1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
86 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
87 supplied as the db file-pair basename. Previously, if a directory
88 path was given, for example via the autoreply "once" option, the DB
89 file.pag and file.dir files would be created in that directory's
96 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
97 related applications. Previously an "H" was used where available info
98 says that "M" should be, so change to match.
100 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
101 as arguments, so an implementation trying to copy these into a local
102 buffer was taking a taint-enforcement trap. Fix by using dynamically
103 created buffers. Similar fix for radius expansion condition.
105 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
106 reasonable, eg. to count headers. Fix by using dynamically created
107 buffers rather than a local. Do similar fixes for ACL actions "dcc",
108 "log_reject_target", "malware" and "spam"; the arguments are expanded
109 so could be handling tainted values.
111 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
112 broken the (no-op) support for this sendmail command. Restore it
113 to doing nothing, silently, and returning good status.
115 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
116 record path was given (or the default used) without a leading directory
117 path, an error occurred on trying to open it. Use the transport's working
120 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
121 transport. Previously it was the name on the DNS A-record; use instead
122 the head of the CNAME chain leading there (if there is one). This seems
123 to align better with RFC 6125.
125 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
126 smtp_accept_max_per_host allocated resources which were not released
127 when the limit was exceeded. This eventually crashed the daemon. Fix
128 by adding a release action in that path.
130 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
131 expanded; previously using tainted values was rejected. Fix by using
132 dynamically-created buffers.
134 JH/09 Relax restrictions on ACL verify condition needing access to message
135 headers. Previously they were only permitted in data and non-smtp ACLs;
136 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
137 not_blind, header_sender and header_names_ascii verification.
139 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
140 Previously a macro used one argument twice; when called with the
141 argument as an expression having side-effects, incorrect operation
142 resulted. Use an inlineable function.
144 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
145 held open for a verify callout. Previously this wan not accounted for
146 and a corrupt onward SMTP conversation resulted.
148 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
149 Previously they were embedded naively in the constructed address; when
150 needed, strip the quoting and quote the entire local_part.
151 Also make the inbound_srs expansion condition handle quoting.
153 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
154 excluded, not matching the documentation.
156 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
157 was given for the sqlite_dbfile a trap resulted.
159 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
160 "name" argument resulted in a trap. There is no reason to disallow such;
161 this was a coding error.
163 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
164 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
165 spent suspended, ignoring the POSIX definition. Previously we assumed
166 it did and a constant offset from real time could be used as a correction.
167 Change to using the same clock source for the start-of-message and the
168 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
169 exists, just to get a clock slightly more aligned to reality.
171 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
172 RFC says it is optional some validators care. The missing char was not
173 intended but triggered by a line-wrap alignment. Discovery and fix by
174 Guillaume Outters, hacked on by JH.
176 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
177 name being quoted was tainted a trap would be taken. Fix by using
178 dynamically created buffers. The routine could have been called by a
179 rewrite with the "h" flag, by using the "-F" command-line option, or
180 by using a "name=" option on a control=submission ACL modifier.
182 JH/19 SPF: change the Authentication-Results expansion component to give
183 smtp.helo when the sender domain is empty. Previously it gave
186 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
187 not in 127.0.0.0/8 to help in spotting list domains taken over by a
188 domain-parking registrar.
190 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
191 Previously when a whitespace character was specified it was not inserted
192 after removing the newline.
194 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
195 the domain part of the recipient address. This overrides any tls_sni
196 option set, which was previously used.
198 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
201 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
202 is_tainted() had an off-by-one error in the overenthusiastic direction.
203 Find and fix by Gavan. Although NetBSD is not a supported platform for
204 4.94 this bug could affect other platforms.
206 PP/01 Fix default prime selection to be consistent.
207 One path used ike23 still, instead of exim.dev.20160529.3; now both
208 execution flows will use the same DH primes (currently
209 exim.dev.20160529.3).
