 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5896 the function were treating the return as a boolean value, which happened to
5897 work because 0=false and not-0=true, but is not correct code.
5899 68. The host_aton() function was not handling scoped IPv6 addresses (those
5900 with, for example, "%eth0" on the end) correctly.
5902 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5903 negated items (that is, ~something) in unsigned ints. Some compilers
5904 apparently mutter when there is no cast.
5906 70. If an address verification called from an ACL failed, and did not produce a
5907 user-specific message (i.e. there was only a "system" message), nothing was
5908 put in $acl_verify_message. In this situation, it now puts the system
5911 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5912 session; change 4.31/43 added the unwanted input to the log line - except
5913 that it did not do this in the start of session case. It now does.
5915 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5916 This is wrong and can cause the other end to generate a synchronization
5917 error if it is another Exim or anything else that does the synchronization
5918 check. A QUIT command is no longer sent after a timeout.
5920 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5921 during host lookups.
5923 74. The defer_ok option of callout verification was not working if it was used
5924 when verifying addresses in header lines, that is, for this case:
5926 verify = header_sender/callout=defer_ok
5928 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5929 those file descriptors could be used for SMTP connections. If anything
5930 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5931 could be sent to the SMTP client, causing chaos. The daemon now opens
5932 stdin, stdout, and stderr to /dev/null when it puts itself into the
5935 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5936 log by default. The user can override this with suitable Perl magic.
5938 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5939 the log message when discard triggers, was not working for the DATA ACL or
5940 for the non-SMTP ACL.
5942 78. Error message wording change in sieve.c.
5944 79. If smtp_accept_max_per_host was set, the number of connections could be
5945 restricted to fewer than expected, because the daemon was trying to set up
5946 a new connection before checking whether the processes handling previous
5947 connections had finished. The check for completed processes is now done
5948 earlier. On busy systems, this bug wouldn't be noticed because something
5949 else would have woken the daemon, and it would have reaped the completed
5952 80. If a message was submitted locally by a user whose login name contained one
5953 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5954 It caused a spool format error. I have fixed the spool reading code. A
5955 related problem was that the "from" clause in the Received: line became
5956 illegal because of the space(s). It is now covered by ${quote_local_part.
5958 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5961 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5963 83. Patch from Sieve maintainer to fix unterminated string problem in
5964 "vacation" handling.
5966 84. Some minor changes to the Linux configuration files to help with other
5967 OS variants using glibc.
5969 85. One more patch for Sieve to update vacation handling to latest spec.
5972 ----------------------------------------------------
5973 See the note above about the 4.44 and 4.50 releases.
5974 ----------------------------------------------------
5980 1. Change 4.43/35 introduced a bug that caused file counts to be
5981 incorrectly computed when quota_filecount was set in an appendfile
5984 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5985 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5988 3. Give more explanation in the error message when the command for a transport
5989 filter fails to execute.
5991 4. There are several places where Exim runs a non-Exim command in a
5992 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5993 was being done only for the command run by the queryprogram router. It is
5994 now done for all such subprocesses. The other cases are: ${run, transport
5995 filters, and the commands run by the lmtp and pipe transports.
5997 5. Some older OS have a limit of 256 on the maximum number of file
5998 descriptors. Exim was using setrlimit() to set 1000 as a large value
5999 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6000 systems. I've change it so that if it can't get 1000, it tries for 256.
6002 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6003 was an oversight, and furthermore, ever since the addition of extra
6004 controls (e.g. 4.43/32), the checks on when to allow different forms of
6005 "control" were broken. There should now be diagnostics for all cases when a
6006 control that does not make sense is encountered.
6008 7. $recipients is now available in the predata ACL (oversight).
6010 8. Tidy the search cache before the fork to do a delivery from a message
6011 received from the command line. Otherwise the child will trigger a lookup
6012 failure and thereby defer the delivery if it tries to use (for example) a
6013 cached ldap connection that the parent has called unbind on.
6015 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6016 of $address_data from the recipient verification was clobbered by the
6017 sender verification.
6019 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6020 was its contents. (It was OK if the option was not defined at all.)
6022 11. A "Completed" log line is now written for messages that are removed from
6023 the spool by the -Mrm option.
6025 12. $host_address is now set to the target address during the checking of
6026 ignore_target_hosts.
6028 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6029 being passed; this would have caused $sender_host_name to have been used if
6030 matching the list had actually called for a host name (not very likely,
6031 since this list is usually IP addresses). A host name is now passed as
6034 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6035 code that specifies a non-restarting handler (typically sigaction() in
6036 modern systems) in an attempt to fix a rare and obscure crash bug.
6038 15. Narrowed the window for a race in the daemon that could cause it to ignore
6039 SIGCHLD signals. This is not a major problem, because they are used only to
6040 wake it up if nothing else does.
6042 16. A malformed maildirsize file could cause Exim to calculate negative values
6043 for the mailbox size or file count. Odd effects could occur as a result.
6044 The maildirsize information is now recalculated if the size or filecount
6047 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6048 support for a long time. Removed HAVE_SYS_VFS_H.
6050 18. Updated exipick to current release from John Jetmore.
6052 19. Allow an empty sender to be matched against a lookup in an address list.
6053 Previously the only cases considered were a regular expression, or an
6056 20. Exim went into a mad DNS lookup loop when doing a callout where the
6057 host was specified on the transport, if the DNS lookup yielded more than
6060 21. The RFC2047 encoding function was originally intended for short strings
6061 such as real names; it was not keeping to the 75-character limit for
6062 encoded words that the RFC imposes. It now respects the limit, and
6063 generates multiple encoded words if necessary. To be on the safe side, I
6064 have increased the buffer size for the ${rfc2047: expansion operator from
6067 22. Failure to deliver a bounce message always caused it to be frozen, even if
6068 there was an errors_to setting on the router. The errors_to setting is now
6071 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6072 canonical form (fully expanded) before being placed in
6073 $sender_host_address.
6075 24. Updated eximstats to version 1.33
6077 25. Include certificate and key file names in error message when GnuTLS fails
6078 to set them up, because the GnuTLS error message doesn't include the name
6079 of the failing file when there is a problem reading it.
6081 26. Expand error message when OpenSSL has problems setting up cert/key files.
6084 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6085 (this can affect the format of dates).
6087 28. exim_tidydb, when checking for the continued existence of a message for
6088 which it has found a message-specific retry record, was not finding
6089 messages that were in split spool directories. Consequently, it was
6090 deleting retry records that should have stayed in existence.
6092 29. eximstats updated to version 1.35
6093 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6094 1.35 - bugfix such that pie charts by volume are generated correctly
6096 30. The SPA authentication driver was not abandoning authentication and moving
6097 on to the next authenticator when an expansion was forced to fail,
6098 contradicting the general specification for all authenticators. Instead it
6099 was generating a temporary error. It now behaves as specified.
6101 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6102 (the order specifies the preference for clients). The order is now AES256,
6103 AES128, 3DES, ARCFOUR128.
6105 31. Small patch to Sieve code - explicitly set From: when generating an
6108 32. Exim crashed if a remote delivery caused a very long error message to be
6109 recorded - for instance if somebody sent an entire SpamAssassin report back
6110 as a large number of 550 error lines. This bug was coincidentally fixed by
6111 increasing the size of one of Exim's internal buffers (big_buffer) that
6112 happened as part of the Exiscan merge. However, to be on the safe side, I
6113 have made the code more robust (and fixed the comments that describe what
6116 33. Some experimental protocols are using DNS PTR records for new purposes. The
6117 keys for these records are domain names, not reversed IP addresses. The
6118 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6119 leaves it alone. Component reversal etc. now happens only for IP addresses.
6122 34. The host_aton() function is supposed to be passed a string that is known
6123 to be a valid IP address. However, in the case of IPv6 addresses, it was
6124 not checking this. This is a hostage to fortune. Exim now panics and dies
6125 if the condition is not met. A case was found where this could be provoked
6126 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6127 components; fortuitously, this particular loophole had already been fixed
6128 by change 4.50/55 or 4.44/33 above.
6130 If there are any other similar loopholes, the new check in host_aton()
6131 itself should stop them being exploited. The report I received stated that
6132 data on the command line could provoke the exploit when Exim was running as
6133 exim, but did not say which command line option was involved. All I could
6134 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6135 running as the user.
6138 35. There was a buffer overflow vulnerability in the SPA authentication code
6139 (which came originally from the Samba project). I have added a test to the
6140 spa_base64_to_bits() function which I hope fixes it.
6143 36. The daemon start-up calls getloadavg() while still root for those OS that
6144 need the first call to be done as root, but it missed one case: when
6145 deliver_queue_load_max is set with deliver_drop_privilege. This is
6146 necessary for the benefit of the queue runner, because there is no re-exec
6147 when deliver_drop_privilege is set.
6149 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6150 was in use, was not putting the data itself into the right store pool;
6151 consequently, it could be overwritten for a subsequent message in the same
6152 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6155 38. Sometimes the final signoff response after QUIT could fail to get
6156 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6157 < 0 before doing a fflush(). This bug looks as though it goes back to the
6158 introduction of TLS in release 3.20, but "sometimes" must have been rare
6159 because the tests only now provoked it.
6165 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6166 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6167 Now it returns an integer. A number of calls were still expecting a T/F
6168 return. Fortuitously, in all cases, the tests worked in OK situations,
6169 which is the norm. However, things would have gone wrong on any write
6170 failures on the smtp file descriptor. This function is used when sending
6171 messages over SMTP and also when doing verify callouts.
6173 2. When Exim is called to do synchronous delivery of a locally submitted
6174 message (the -odf or -odi options), it no longer closes stderr before doing
6177 3. Implemented the mua_wrapper option.
6179 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6181 5. Implemented the functions header_remove(), header_testname(),
6182 header_add_at_position(), and receive_remove_recipient(), and exported them
6185 6. If an ACL "warn" statement specified the addition of headers, Exim already
6186 inserted X-ACL-Warn: at the start if there was no header name. However, it
6187 was not making this test for the second and subsequent header lines if
6188 there were newlines in the string. This meant that an invalid header could
6189 be inserted if Exim was badly configured.
6191 7. Allow an ACL "warn" statement to add header lines at the start or after all
6192 the Received: headers, as well as at the end.
6194 8. Added the rcpt_4xx retry error code.
6196 9. Added postmaster_mailfrom=xxx to callout verification option.
6198 10. Added mailfrom=xxxx to the callout verification option, for verify=
6201 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6202 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6204 12. Inserted some casts to stop certain compilers warning when using pointer
6205 differences as field lengths or precisions in printf-type calls (mostly
6206 affecting debugging statements).
6208 13. Added optional readline() support for -be (dynamically loaded).
6210 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6211 same clock tick as a message's arrival, so that its received time was the
6212 same as the "first fail" time on the retry record, and that message
6213 remained on the queue past the ultimate address timeout, every queue runner
6214 would try a delivery (because it was past the ultimate address timeout) but
6215 after another failure, the ultimate address timeout, which should have then
6216 bounced the address, did not kick in. This was a "< instead of <=" error;
6217 in most cases the first failure would have been in the next clock tick
6218 after the received time, and all would be well.
6220 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6221 being recognized when the domain list was tested by the match_domain
6222 condition in an expansion string.
6224 16. Added the ${str2b64: operator.
6226 17. Exim was always calling setrlimit() to set a large limit for the number of
6227 processes, without checking whether the existing limit was already
6228 adequate. (It did check for the limit on file descriptors.) Furthermore,
6229 errors from getrlimit() and setrlimit() were being ignored. Now they are
6230 logged to the main and panic logs, but Exim does carry on, to try to do its
6231 job under whatever limits there are.
6233 18. Imported PCRE 5.0.
6235 19. Trivial typo in log message " temporarily refused connection" (the leading
6238 20. If the log selector return_path_on_delivery was set and an address was
6239 redirected to /dev/null, the delivery process crashed because it assumed
6240 that a return path would always be set for a "successful" delivery. In this
6241 case, the whole delivery is bypassed as an optimization, and therefore no
6244 21. Internal re-arrangement: the function for sending a challenge and reading
6245 a response while authentication was assuming a zero-terminated challenge
6246 string. It's now changed to take a pointer and a length, to allow for
6247 binary data in such strings.
6249 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6251 23. Exim was not respecting finduser_retries when seeking the login of the
6252 uid under which it was called; it was always trying 10 times. (The default
6253 setting of finduser_retries is zero.) Also, it was sleeping after the final
6254 failure, which is pointless.
6256 24. Implemented tls_on_connect_ports.
6258 25. Implemented acl_smtp_predata.
6260 26. If the domain in control=submission is set empty, Exim assumes that the
6261 authenticated id is a complete email address when it generates From: or
6262 Sender: header lines.
6264 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6265 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6266 chown and chgrp in /bin and hostname in /usr/bin.
6268 28. Exim was keeping the "process log" file open after each use, just as it
6269 does for the main log. This opens the possibility of it remaining open for
6270 long periods when the USR1 signal hits a daemon. Occasional processlog
6271 errors were reported, that could have been caused by this. Anyway, it seems
6272 much more sensible not to leave this file open at all, so that is what now
6275 29. The long-running daemon process does not normally write to the log once it
6276 has entered its main loop, and it closes the log before doing so. This is
6277 so that log files can straightforwardly be renamed and moved. However,
6278 there are a couple of unusual error situations where the daemon does write
6279 log entries, and I had neglected to close the log afterwards.
6281 30. The text of an SMTP error response that was received during a remote
6282 delivery was being truncated at 512 bytes. This is too short for some of
6283 the long messages that one sometimes sees. I've increased the limit to
6286 31. It is now possible to make retry rules that apply only when a message has a
6287 specific sender, in particular, an empty sender.
6289 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6290 it possible to be selective about when SMTP synchronization is enforced.
6292 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6294 32. Implemented hosts_connection_nolog.
6296 33. Added an ACL for QUIT.
6298 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6301 35. Added mailbox_size and mailbox_filecount to appendfile.
6303 36. Added control = no_multiline_responses to ACLs.
6305 37. There was a bug in the logic of the code that waits for the clock to tick
6306 in the case where the clock went backwards by a substantial amount such
6307 that the microsecond fraction of "now" was more than the microsecond
6308 fraction of "then" (but the whole seconds number was less).
6310 38. Added support for the libradius Radius client library this is found on
6311 FreeBSD (previously only the radiusclient library was supported).
6317 1. When certain lookups returned multiple values in the form name=value, the
6318 quoting of the values was not always being done properly. Specifically:
6319 (a) If the value started with a double quote, but contained no whitespace,
6321 (b) If the value contained whitespace other than a space character (i.e.
6322 tabs or newlines or carriage returns) it was not quoted.
6323 This fix has been applied to the mysql and pgsql lookups by writing a
6324 separate quoting function and calling it from the lookup code. The fix
6325 should probably also be applied to nisplus, ibase and oracle lookups, but
6326 since I cannot test any of those, I have not disturbed their existing code.
6328 2. A hit in the callout cache for a specific address caused a log line with no
6329 reason for rejecting RCPT. Now it says "Previous (cached) callout
6330 verification failure".
6332 3. There was an off-by-one bug in the queryprogram router. An over-long
6333 return line was truncated at 256 instead of 255 characters, thereby
6334 overflowing its buffer with the terminating zero. As well as fixing this, I
6335 have increased the buffer size to 1024 (and made a note to document this).
6337 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6338 when Exim is waiting for an SMTP response from a remote server, Exim
6339 restarts its select() call on the socket, thereby resetting its timeout.
6340 This is not a problem when such interrupts are rare. Somebody set up a cron
6341 job to run exiwhat every 2 minutes, which is less than the normal select()
6342 timeout (5 or 10 minutes). This meant that the select() timeout never
6343 kicked in because it was always reset. I have fixed this by comparing the
6344 time when an interrupt arrives with the time at the start of the first call
6345 to select(). If more time than the timeout has elapsed, the interrupt is
6346 treated as a timeout.
6348 5. Some internal re-factoring in preparation for the addition of Sieve
6349 extensions (by MH). In particular, the "personal" test is moved to a
6350 separate function, and given an option for scanning Cc: and Bcc: (which is
6351 not set for Exim filters).
6353 6. When Exim created an email address using the login of the caller as the
6354 local part (e.g. when creating a From: or Sender: header line), it was not
6355 quoting the local part when it contained special characters such as @.
6357 7. Installed new OpenBSD configuration files.
6359 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6360 try to make them clearer.
6362 9. Callout options, other than the timeout value, were being ignored when
6363 verifying sender addresses in header lines. For example, when using
6365 verify = header_sender/callout=no_cache
6367 the cache was (incorrectly) being used.
6369 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6370 only the Cygwin environment.
6372 11. When return_path_on_delivery was set as a log selector, if different remote
6373 addresses in the same message used different return paths and parallel
6374 remote delivery occurred, the wrong values would sometimes be logged.
6375 (Whenever a remote delivery process finished, the return path value from
6376 the most recently started remote delivery process was logged.)
6378 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6379 lines when AUTH and/or TLS are in use. This is the "received protocol"
6380 field. Exim used to use "asmtp" for authenticated SMTP, without any
6381 indication (in the protocol name) for TLS use. Now it follows the RFC and
6382 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6383 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6384 names appear in log lines as well as in Received: header lines.
6386 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6387 extensions, and comparison tests, and to fix some bugs.
6389 14. Changes to the "personal" filter test:
6391 (1) The test was buggy in that it was just doing the equivalent of
6392 "contains" tests on header lines. For example, if a user's address was
6393 anne@some.where, the "personal" test would incorrectly be true for
6395 To: susanne@some.where
6397 This test is now done by extracting each address from the header in turn,
6398 and checking the entire address. Other tests that are part of "personal"
6399 are now done using regular expressions (for example, to check local parts
6400 of addresses in From: header lines).
6402 (2) The list of non-personal local parts in From: addresses has been
6403 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6404 taken from the Sieve specification recommendations.
6406 (3) If the message contains any header line starting with "List-" it is
6407 treated as non-personal.
6409 (4) The test for "circular" in the Subject: header line has been removed
6410 because it now seems ill-conceived.
6412 15. Minor typos in src/EDITME comments corrected.
6414 16. Installed latest exipick from John Jetmore.
6416 17. If headers_add on a router specified a text string that was too long for
6417 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6418 of string_sprintf() is now avoided.
6420 18. $message_body_size was not set (it was always zero) when running the DATA
6421 ACL and the local_scan() function.
6423 19. For the "mail" command in an Exim filter, no default was being set for
6424 the once_repeat time, causing a random time value to be used if "once" was
6425 specified. (If the value happened to be <= 0, no repeat happened.) The
6426 default is now 0s, meaning "never repeat". The "vacation" command was OK
6427 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6428 (I found it when inspecting the code).
6430 20. There is now an overall timeout for performing a callout verification. It
6431 defaults to 4 times the callout timeout, which applies to individual SMTP
6432 commands during the callout. The overall timeout applies when there is more
6433 than one host that can be tried. The timeout is checked before trying the
6434 next host. This prevents very long delays if there are a large number of
6435 hosts and all are timing out (e.g. when the network connections are timing
6436 out). The value of the overall timeout can be changed by specifying an
6437 additional sub-option for "callout", called "maxwait". For example:
6439 verify = sender/callout=5s,maxwait=20s
6441 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6442 to the end before writing, but this should make it even safer).
6444 22. Exim was forgetting that it had advertised PIPELINING for the second and
6445 subsequent messages on an SMTP connection. It was also not resetting its
6446 memory on STARTTLS and an internal HELO.
6448 23. When Exim logs an SMTP synchronization error within a session, it now
6449 records whether PIPELINING has been advertised or not.
6451 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6452 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6453 rather than long int.
6455 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6457 26. Added the never_mail option to autoreply.
6463 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6464 crash if the getsockname() call failed; this can happen if a connection is
6465 closed very soon after it is established. The problem was simply in the
6466 order in which certain operations were done, causing Exim to try to write
6467 to the SMTP stream before it had set up the file descriptor. The bug has
6468 been fixed by making things happen in the correct order.