211 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
212 option in smtp transport, to match the documentation. Previously
213 verification was not being done in some cases where it should have been.
215 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
216 than one server was defined and depending on the platform memory layout
217 details, an internal consistency trap could be hit while walking the list
220 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
221 files. The value is used by the authresults expansion item. Previously
222 if this was used in a router or transport, a crash could result.
224 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
225 have a sufficient incidence of EINTR returns from select that an
226 interaction with other operations done by the main daemon loop exposed
227 a bug in the error-handling. This was benign apart from the log
230 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
231 for consistency with delivered "=>" and failed "**" lines. While we're
232 there, handle PRX and TFO.
234 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
235 applied. This resulted, if any header-line rewrite rules were configured,
236 in a panic-log triggerable by sending a message with a long address in
237 a header. Fix by increasing the arbitrary limit to larger than a single
238 (dewrapped) 5322 header line maximum size.
240 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
241 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
242 RFC 6648 which deprecates X- options in protocols as a general practice.
243 Changeover between the implementations is handled by the mechanisms
246 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
247 a remote which uses callout/hold. Previously the local was queued.
249 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
252 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
253 log line. The asterisk is now consistently appended to the remote IP
254 (and port, if given), and will also be provided on defer and fail log
255 lines. Previously it could be placed on the local IP if that was being
256 logged, and was only provided on delivery lines.
258 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
260 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
261 authenticator client_send option. Previously the next char, after a pair
262 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
263 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
264 documentation. There is still no way to get a leading ^ immediately
265 after a NUL (ie. for the password of a PLAIN method authenticator.
267 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
268 files. For bad sizes read, delete the record and whine to paniclog.
270 JH/38 When logging an AUTH failure, as server, do not include sensitive
271 information. Previously, the credentials would be included if given
272 as part of the AUTH command line and an ACL denied authentication.
274 JH/39 Bug 2691: fix $local_part_data. When the matching list element
275 referred to a file, bad data was returned. This likely also affected
278 JH/40 The gsasl authenticator now supports caching of the salted password
279 generated by the client-side implementation. This required the addition
280 of a new variable: $auth4.
282 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
283 left undeleted; the attempt to re-create it then failed - resulting in
284 the usual "SIGHUP tp have daemon reload configuration" to not work.
285 This affected any platform not supporting "abstract" Unix-domain
286 sockets (i.e. not Linux).
288 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
289 recipients" response to RCPT in a later response, with a 250. The
290 previous coding assumed this would not happen, and under PIPELINING
291 would result in both lost and duplicate recipients for a message.
293 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
294 Previously the weighting was incorrectly applied. Similar fix for socks
295 proxies. Found and fixed by Heiko Schlichting.
297 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
298 not handle sub-lists included using the +namedlist syntax. While
299 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
300 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
303 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
304 Previously the permanent pool was used, so the sore could not be freed.
305 This meant a connection with many messages would use continually-growing
308 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
309 per-pool so as not to waste too much space. Previously a constant size
310 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
311 DOS attacks harder. The cost is wasted memory use in the larger blocks.
313 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
314 same space repeatedly, and a smaller process footprint.
316 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
317 for GnuTLS. Previously the name started "%s" which, while not a bug,
318 looked as if if might be one.
320 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
321 "continued connection") make the $proxy_* variables available. Previously
322 the information was not passed across the exec() call for subsequent
323 transport executions. This also mean that the log lines for the
324 messages can show the proxy information.
326 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
327 receive time. With modern systems the difference is significant.
328 The historical behaviour can be restored by disabling (a new) log_selector
329 "queue_time_exclusive".
331 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
332 content) were specifically tested for. Now, also cover expansions
333 resulting in ACL names and inline ACL content.
335 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
336 making it unusable in complex expressions.