6474 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6475 before the connection was closed, thus losing the rejection response.
6477 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6478 some early Solaris releases, but causes trouble in current releases where
6479 socklen_t is defined.
6481 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6484 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6487 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6488 that releasing the top of it at the end releases what was used for sub-
6489 expansions (unless the block got too big). However, discard this block if
6490 the first thing is a variable or header, so that we can use its block when
6491 it is dynamic (useful for very large $message_headers, for example).
6493 6. Lookups now cache *every* query, not just the most recent. A new, separate
6494 store pool is used for this. It can be recovered when all lookup caches are
6495 flushed. Lookups now release memory at the end of their result strings.
6496 This has involved some general refactoring of the lookup sources.
6498 7. Some code has been added to the store_xxx() functions to reduce the amount
6499 of flapping under certain conditions.
6501 8. log_incoming_interface used to affect only the <= reception log lines. Now
6502 it causes the local interface and port to be added to several more SMTP log
6503 lines, for example "SMTP connection from", and rejection lines.
6505 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6507 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6509 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6510 could be overwritten at the end of the current message (or the start of a
6511 new message if it was set in a HELO ACL). The value is now preserved for
6512 the duration of the SMTP connection.
6514 12. If a transport had a headers_rewrite setting, and a matching header line
6515 contained an unqualified address, that address was qualified, even if it
6516 did not match any rewriting rules. The underlying bug was that the values
6517 of the flags that permit the existence of unqualified sender and recipient
6518 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6519 non-local messages, and by -bnq for local messages) were not being
6520 preserved with the message after it was received.
6522 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6523 "next input=" as part of the text comprising the host identity instead of
6524 the correct text. The code was using the same buffer for two different
6525 strings. However, depending on which order the printing function evaluated
6526 its arguments, the bug did not always show up. Under Linux, for example, my
6527 test suite worked just fine.
6529 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6530 doesn't work with some older versions of Perl. It has been changed to "my",
6531 which in any case is probably the better facility to use.
6533 15. A really picky compiler found some instances of statements for creating
6534 error messages that either had too many or two few arguments for the format
6537 16. The size of the buffer for calls to the DNS resolver has been increased
6538 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6539 for addresses that have a lot of PTR records. This alleviates a problem; it
6540 does not fully solve it.
6542 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6543 buffer now truncates the list and logs the incident, which is the same
6544 action as happens when Exim is looking up a host name and its aliases.
6545 Previously in this situation something unpredictable would happen;
6546 sometimes it was "internal error: store_reset failed".
6548 18. If a server dropped the connection unexpectedly when an Exim client was
6549 using GnuTLS and trying to read a response, the client delivery process
6550 crashed while trying to generate an error log message.
6552 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6553 string, for example:
6555 warn message = H1: something\nH2: something
6557 the text was added as a single header line from Exim's point of view
6558 though it ended up OK in the delivered message. However, searching for the
6559 second and subsequent header lines using $h_h2: did not work. This has been
6560 fixed. Similarly, if a system filter added multiple headers in this way,
6561 the routers could not see them.
6563 20. Expanded the error message when iplsearch is called with an invalid key to
6564 suggest using net-iplsearch in a host list.
6566 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6567 ACLs are no longer actually imposed (and a message to that effect is
6570 22. If a "gecos" field in a passwd entry contained escaped characters, in
6571 particular, if it contained a \" sequence, Exim got it wrong when building
6572 a From: or a Sender: header from that name. A second bug also caused
6573 incorrect handling when an unquoted " was present following a character
6574 that needed quoting.
6576 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6577 was not being matched caselessly.
6579 24. Arranged for all hyphens in the exim.8 source to be escaped with
6582 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6583 properly. Recipient callout cache records were still being keyed to include
6584 the sender, even when use_sender was set false. This led to far more
6585 callouts that were necessary. The sender is no longer included in the key
6586 when use_sender is false.
6588 26. Added "control = submission" modifier to ACLs.
6590 27. Added the ${base62d: operator to decode base 62 numbers.
6592 28. dnsdb lookups can now access SRV records.
6594 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6595 the configuration file.
6597 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6598 (-v) mode. This makes the output for a verbose queue run more intelligible.
6600 31. Added a use_postmaster feature to recipient callouts.
6602 32. Added the $body_zerocount variable, containing the number of binary zero
6603 bytes in the message body.
6605 33. The time of last modification of the "new" subdirectory is now used as the
6606 "mailbox time last read" when there is a quota error for a maildir
6609 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6611 35. Added +ignore_unknown as a special item in host lists.
6613 36. Code for decoding IPv6 addresses in host lists is now included, even if
6614 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6615 address was recognized as an IP address, but was then not correctly decoded
6616 into binary, causing unexpected and incorrect effects when compared with
6623 1. Very minor rewording of debugging text in manualroute to say "list of
6624 hosts" instead of "hostlist".
6626 2. If verify=header_syntax was set, and a header line with an unqualified
6627 address (no domain) and a large number of spaces between the end of the
6628 name and the colon was received, the reception process suffered a buffer
6629 overflow, and (when I tested it) crashed. This was caused by some obsolete
6630 code that should have been removed. The fix is to remove it!
6632 3. When running in the test harness, delay a bit after writing a bounce
6633 message to get a bit more predictability in the log output.
6635 4. Added a call to search_tidyup() just before forking a reception process. In
6636 theory, someone could use a lookup in the expansion of smtp_accept_max_
6637 per_host which, without the tidyup, could leave open a database connection.
6639 5. Added the variables $recipient_data and $sender_data which get set from a
6640 lookup success in an ACL "recipients" or "senders" condition, or a router
6641 "senders" option, similar to $domain_data and $local_part_data.
6643 6. Moved the writing of debug_print from before to after the "senders" test
6646 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6647 problems for message scanning, either using a data ACL, or using
6648 local_scan() because the Received: header was not generated till after they
6649 were called (in order to set the time as the time of reception completion).
6650 I have revised the way this works. The header is now generated after the
6651 body is received, but before the ACL or local_scan() are called. After they
6652 are run, the timestamp in the header is updated.
6658 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6659 before starting a queue runner without re-exec. This happened only when
6660 deliver_drop_privilege was set or when the Exim user was set to root. The
6661 effect of the bug was that timeouts during subsequent deliveries caused
6662 crashes instead of being properly handled. The handler is now left at its
6663 default (and expected) setting.
6665 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6666 message, again when deliver_drop_privilege is set or Exim is run as root.
6667 The bug described in (1) was not present in this case, but the tidying up
6668 of the other signals was missing. I have made the two cases consistent.
6670 3. The ignore_target_hosts setting on a manualroute router was being ignored
6671 for hosts that were looked up using the /MX notation.
6673 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6676 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6677 operated on the sender address. After changing the $sender_address to <>
6678 for the sender address verify, Exim was re-instated it as the original
6679 (before rewriting) address, but remembering that it had rewritten it, so it
6680 wasn't rewriting it again. This bug also had the effect of breaking the
6681 sender address verification caching when the sender address was rewritten.
6683 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6684 This has been changed so that if the ip literal address matches
6685 ignore_target_hosts, the router declines.
6687 7. Added expansion conditions match_domain, match_address, and match_local_
6688 part (NOT match_host).
6690 8. The placeholder for the Received: header didn't have a length field set.
6692 9. Added code to Exim itself and to exim_lock to test for a specific race
6693 condition that could lead to file corruption when using MBX delivery. The
6694 issue is with the lockfile that is created in /tmp. If this file is removed
6695 after a process has opened it but before that process has acquired a lock,
6696 there is the potential for a second process to recreate the file and also
6697 acquire a lock. This could lead to two Exim processes writing to the file
6698 at the same time. The added code performs the same test as UW imapd; it
6699 checks after acquiring the lock that its file descriptor still refers to
6700 the same named file.
6702 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6703 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6706 11. Added the smtp_active_hostname option. If used, this will typically be made
6707 to depend on the incoming interface address. Because $interface_address is
6708 not set up until the daemon has forked a reception process, error responses
6709 that can happen earlier (such as "too many connections") no longer contain
6712 12. If an expansion in a condition on a "warn" statement fails because a lookup
6713 defers, the "warn" statement is abandoned, and the next ACL statement is
6714 processed. Previously this caused the whole ACL to be aborted.
6716 13. Added the iplsearch lookup type.
6718 14. Added ident_timeout as a log selector.
6720 15. Added tls_certificate_verified as a log selector.
6722 16. Added a global option tls_require_ciphers (compare the smtp transport
6723 option of the same name). This controls incoming TLS connections.
6725 17. I finally figured out how to make tls_require_ciphers do a similar thing
6726 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6727 before starting the TLS session.
6729 18. Tabs are now shown as \t in -bP output.
6731 19. If the log selector return_path_on_delivery was set, Exim crashed when
6732 bouncing a message because it had too many Received: header lines.
6734 20. If two routers both had headers_remove settings, and the first one included
6735 a superfluous trailing colon, the final name in the first list and the
6736 first name in the second list were incorrectly joined into one item (with a
6737 colon in the middle).
6743 1. Added -C and -D options to the exinext utility, mainly to make it easier
6744 to include in the automated testing, but these could be helpful when
6745 multiple configurations are in use.
6747 2. The exinext utility was not formatting the output nicely when there was
6748 an alternate port involved in the retry record key, nor when there was a
6749 message id as well (for retries that were specific to a specific message
6750 and a specific host). It was also confused by IPv6 addresses, because of
6751 the additional colons they contain. I have fixed the IPv4 problem, and
6752 patched it up to do a reasonable job for IPv6.
6754 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6755 delivery, the log line now contains "pipelined" if PIPELINING was used.
6757 4. An SMTP transport process used to panic and die if the bind() call to set
6758 an explicit outgoing interface failed. This has been changed; it is now
6759 treated in the same way as a connect() failure.
6761 5. A reference to $sender_host_name in the part of a conditional expansion
6762 that was being skipped was still causing a DNS lookup. This no longer
6765 6. The def: expansion condition was not recognizing references to header lines
6766 that used bh_ and bheader_.
6768 7. Added the _cache feature to named lists.
6770 8. The code for checking quota_filecount in the appendfile transport was
6771 allowing one more file than it should have been.
6773 9. For compatibility with Sendmail, the command line option
6781 and sets the incoming protocol and host name (for trusted callers). The
6782 host name and its colon can be omitted when only the protocol is to be set.
6783 Note the Exim already has two private options, -pd and -ps, that refer to
6784 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6785 "s", but I don't think that's a major issue.
6787 10. A number of refactoring changes to the code, none of which should affect
6790 (a) The number of logging options was getting close to filling up the
6791 32-bit word that was used as a bit map. I have split them into two classes:
6792 those that are passed in the argument to log_write(), and those that are
6793 only ever tested independently outside of that function. These are now in
6794 separate 32-bit words, so there is plenty of room for expansion again.
6795 There is no change in the user interface or the logging behaviour.
6797 (b) When building, for example, log lines, the code previously used a
6798 macro that called string_cat() twice, in order to add two strings. This is
6799 not really sufficiently general. Furthermore, there was one instance where
6800 it was actually wrong because one of the argument was used twice, and in
6801 one call a function was used. (As it happened, calling the function twice
6802 did not affect the overall behaviour.) The macro has been replaced by a
6803 function that can join an arbitrary number of extra strings onto a growing
6806 (c) The code for expansion conditions now uses a table and a binary chop
6807 instead of a serial search (which was left over from when there were very
6808 few conditions). Also, it now recognizes conditions like "pam" even when
6809 the relevant support is not compiled in: a suitably worded error message is
6810 given if an attempt is made to use such a condition.
6812 11. Added ${time_interval:xxxxx}.
6814 12. A bug was causing one of the ddress fields not to be passed back correctly
6815 from remote delivery subprocesses. The field in question was not being
6816 subsequently used, so this caused to problems in practice.
6818 13. Added new log selectors queue_time and deliver_time.
6820 14. Might have fixed a bug in maildirsizefile handling that threw up
6821 "unexpected character" debug warnings, and recalculated the data
6822 unnecessarily. In any case, I expanded the warning message to give more
6825 15. Added the message "Restricted characters in address" to the statements in
6826 the default ACL that block characters like @ and % in local parts.
6828 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6829 Three changes have been made:
6831 (a) There was a serious bug; a negative response to MAIL caused the whole
6832 recipient domain to be cached as invalid, thereby blocking all messages
6833 to all local parts at the same domain, from all senders. This bug has
6834 been fixed. The domain is no longer cached after a negative response to
6835 MAIL if the sender used is not empty.
6837 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6840 (c) A new callout option, "use_sender" has been added for people who want
6841 the modified behaviour.
6847 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6850 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6851 indeed breaks things for older releases.
6853 3. Added additional logging to the case where there is a problem reading data
6854 from a filter that is running in a subprocess using a pipe, in order to
6855 try to track down a specific problem.
6857 4. Testing facility fudge: when running in the test harness and attempting
6858 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6859 getting "No route to host". Convert this to a timeout.
6861 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6864 6. Some OS don't have socklen_t but use size_t instead. This affects the
6865 fifth argument of getsockopt() amongst other things. This is now
6866 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6867 can be set for individual OS. I have set it for SunOS5, OSF1, and
6868 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6869 some earlier ones do not.
6871 7. Change 4.30/15 was not doing the test caselessly.
6873 8. The standard form for an IPv6 address literal was being rejected by address
6874 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6875 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6876 this, as well as the form without the "IPv6" on the front (but only when
6877 address literals are enabled, of course).
6879 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6881 10. Exim crashed if a message with an empty sender address specified by -f
6882 encountered a router with an errors_to setting. This could be provoked only
6883 by a command such as
6887 where an empty string was supplied; "<>" did not hit this bug.
6889 11. Installed PCRE release 4.5.
6891 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6892 remained set. It is now erased.
6894 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6895 times from message ids (which are base 36 rather than the normal 62).
6897 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6898 were being counted as actual protocol errors, and logged if the log
6899 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6900 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6901 and DATA following a set of rejected RCPTs do not count as protocol errors.
6902 In other words, Exim assumes they were pipelined, though this may not
6903 actually be the case. Of course, in all cases the client gets an
6904 appropriate error code.
6906 15. If a lookup fails in an ACL condition, a message about the failure may
6907 be available; it is used if testing the ACL cannot continue, because most
6908 such messages specify what the cause of the deferral is. However, some
6909 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6910 that caused an old message to be retained and used if a later statement
6911 caused a defer, replacing the real cause of the deferral.
6913 16. If an IP address had so many PTR records that the DNS lookup buffer
6914 was not large enough to hold them, Exim could crash while trying to process
6915 the truncated data. It now detects and logs this case.
6917 17. Further to 4.21/58, another change has been made: if (and only if) the
6918 first line of a message (the first header line) ends with CRLF, a bare LF
6919 in a subsequent header line has a space inserted after it, so as not to
6920 terminate the header.
6922 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6923 unnecessarily, used atoi() instead of strtol(), and didn't check the
6924 termination when getting file sizes from file names by regex.
6926 19. Completely re-implemented the support for maildirsize files, in the light
6927 of a number of problems with the previous contributed implementation
6928 (4.30/29). In particular:
6930 . If the quota is zero, the maildirsize file is maintained, but no quota is
6933 . If the maildir directory does not exist, it is created before any attempt
6934 to write a maildirsize file.
6936 . The quota value in the file is just a cache; if the quota is changed in
6937 the transport, the new value overrides.
6939 . A regular expression is available for excluding directories from the
6942 20. The autoreply transport checks the characters in options that define the
6943 message's headers; it allows continued headers, but it was checking with
6944 isspace() after an embedded newline instead of explicitly looking for a
6947 21. If all the "regular" hosts to which an address was routed had passed their
6948 expiry times, and had not reached their retry times, the address was
6949 bounced, even if fallback hosts were defined. Now Exim should go on to try
6952 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6953 equivalent code in the SMTP transport. Some hosts send humungous responses
6954 to HELO/EHLO, more than 1024 it seems.
6956 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6957 gives compiler warnings in some environments. I've now done it "properly",
6960 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6961 (because of problems with the built-in one) was declared to return uschar *
6962 instead of char *, causing compiler failure.
6964 25. Fixed a file descriptor leak when processing alias/forward files.
6966 26. Fixed a minor format string issue in dbfn.c.
6968 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6970 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6971 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6974 29. When a sender address is verified, it is cached, to save repeating the test
6975 when there is more than one recipient in a message. However, when the
6976 verification involves a callout, it is possible for different callout
6977 options to be set for different recipients. It is too complicated to keep
6978 track of this in the cache, so now Exim always runs a verification when a
6979 callout is required, relying on the callout cache for the optimization.
6980 The overhead is duplication of the address routing, but this should not be
6983 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6984 to be verified with callout=postmaster, and the main callout worked but the
6985 postmaster check failed, the verification correctly failed. However, if a
6986 subsequent RCPT command asked for sender verification *without* the
6987 postmaster check, incorrect caching caused this verification also to fail,
6990 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6991 it was not caching the DNS options (qualify_single, search_parents) that
6992 were used when the lookup failed. A subsequent lookup with different
6993 options therefore always gave the same answer, though there were cases
6994 where it should not have. (Example: a "domains = !$mx_any" option on a
6995 dnslookup router: the "domains" option is always processed without any
6996 widening, but the router might have qualify_single set.) Now Exim uses the
6997 cached value only when the same options are set.
6999 32. Added John Jetmore's "exipick" utility to the distribution.
7001 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7002 than a timeout (e.g. a certificate is required, and is not provided), an
7003 Exim server now closes the connection immediately. Previously it waited for
7004 the client to close - but if the client is SSL, it seems that they each
7005 wait for each other, leading to a delay before one of them times out.
7007 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7008 maintained 0.8.x compatibility because I don't think many are using it, and
7009 it is clearly obsolete.
7011 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7014 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7015 client certificate was expired. A simple patch fixes this, though I don't
7016 understand the full logic of why the verify callback is called multiple
7019 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7020 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7021 which causes problems with some clients (such as the Certicom SSL Plus
7022 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7023 disables the coutermeasure allowing Eudora to connect."
7025 38. Exim was not checking that a write() to a log file succeeded. This could
7026 lead to Bad Things if a log got too big, in particular if it hit a file
7027 size limit. Exim now panics and dies if it cannot write to a log file, just
7028 as it does if it cannot open a log file.
7030 39. Modified OS/Makefile-Linux so that it now contains
7032 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7034 The two -D definitions ensure that Exim is compiled with large file
7035 support, which makes it possible to handle log files that are bigger than
7038 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7039 instance) a domain was checked against a named list that involved a lookup,
7040 causing $domain_data to be set, then another domain was checked against the
7041 same list, then the first domain was re-checked, the value of $domain_data
7042 after the final check could be wrong. In particular, if the second check
7043 failed, it could be set empty. This bug probably also applied to
7046 41. The strip_trailing_dot option was not being applied to the address given
7047 with the -f command-line option.
7049 42. The code for reading a message's header from the spool was incrementing
7050 $received_count, but never initializing it. This meant that the value was
7051 incorrect (doubled) while delivering a message in the same process in which
7052 it was received. In the most common configuration of Exim, this never
7053 happens - a fresh exec is done - but it can happen when
7054 deliver_drop_privilege is set.
7056 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7057 it now includes up to 150 characters of the unexpected data in the log
7060 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7061 and building data strings. The size of both of these buffers was 10 000
7062 bytes - far larger than anybody would *ever* want, thought I. Needless to
7063 say, somebody hit the limit. I have increased the maximum line length to
7064 20 000 and the maximum data length of concatenated lines to 100 000. I have
7065 also fixed two bugs, because there was no checking on these buffers. Tsk,
7066 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7067 buffer is too small.
7069 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7070 lsearch lookups. Now it does.