338 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
339 fail with a taint-check on the spoolfile name, and leave the message
342 HS/01 Enforce absolute PID file path name.
344 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
346 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
348 PP/02 Bug 2643: Correct TLS DH constants.
349 A missing NUL termination in our code-generation tool had led to some
350 incorrect Diffie-Hellman constants in the Exim source.
351 Reported by kylon94, code-gen tool fix by Simon Arlott.
353 PP/03 Impose security length checks on various command-line options.
354 Fixes CVE-2020-SPRSS reported by Qualys.
356 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
357 better. Reported by Qualys.
359 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
360 providing a particularly obnoxious sender full name.
363 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
365 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
368 PP/08 Change default for recipients_max from unlimited to 50,000.
370 PP/09 Fix security issue with too many recipients on a message (to remove a
371 known security problem if someone does set recipients_max to unlimited,
372 or if local additions add to the recipient list).
373 Fixes CVE-2020-RCPTL reported by Qualys.
375 PP/10 Fix security issue in SMTP verb option parsing
376 Fixes CVE-2020-EXOPT reported by Qualys.
378 PP/11 Fix security issue in BDAT state confusion.
379 Ensure we reset known-good where we know we need to not be reading BDAT
380 data, as a general case fix, and move the places where we switch to BDAT
381 mode until after various protocol state checks.
382 Fixes CVE-2020-BDATA reported by Qualys.
384 HS/03 Die on "/../" in msglog file names
386 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
387 the Exim runtime user are allowed to create files.
389 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
392 QS/03 When reading the output from interpreted forward files we do not
393 pass the pipe between the parent and the interpreting process to
394 executed child processes (if any).
396 QS/04 Always die if requested from internal logging, even is logging is
399 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
400 the API; compilation noo longer completes with DMARC support included.
401 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
402 been reported on other platforms.
404 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
406 JH/56 Make the majority of info read from config files readonly, for defence-in-
407 depth against exploits. Suggestion by Qualys.
408 Not supported on Solaris 10.
410 JH/57 Fix control=fakreject for a custom message containing tainted data.
411 Previously this resulted in a log complaint, due to a re-expansion present
412 since fakereject was originally introduced.
414 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
415 modifier was given, a loop resulted.
417 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
418 pipelined SMTP command followed the BDAT LAST then it would be
419 incorrectly treated as part of the message body, causing a verification
422 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
423 require looser validation rules than those for 821-level addresses,
424 which only permit IP addresses.
430 JH/01 Avoid costly startup code when not strictly needed. This reduces time
431 for some exim process initialisations. It does mean that the logging
432 of TLS configuration problems is only done for the daemon startup.
434 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
436 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
437 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
440 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
441 DKIM signing is being done. Previously a transport_filter would always
442 disable CHUNKING, falling back to traditional DATA.
444 JH/05 Regard command-line recipients as tainted.
446 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
448 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
449 PAM library frees one of the arguments given to it, despite the
450 documentation. Therefore a plain malloc must be used.
452 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
453 on-stack buffers were used, resulting in a taint trap when DSN information
454 copied from a received message was written into the buffer.
456 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
457 the ordering of its ARC headers. This caused a crash.
459 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
460 a new record was being constructed with information from the peer, a trap
463 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
464 installation would get error messages from DMARC verify, when it hit the
465 nonexistent file indicated by the default. Distros wanting DMARC enabled
466 should both provide the file and set the option.
467 Also enforce no DMARC verification for command-line sourced messages.
469 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
470 could, depending on the platform, hang at the STARTTLS response.
472 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
473 message on a connection. Previously if one message had ARC headers and
474 the following one did not, a crash could result when adding an
475 Authentication-Results: header.
477 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
478 Exim main code and Exim-related utities. The introduction of taint
479 tracking also did many adjustments to string handling. Since then, eximon
480 frequently terminated with an assert failure.
482 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
483 check for 452 responses. This slightly helps the inefficieny of doing
484 a large alias-expansion into a recipient-limited target. The max_rcpt
485 transport option still applies (and at the current default