7072 46. When parsing a route_list item in a manualroute router, a fixed-length
7073 buffer was used for the list of hosts. I made this 1024 bytes long,
7074 thinking that nobody would ever have a list of hosts that long. Wrong.
7075 Somebody had a whole pile of complicated expansion conditions, and the
7076 string was silently truncated, leading to an expansion error. It turns out
7077 that it is easier to change to an unlimited length (owing to other changes
7078 that have happened since this code was originally written) than to build
7079 structure for giving a limitation error. The length of the item that
7080 expands into the list of hosts is now unlimited.
7082 47. The lsearch lookup could not handle data where the length of text line was
7083 more than 4095 characters. Such lines were truncated, leading to shortened
7084 data being returned. It should now handle lines of any length.
7086 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7087 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7090 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7091 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7092 no longer happens in comment lines. A list of replacements is now placed
7093 at the head of all of the source files, except those whose only change is
7094 to replace PERL_COMMAND in the very first #! line.
7096 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7097 messages on the queue, with a bottom-up merge sort, using code contributed
7098 by Michael Haardt. This should make operations like -bp somewhat faster on
7099 large queues. It won't affect queue runners, except when queue_run_in_order
7102 51. Installed eximstats 1.31 in the distribution.
7104 52. Added support for SRV lookups to the dnslookup router.
7106 53. If an ACL referred to $message_body or $message_body_end, the value was not
7107 reset for any messages that followed in the same SMTP session.
7109 54. The store-handling optimization for building very long strings was not
7110 differentiating between the different store pools. I don't think this
7111 actually made any difference in practice, but I've tidied it.
7113 55. While running the routers to verify a sender address, $sender_address
7114 was still set to the sender address. This is wrong, because when routing to
7115 send a bounce to the sender, it would be empty. Therefore, I have changed
7116 it so that, while verifying a sender address, $sender_address is set to <>.
7117 (There is no change to what happens when verifying a recipient address.)
7119 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7120 target A or AAAA records (if not already returned) without resetting the
7121 qualify_single or search_parents options of the DNS resolver. These are
7122 inappropriate in this case because the targets of MX and SRV records must
7123 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7124 target that, when qualified, matched something in the local domain. These
7125 two options are now turned off when doing these lookups.
7127 57. It seems that at least some releases of Reiserfs (which does not have the
7128 concept of a fixed number of inodes) returns zero and not -1 for the
7129 number of available inodes. This interacted badly with check_spool_inodes,
7130 which assumed that -1 was the "no such thing" setting. What I have done is
7131 to check that the total number of inodes is greater than zero before doing
7132 the test of how many are available.
7134 58. When a "warn" ACL statement has a log_message modifier, the message is
7135 remembered, and not repeated. This is to avoid a lot of repetition when a
7136 message has many recipients that cause the same warning to be written.
7137 However, Exim was preserving the list of already written lines for an
7138 entire SMTP session, which doesn't seem right. The memory is now reset if a
7139 new message is started.
7141 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7142 debugging output unless log_rewrite was also set.
7144 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7145 of (int)(handle) when we know that handle contains (void *)(-1).
7147 61. The Exim daemon panic-logs an error return when it closes the incoming
7148 connection. However "connection reset by peer" seems to be common, and
7149 isn't really an error worthy of noting specially, so that particular error
7152 62. When Exim is trying to find all the local interfaces, it used to panic and
7153 die if the ioctl to get the interface flags failed. However, it seems that
7154 on at least one OS (Solaris 9) it is possible to have an interface that is
7155 included in the list of interfaces, but for which you get a failure error
7156 for this call. This happens when the interface is not "plumbed" into a
7157 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7158 failure of the "get flags" call assumes that the interface is down.
7160 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7161 makes life easier for people who are doing arithmetic on fields extracted
7162 from dates, where you often get leading zeros that should not be
7163 interpreted as octal.
7165 64. Added qualify_domain to the redirect router, to override the global
7168 65. If a pathologically long header line contained very many addresses (the
7169 report of this problem mentioned 10 000) and each of them was rewritten,
7170 Exim could use up a very large amount of memory. (It kept on making new
7171 copies of the header line as it rewrote, and never released the old ones.)
7172 At the expense of a bit more processing, the header rewriting function has
7173 been changed so that it no longer eats memory in this way.
7175 66. The generation of the Received: header has been moved from the time that a
7176 message starts to be received, to the time that it finishes. The timestamp
7177 in the Received: header should now be very close to that of the <= log
7178 line. There are two side-effects of this change:
7180 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7181 logged header lines no longer include the local Received: line, because
7182 it has not yet been created. The same applies to a copy of the message
7183 that is returned to a non-SMTP sender when a message is rejected.
7185 (b) When a filter file is tested using -bf, no additional Received: header
7186 is added to the test message. After some thought, I decided that this
7189 This change does not affect the value of $received_for. It is still set
7190 after address rewriting, but before local_scan() is called.
7192 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7194 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7195 gave an unhelpful panic error message, and a defer error. I have managed to
7196 change this behaviour so that it now rejects any supplied certificate,
7197 which seems right, as the list of acceptable certificates is empty.
7199 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7200 gave an unhelpful defer error. I have not managed to make this reject any
7201 supplied certificates, but the error message it gives is "no certificate
7202 supplied", which is not helpful.
7204 70. exigrep's output now also includes lines that are not associated with any
7205 message, but which match the given pattern. Implemented by a patch from
7206 Martin Sluka, which also tidied up the Perl a bit.
7208 71. Recipient callout verification, like sender verification, was using <> in
7209 the MAIL FROM command. This isn't really the right thing, since the actual
7210 sender may affect whether the remote host accepts the recipient or not. I
7211 have changed it to use the actual sender in the callout; this means that
7212 the cache record is now keyed on a recipient/sender pair, not just the
7213 recipient address. There doesn't seem to be a real danger of callout loops,
7214 since a callout by the remote host to check the sender would use <>.
7215 [SEE ABOVE: changed after hitting problems.]
7217 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7218 temporary errors. However, in the case of such a code being given after
7219 the end of a data transmission (i.e. after ".") Exim was failing to write
7220 a retry record for the message. (Yes, there was some broken host that was
7221 actually sending 8xx at this point.)
7223 73. An unknown lookup type in a host list could cause Exim to panic-die when
7224 the list was checked. (An example that provoked this was putting <; in the
7225 middle of a list instead of at the start.) If this happened during a DATA
7226 ACL check, a -D file could be left lying around. This kind of configuration
7227 error no longer causes Exim to die; instead it causes a defer error. The
7228 incident is still logged to the main and panic logs.
7230 74. Buglet left over from Exim 3 conversion. The message "too many messages
7231 in one connection" was written to the rejectlog but not the mainlog, except
7232 when address rewriting (yes!) was being logged.
7234 75. Added write_rejectlog option.
7236 76. When a system filter was run not as root (that is, when system_filter_user
7237 was set), the values of the $n variables were not being returned to the
7238 main process; thus, they were not subsequently available in the $sn
7241 77. Added +return_path_on_delivery log selector.
7243 78. A connection timeout was being treated differently from recipients deferred
7244 when testing hosts_max_try with a message that was older than the host's
7245 retry timeout. (The host should not be counted, thus allowing all hosts to
7246 be tried at least once before bouncing.) This may have been the cause of an
7247 occasionally reported bug whereby a message would remain on the queue
7248 longer than the retry timeout, but would be bounced if a delivery was
7249 forced. I say "may" because I never totally pinned down the problem;
7250 setting up timeout/retry tests is difficult. See also the next item.
7252 79. The ultimate address timeout was not being applied to errors that involved
7253 a combination of host plus message (for example, a timeout on a MAIL
7254 command). When an address resolved to a number of possible hosts, and they
7255 were not all tried for each delivery (e.g. because of hosts_max_try), a
7256 message could remain on the queue longer than the retry timeout.
7258 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7261 81. Fixed an obscure SMTP outgoing bug which required at least the following
7262 conditions: (a) there was another message waiting for the same server;
7263 (b) the server returned 5xx to all RCPT commands in the first message so
7264 that the message was not completed; (c) the server dropped the connection
7265 or gave a negative response to the RSET that Exim sends to abort the
7266 transaction. The observed case was a dropped connection after DATA that had
7267 been sent in pipelining mode. That is, the server had advertised PIPELINING
7268 but was not implementing it correctly. The effect of the bug was incorrect
7269 behaviour, such as trying another host, and this could lead to a crash.
7275 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7276 and daemon.c were passed as pointers to ints; they should have been
7277 pointers to socklen_t variables (which are typically unsigned ints).
7279 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7282 3. Fixed a really odd bug that affected only the testing scheme; patching a
7283 certain fixed string in the binary changed the value of another string that
7284 happened to be identical to the end of the original first string.
7286 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7287 name", it returns that address as the IP address. On some operating
7288 systems (e.g. Solaris), it also passes back the IP address string as the
7289 "host name". However, on others (e.g. Linux), it passes back an empty
7290 string. Exim wasn't checking for this, and was changing the host name to an
7291 empty string, assuming it had been canonicalized.
7293 5. Although rare, it is permitted to have more than one PTR record for a given
7294 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7295 all the names associated with an address, because they do in Solaris.
7296 However, it seems that they do not in Linux for data that comes from the
7297 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7298 I found this out when I moved to a new Linux workstation and tried to run
7299 the Exim test suite.
7301 To get round this problem I have changed the code so that it now does its
7302 own call to the DNS to look up PTR records when searching for a host name.
7303 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7304 addresses that are only in /etc/hosts are still found.
7306 This behaviour is, however, controlled by an option called host_lookup_
7307 order, which defaults to "bydns:byaddr". If people want to use the other
7308 order, or indeed, just use one or the other means of lookup, they can
7309 specify it in this variable.
7311 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7312 some operating systems, this comes back from gethostbyaddr() as an empty
7313 string, and this is what Exim used to test for. However, it seems that in
7314 other systems, "." is yielded. Exim now tests for this case too.
7316 7. The values of check_spool_space and check_log_space are now held internally
7317 as a number of kilobytes instead of an absolute number of bytes. If a
7318 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7319 kilobyte. This means that much larger values can be stored.
7321 8. Exim monitor: an attempt to get the action menu when not actually pointing
7322 at a message produces an empty menu entitled "No message selected". This
7323 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7324 no entries in it ("Shell widget menu has zero width and/or height"). So I
7325 have added a single, blank menu entry in this case.
7327 9. Added ${quote_local_part.
7329 10. MIME decoding is now applied to the contents of Subject: header lines when
7332 11. Now that a reference to $sender_host_address automatically causes a reverse
7333 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7334 host lookup before query-style lookups in lists that might use this
7335 variable. This has therefore been abolished, and the "net-" prefix is no
7336 longer necessary for query-style lookups.
7338 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7339 have been a typo for LFLAGS, so it has been changed.
7341 13. The install script calls Exim with "-C /dev/null" in order to find the
7342 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7343 to be output. However, since Exim outputs its version number before the
7344 error, it didn't break the script. It just looked ugly. I fixed this by
7345 always allowing "-C /dev/null" if the caller is root.
7347 14. Ignore overlarge ACL variable number when reading spool file - insurance
7348 against a later release with more variables having written the file.
7350 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7351 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7352 this, as well as the form without the "IPv6" on the front.
7354 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7355 OS/Makefile-Darwin file.
7357 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7358 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7360 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7361 variable. However, some OS use ints for inodes. I've added cast to long int
7362 to get rid of the compiler warning.
7364 19. I had forgotten to lock out "/../" in configuration file names when
7365 ALT_CONFIG_PREFIX was set.
7367 20. Routers used for verification do not need to specify transports. However,
7368 if such a router generated a host list, and callout was configured, Exim
7369 crashed, because it could not find a port number from the (non-existent)
7370 transport. It now assumes port 25 in this circumstance.
7372 21. Added the -t option to exigrep.
7374 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7375 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7376 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7377 them is set, LOOKUP_LSEARCH is forced.
7379 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7381 24. Added sender and host information to the "rejected by local_scan()" log
7382 line; previously there was no indication of these.
7384 25. Added .include_if_exists.
7386 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7387 Linux. It turns out that not all file systems support this. Apparently some
7388 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7389 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7390 not supported on the file descriptor, is now ignored when Exim is trying to
7391 sync a directory. This applies only to Linux.
7393 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7395 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7396 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7399 29. Added support for maildirsize files from supplied patch (modified a bit).
7401 30. The use of :fail: followed by an empty string could lead Exim to respond to
7402 sender verification failures with (e.g.):
7404 550 Verification failed for <xxx>
7405 550 Sender verify failed
7407 where the first response line was missing the '-' that indicates it is not
7408 the final line of the response.
7410 31. The loop for finding the name of the user that called Exim had a hardwired
7411 limit of 10; it now uses the value of finduser_retries, which is used for
7412 all other user lookups.
7414 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7417 33. Exim was neglecting to zero errno before one call of strtol() when
7418 expanding a string and expecting an integer value. On some systems this
7419 resulted in spurious "integer overflow" errors. Also, it was casting the
7420 result into an int without checking.
7422 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7423 did not work. The code looks as if it has *never* worked, though it appears
7424 to have been documented since at least release 1.62. I have made it work.
7426 35. The "timeout_DNS" error in retry rules, also documented since at least
7427 1.62, also never worked. As it isn't clear exactly what this means, and
7428 clearly it isn't a major issue, I have abolished the feature by treating it
7429 as "timeout", and writing a warning to the main and panic logs.
7431 36. The display of retry rules for -brt wasn't always showing the error code
7434 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7435 timeout_connect_A, timeout_connect_MX.
7437 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7438 to the empty sender.
7440 39. The daemon was not analysing the content of -oX till after it had closed
7441 stderr and disconnected from the controlling terminal. This meant that any
7442 syntax errors were only noted on the panic log, and the return code from
7443 the command was 0. By re-arranging the code a little, I've made the
7444 decoding happen first, so such errors now appear on stderr, and the return
7445 code is 1. However, the actual setting up of the sockets still happens in
7446 the disconnected process, so errors there are still only recorded on the
7449 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7450 connections (as happens on some IP stacks) was logged at start up time as
7451 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7452 it from "IPv6 and IPv4", which means that two separate sockets are being
7455 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7456 says whether AF_INET or AF_INET6 was passed as an argument.
7458 42. Exiwhat output was messed up when time zones were included in log
7461 43. Exiwhat now gives more information about the daemon's listening ports,
7462 and whether -tls-on-connect was used.
7464 44. The "port" option of the smtp transport is now expanded.
7466 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7467 silently ignored. Now an error message is written to the main and panic
7470 46. There's a new ACL modifier called "logwrite" which writes to a log file
7471 as soon as it is encountered.
7473 47. Added $local_user_uid and $local_user_gid at routing time.
7475 48. Exim crashed when trying to verify a sender address that was being
7478 49. Exim was recognizing only a space character after ".include". It now also
7479 recognizes a tab character.
7481 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7482 extracting the relevant information from the specification. The man page no
7483 longer contains scrambled data for the -d option, and I've added a section
7484 at the front about calling Exim under different names.
7486 51. Added "extra_headers" argument to the "mail" command in filter files.
7488 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7491 53. Installed eximstats 1.29.
7493 54. Added transport_filter_timeout as a generic transport option.
7495 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7496 Cc: header lines. This was required by RFC 822, but it not required by RFC
7499 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7500 incoming messages that did not have them. Now it does so only if the
7501 message originates locally, that is, if there is no associated remote host
7502 address. When Resent- header lines are present, this applies to the Resent-
7503 lines rather than the non-Resent- lines.
7505 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7506 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7508 58. Messages for configuration errors now include the name of the main
7509 configuration file - useful now that there may be more than one file in a
7510 list (.included file names were always shown).
7512 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7513 for those rare installations that do not start the daemon as root or run it
7514 setuid root. I've cut out the call to initgroups() if the daemon is not
7517 60. The Exim user and group can now be bound into the binary as text strings
7518 that are looked up at the start of Exim's processing.
7520 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7522 62. Added $mailstore_basename variable.
7524 63. Installed patch to sieve.c from Michael Haardt.
7526 64. When Exim failed to open the panic log after failing to open the main log,
7527 the original message it was trying to log was written to stderr and debug
7528 output, but if they were not available (the usual case in production), it
7529 was lost. Now it is written to syslog before the two lines that record the
7530 failures to open the logs.
7532 65. Users' Exim filters run in subprocesses under the user's uid. It is
7533 possible for a "deliver" command or an alias in a "personal" command to
7534 provoke an address rewrite. If logging of address rewriting is configured,
7535 this fails because the process is not running as root or exim. There may be
7536 a better way of dealing with this, but for the moment (because 4.30 needs
7537 to be released), I have disabled address rewrite logging when running a
7538 filter in a non-root, non-exim process.
7544 1. The buildconfig auxiliary program wasn't quoting the value set for
7545 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7546 not defined. This bug was masked in 4.22 by the effect that was fixed in
7549 2. Some messages that were rejected after a message id was allocated were
7550 shown as "incomplete" by exigrep. It no longer does this for messages that
7551 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7553 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7554 have allow_domain_literals set, the ID did not get logged in the <= line.
7555 Domain literals are now always recognized in Message-ID: header lines.
7557 4. The first argument for a ${extract expansion item is the key name or field
7558 number. Leading and trailing spaces in this item were not being ignored,
7559 causing some misleading effects.
7561 5. When deliver_drop_privilege was set, single queue runner processes started
7562 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7563 same command in the process it spins off) were not dropping privilege.
7565 6. When the daemon running as "exim" started a queue runner, it always
7566 re-executed Exim in the spun-off process. This is a waste of effort when
7567 deliver_drop_privilege is set. The new process now just calls the
7568 queue-runner function directly.
7574 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7577 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7578 ignored. Though the use of -oP was forcing the writing of a pid file, it
7579 was always written to the default place.
7581 3. If the message "no IP address found for host xxxx" is generated during
7582 incoming verification, it is now followed by identification of the incoming
7583 connection (so you can more easily find what provoked it).
7585 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7587 5. Added some features to "harden" Exim a bit more against certain attacks:
7589 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7590 be put in Local/Makefile. This is like the never_users runtime option,
7591 but it cannot be overridden. The default setting is "root".
7593 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7594 prefix string with which any file named in a -C command line option
7597 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7598 is retained for -C and -D only if the caller of Exim is root. Without
7599 it, the exim user may also use -C and -D and retain privilege.
7601 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7602 command line option is disabled.
7604 6. Macro names set by the -D option must start with an upper case letter, just
7605 like macro names defined in the configuration file.
7607 7. Added "dereference=" facility to LDAP.
7609 8. Two instances of the typo "uknown" in the source files are fixed.
7611 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7612 the Configure-Makefile script screwed up while processing it.
7614 10. Incorporated PCRE 4.4.
7616 11. The SMTP synchronization check was not operating right at the start of an
7617 SMTP session. For example, it could not catch a HELO sent before the client
7618 waited for the greeting. There is now a check for outstanding input at the
7619 point when the greeting is written. Because of the duplex, asynchronous
7620 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7621 way, but not yet received, when the check is performed.
7623 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7624 on TCP/IP sockets, because this apparently causes some broken clients to
7627 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7628 unchanged) from the Cygwin maintainer.
7630 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7631 was supported instead of testing for mbx. Effectively a typo.
7633 15. The spa authenticator server code was not checking that the input it
7634 received was valid base64.
7636 16. The debug output line for the "set" modifier in ACLs was not showing the
7637 name of the variable that was being set.
7639 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7641 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7642 lookup on its own. Something else was needed to trigger the lookup. For
7643 example, a match in host_lookup or the need for a host name in a host list.
7644 Now, if $sender_host_name is referenced and the host name has not yet been
7645 looked up, a lookup is performed. If the lookup fails, the variable remains
7646 empty, and $host_lookup_failed is set to "1".
7648 19. Added "eqi" as a case-independent comparison operator.
7650 20. The saslauthd authentication condition could segfault if neither service
7651 nor realm was specified.
7653 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7654 error message that was logged was misleading, and incoming SMTP
7655 connections were dropped. The message is now more accurate, and temporary
7656 errors are given to SMTP connections.
7658 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7659 (except RSET) with a 421 error, until QUIT is received. However, it was
7660 failing to send a response to QUIT.
7662 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7663 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7664 ACL is now run after the helo_try_verify_hosts code.
7666 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7667 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7668 case-independent, so other case variants are also recognized). Apparently
7669 some systems use these upper case variants.
7671 25. If more than two messages were waiting for the same host, and a transport
7672 filter was specified for the transport, Exim sent two messages over the
7673 same TCP/IP connection, and then failed with "socket operation on non-
7674 socket" when it tried to send the third.
7676 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7678 27. The extern definition of crypt16() in expand.c was not being excluded when
7679 the OS had its own crypt16() function.
7681 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7682 as a preferred synonym for return_size_limit, both as an option and as an
7685 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7687 30. Changed the default configuration ACL to relax the local part checking rule
7688 for addresses that are not in any local domains. For these addresses,
7689 slashes and pipe symbols are allowed within local parts, but the sequence
7690 /../ is explicitly forbidden.
7692 31. SPA server authentication was not clearing the challenge buffer before
7695 32. log_message in a "warn" ACL statement was writing to the reject log as
7696 well as to the main log, which contradicts the documentation and doesn't
7697 seem right (because no rejection is happening). So I have stopped it.
7699 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7700 However, I am unable to do any testing of this.
7702 34. Fixed an infelicity in the appendfile transport. When checking directories
7703 for a mailbox, to see if any needed to be created, it was accidentally
7704 using path names with one or more superfluous leading slashes; tracing
7705 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7707 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7708 contents are added to the log line that is written for every discarded
7709 recipient. (Previously a log_message setting was ignored.)
7711 36. The ${quote: operator now quotes the string if it is empty.
7713 37. The install script runs exim in order to find its version number. If for
7714 some reason other than non-existence or emptiness, which it checks, it
7715 could not run './exim', it was installing it with an empty version number,
7716 i.e. as "exim-". This error state is now caught, and the installation is
7719 38. An argument was missing from the function that creates an error message
7720 when Exim fails to connect to the socket for saslauthd authentication.
7721 This could cause Exim to crash, or give a corrupted message.
7723 39. Added isip, isip4, and isip6 to ${if conditions.
7725 40. The ACL variables $acl_xx are now saved with the message, and can be
7726 accessed later in routers, transports, and filters.
7728 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7729 strings in the file are not string-expanded.
7731 42. If a MAIL command specified a SIZE value that was too large to fit into an
7732 int variable, the check against message_size_limit failed. Such values are
7733 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7734 day this will have to be increased, but I don't think I want to be around
7735 when emails are that large.
7742 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7743 iconv() is not standard in FreeBSD.
7745 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7746 IPv6 enabled. The observed symptom was a segmentation fault on return from
7747 the function os_common_find_running_interfaces() in src/os.c.
7749 3. In the check_special_case() function in daemon.c I had used "errno" as an
7750 argument name, which causes warnings on some systems. This was basically a
7751 typo, since it was named "eno" in the comments!
7753 4. The code that waits for the clock to tick (at a resolution of some fraction
7754 of a second) so as to ensure message-id uniqueness was always waiting for
7755 at least one whole tick, when it could have waited for less. [This is
7756 almost certainly not relevant at current processor speeds, where it is
7757 unlikely to ever wait at all. But we try to future-proof.]
7759 5. The function that sleeps for a time interval that includes fractions of a
7760 second contained a race. It did not block SIGALRM between setting the
7761 timer, and suspending (a couple of lines later). If the interval was short
7762 and the sigsuspend() was delayed until after it had expired, the suspension
7763 never ended. On busy systems this could lead to processes getting stuck for
7766 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7767 process, before it forks any delivery processes. The open lookup caching
7768 mechanism meant that the open file or database connection was passed into
7769 the delivery process. The problem was that delivery processes always tidy
7770 up cached lookup data. This could cause a problem for the next delivery
7771 process started by the queue runner, because the external queue runner
7772 process does not know about the closure. So the next delivery process
7773 still has data in the lookup cache. In the case of a file lookup, there was
7774 no problem because closing a file descriptor in a subprocess doesn't affect
7775 the parent. However, if the lookup was caching a connection to a database,
7776 the connection was closed, and the second delivery process was likely to
7777 see errors such as "PGSQL: query failed: server closed the connection
7778 unexpectedly". The problem has been fixed by closing all cached lookups
7779 in a queue runner before running a delivery process.
7781 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7782 seem to have the "const" qualifier which it has on other OS. I've
7785 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7786 *of the same type* (client or server) with the same public name that an
7787 error should be diagnosed.
7789 9. When Exim looked up a host name for an IP address, but failed to find the
7790 original IP address when looking up the host name (a safety check), it
7791 output the message "<ip address> does not match any IP for NULL", which was
7792 confusing, to say the least. The bug was that the host name should have
7793 appeared instead of "NULL".
7795 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7796 user that is built into the binary, and the -C or -D options is used, root
7797 privilege is dropped before the configuration file is read. In addition,
7798 logging is switched to stderr instead of the normal log files. If the
7799 configuration then re-defines the Exim user, the unprivileged environment
7800 is probably not what is expected, so Exim logs a panic warning message (but
7803 However, if deliver_drop_privilege is set, the unprivileged state may well
7804 be exactly what is intended, so the warning has been cut out in that case,
7805 and Exim is allowed to try to write to its normal log files.
7811 1. smtp_return_error_details was not giving details for temporary sender
7812 or receiver verification errors.
7814 2. Diagnose a configuration error if two authenticators have the same public
7817 3. Exim used not to create the message log file for a message until the first
7818 delivery attempt. This could be confusing when incoming messages were held
7819 for policy or load reasons. The message log file is now created at the time
7820 the message is received, and an initial "Received" line is written to it.
7822 4. The automatically generated man page for command line options had a minor
7823 bug that caused no ill effects; however, a more serious problem was that
7824 the procedure for building the man page automatically didn't always
7825 operate. Consequently, release 4.20 contains an out-of-date version. This
7826 shouldn't happen again.
7828 5. When building Exim with embedded Perl support, the script that builds the
7829 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7830 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7832 6. The freeze_tell option was not being used for messages that were frozen on
7833 arrival, either by an ACL or by local_scan().
7835 7. Added the smtp_incomplete_transaction log selector.
7837 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7838 was accepting AUTH without a new EHLO.
7840 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7841 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7844 10. Logging of TCP/IP connections (when configured) now happens in the main
7845 daemon process instead of the child process, so that the TCP/IP connection
7846 count is more accurate (but it can never be perfect).
7848 11. The use of "drop" in a nested ACL was not being handled correctly in the
7849 outer ACL. Now, if condition failure induced by the nested "drop" causes
7850 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7851 or "require"), the connection is dropped.
7853 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7854 that yield "discard" can now be used with an "accept" or a "discard" verb,
7855 but an error is generated for any others (because I can't see a useful way
7856 to define what should happen).
7858 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7859 are now processed in the same way as lines in the Exim configuration file.
7860 In particular, continuation lines are supported.
7862 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7864 15. Added -ti meaning -t -i.
7866 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7867 domains, and warn by logging if others are found.
7869 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7870 returned from GIFCONF when Exim is trying to find the list of interfaces on
7871 a host. The code in os.c has been modified to copy each ifreq to an aligned
7872 structure in all cases.
7874 Also, in some cases, the returned ifreq's were being copied to a 'struct
7875 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7876 means the last couple of bytes of an IPv6 address could be chopped if the
7877 ifreq contained only a normal sockaddr (14 bytes storage).
7879 18. Named domain lists were not supported in the hosts_treat_as_local option.
7880 An entry such as +xxxx was not recognized, and was treated as a literal
7883 19. Ensure that header lines added by a DATA ACL are included in the reject log
7884 if the ACL subsequently rejects the message.
7886 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7887 MD5 (which is deprecated).
7889 21. When testing a filter file using -bf, Exim was writing a message when it
7890 took the sender from a "From " line in the message, but it was not doing so
7891 when it took $return_path from a Return-Path: header line. It now does.
7893 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7894 with a valid header line field name (a series of printing characters
7895 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7897 23. Changed "disc" in the source to "disk" to conform to the documentation and
7898 the book and for uniformity.
7900 24. Ignore Sendmail's -Ooption=value command line item.
7902 25. When execve() failed while trying to run a command in a pipe transport,
7903 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7904 could be confused with a return value of 69 from the command itself. This
7905 has been changed to 127, the value the shell returns if it is asked to run
7906 a non-existent command. The wording for the related log line suggests a
7907 non-existent command as the problem.
7909 26. If received_header_text expands to an empty string, do not add a Received:
7910 header line to the message. (Well, it adds a token one on the spool, but
7911 marks it "old" so that it doesn't get used or transmitted.)
7913 27. Installed eximstats 1.28 (addition of -nt option).
7915 28. There was no check for failure on the call to getsockname() in the daemon
7916 code. This can fail if there is a shortage of resources on the system, with
7917 ENOMEM, for example. A temporary error is now given on failure.
7919 29. Contrary to the C standard, it seems that in some environments, the
7920 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7921 program. Exim now does this explicitly; it affects the formatting of
7922 timestamps using strftime().
7924 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7925 complaints. I've now initialized all the variables, to avoid this.
7927 32. Header lines added by a system filter were not being "seen" during
7928 transport-time rewrites.
7930 33. The info_callback() function passed to OpenSSL is set up with type void
7931 (*)(SSL *, int, int), as described somewhere. However, when calling the
7932 function (actually a macro) that sets it up, the type void(*)() is
7933 expected. I've put in a cast to prevent warnings from picky compilers.
7935 34. If a DNS black list lookup found a CNAME record, but there were no A
7936 records associated with the domain it pointed at, Exim crashed.
7938 35. If a DNS black list lookup returned more than one A record, Exim ignored
7939 all but the first. It now scans all returned addresses if a particular IP
7940 value is being sought. In this situation, the contents of the
7941 $dnslist_value variable are a list of all the addresses, separated by a
7944 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7945 to accept a host name and all its aliases if the forward lookup for any of
7946 them yielded the IP address of the incoming connection. Now it accepts only
7947 those names whose forward lookup yields the correct IP address. Any other
7948 names are discarded. This closes a loophole whereby a rogue DNS
7949 administrator could create reverse DNS records to break through a
7950 wildcarded host restriction in an ACL.
7952 37. If a user filter or a system filter that ran in a subprocess used any of
7953 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7954 the wrong values were passed to the pipe command ($thisaddress had the
7955 value of $0, $0 had the value of $1, etc). This bug was introduced by
7956 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7958 38. Improved the line breaking for long SMTP error messages from ACLs.
7959 Previously, if there was no break point between 40 and 75 characters, Exim
7960 left the rest of the message alone. Two changes have been made: (a) I've
7961 reduced the minimum length to 35 characters; (b) if it can't find a break
7962 point between 35 and 75 characters, it looks ahead and uses the first one
7963 that it finds. This may give the occasional overlong line, but at least the
7964 remaining text gets split now.
7966 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7967 file descriptors might be low, and that setting 1000 would always raise it.
7968 It turns out that in some environments, the limit is already over 1000 and
7969 that lowering it causes trouble. So now Exim takes care not to decrease it.
7971 40. When delivering a message, the value of $return_path is set to $sender_
7972 address at the start of routing (routers may change the value). By an
7973 oversight, this default was not being set up when an address was tested by
7974 -bt or -bv, which affected the outcome if any router or filter referred to
7977 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7978 log only when "message" or "log_message" are set. However, if one of the
7979 conditions was an address verification, or a call to a nested ACL, the
7980 messages generated by the underlying test were being passed through. This
7981 no longer happens. The underlying message is available in $acl_verify_
7982 message for both "message" and "log_message" expansions, so it can be
7983 passed through if needed.
7985 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7986 new expansion $bh_ to give the encoded byte string without charset
7987 translation. Translation happens only if iconv() is available; HAVE_ICONV
7988 indicates this at build time. HEADERS_CHARSET gives the charset to
7989 translate to; headers_charset can change it in the configuration, and
7990 "headers charset" can change it in an individual filter file.
7992 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7993 that creates RFC 2047 encoded "words" labels them as that charset instead
7994 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7995 expansion operator; (ii) when Exim creates a From: line for a local
7996 message; (iii) when a header line is rewritten to include a "phrase" part.
7998 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7999 buggy, causing it to skip the first lines of messages whose message ID
8000 ended in 'D'. This would not have bitten before Exim release 4.14, because
8001 message IDs were unlikely to end in 'D' before then. The effect was to have
8002 incorrect size information for certain domains.
8004 45. #include "config.h" was missing at the start of the crypt16.c module. This
8005 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8008 46. If there was a timeout during a "random" callout check, Exim treated it as
8009 a failure of the random address, and carried on sending RSET and the real
8010 address. If the delay was just some slowness somewhere, the response to the
8011 original RCPT would be taken as a response to RSET and so on, causing
8012 mayhem of various kinds.
8014 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8015 when I implemented it. It didn't allow for the fact that some option values
8016 may legitimately be negative (e.g. size_addition), and it didn't even do
8017 the right test for positive values.
8019 48. Domain names in DNS records are case-independent. Exim always looks them up
8020 in lower case. Some resolvers return domain names in exactly the case they
8021 appear in the zone file, that is, they may contain uppercase letters. Not
8022 all resolvers do this - some return always lower case. Exim was treating a
8023 change of case by a resolver as a change of domain, similar to a widening
8024 of a domain abbreviation. This triggered its re-routing code and so it was
8025 trying to route what was effectively the same domain again. This normally
8026 caused routing to fail (because the router wouldn't handle the domain
8027 twice). Now Exim checks for this case specially, and just changes the
8028 casing of the domain that it ultimately uses when it transmits the message
8031 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8034 50. If a filter generated a file delivery with a non-absolute name (possible if
8035 no home directory exists for the router), the forbid_file option was not
8038 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8039 the existing equality matching.
8041 52. Exim was using ints instead of ino_t variables in some places where it was
8042 dealing with inode numbers.
8044 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8045 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8046 called TMPDIR, and if it finds it is different, it changes its value.
8048 54. The smtp_printf() function is now made available to local_scan() so
8049 additional output lines can be written before returning. There is also an
8050 smtp_fflush() function to enable the detection of a dropped connection.
8051 The variables smtp_input and smtp_batched_input are exported to
8054 55. Changed the default runtime configuration: the message "Unknown user"
8055 has been removed from the ACL, and instead placed on the localuser router,
8056 using the cannot_route_message feature. This means that any verification
8057 failures that generate their own messages won't get overridden. Similarly,
8058 the "Unrouteable address" message that was in the ACL for unverifiable
8059 relay addresses has also been removed.
8061 56. Added hosts_avoid_esmtp to the smtp transport.
8063 57. The exicyclog script was not checking for the esoteric option
8064 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8065 will work only if exicyclog is run under the appropriate euid.
8067 58. Following a discussion on the list, the rules by which Exim recognises line
8068 endings on incoming messages have been changed. The -dropcr and drop_cr
8069 options are now no-ops, retained only for backwards compatibility. The
8070 following line terminators are recognized: LF CRLF CR. However, special
8071 processing applies to CR:
8073 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8074 nor a local message in the state where . is a terminator.
8076 (ii) If a bare CR is encountered in a header line, an extra space is added
8077 after the line terminator so as not to end the header. The reasoning
8078 behind this is that bare CRs in header lines are most likely either
8079 to be mistakes, or people trying to play silly games.
8081 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8082 was being incorrectly given as 18 bytes larger than it should have been.
8083 This is a VOB (very old bug).
8085 60. This may never have affected anything current, but just in case it has:
8086 When the local host is found other than at the start of a list of hosts,
8087 the local host, those with the same MX, and any that follow, are discarded.
8088 When the list in question was part of a longer list of hosts, the following
8089 hosts (not currently being processed) were also being discarded. This no
8090 longer happens. I'm not sure if this situation could ever has previously
8093 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8096 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8097 header. This is something that is recommended in a new Internet Draft, and
8098 is something that is documented as being done by Sendmail. There are two
8099 possible values. For messages generated by the autoreply transport, Exim
8102 Auto-Submitted: auto-replied
8104 whereas for all other generated messages (e.g. bounces) it adds
8106 Auto-Submitted: auto-generated
8108 63. The "personal" condition in filters now includes a test for the
8109 Auto-Submitted: header. If it contains the string "auto-" the message it
8110 not considered personal.
8112 64. Added rcpt_include_affixes as a generic transport option.
8114 65. Added queue_only_override (default true).
8116 66. Added the syslog_duplication option.
8118 67. If what should have been the first header line of a message consisted of
8119 a space followed by a colon, Exim was mis-interpreting it as a header line.
8120 It isn't of course - it is syntactically invalid and should therefore be
8121 treated as the start of the message body. The misbehaviour could have
8122 caused a number of strange effects, including loss of data in subsequent
8123 header lines, and spool format errors.
8125 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8126 client host had authenticated. This control can now be exercised by an ACL
8127 for more flexibility.
8129 69. By default, callouts do not happen when testing with -bh. There is now a
8130 variant, -bhc, which does actually run the callout code, including
8131 consulting and updating the callout cache.
8133 70. Added support for saslauthd authentication, courtesy of Alexander
8136 71. If statvfs() failed on the spool or log directories while checking their
8137 size for availability, Exim confusingly gave the error "space shortage".
8138 Furthermore, in debugging mode it crashed with a floating point exception.
8139 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8140 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8141 serious problem, Exim now writes to the main and panic logs when this
8142 happens, with details of the failure. It then refuses to accept the
8143 incoming message, giving the message "spool directory problem" or "log
8144 directory problem" with a 421 code for SMTP messages.
8146 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8147 0, 1, and 2 exist, because some OS complain for execs without them (see
8148 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8149 descriptors. However, the code omitted to check that the open succeeded,
8150 causing mysterious errors if for some reason the permissions on /dev/null
8151 got screwed. Now Exim writes a message to the main and panic logs, and
8152 bombs out if it can't open /dev/null.
8154 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8155 interact so that it is all more flexible. It is supposed to remain
8156 backwards compatible. Also added extra_local_interfaces.
8158 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8159 to bomb out with an assertion failure - to the client this appears as a
8160 connection drop. This problem occurs in the part of the code that was taken
8161 from the Samba project. Fortunately, the assertion is in a very simple
8162 function, so I have fixed this by reproducing the function inline in the
8163 one place where it is called, and arranging for authentication to fail
8164 instead of killing the process with assert().
8166 75. The SPA client code was not working when the server requested OEM rather
8167 than Unicode encoding.
8169 76. Added code to make require_files with a specific uid setting more usable in
8170 the case where statting the file as root fails - usually a non-root-mounted
8171 NFS file system. When this happens and the failure is EACCES, Exim now
8172 forks a subprocess and does the per-uid checking as the relevant uid.
8174 77. Added process_log_path.
8176 78. If log_file_path was not explicitly set, a setting of check_log_space or
8177 check_log_inodes was ignored.
8179 79. If a space check for the spool or log partitions fails, the incident is now
8180 logged. Of course, in the latter case the data may get lost...
8182 80. Added the %p formatting code to string_format() so that it can be used to
8183 print addresses in debug_print(). Adjusted all the address printing in the
8184 debugging in store.c to use %p rather than %d.
8186 81. There was a concern that a line of code in smtp_in.c could overflow a
8187 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8188 initially expressed, the concern was not well-founded, because trailing
8189 spaces are removed early. However, if the trailing spaces were followed by
8190 a NULL, they did not get removed, so the overflow was possible. Two fixes
8193 (a) I re-wrote the offending code in a cleaner fashion.
8194 (b) If an incoming SMTP command contains a NULL character, it is rejected
8197 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8198 runs initgroups(), so that if the Exim user is in any additional groups,
8199 they will be used during message reception.
8205 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob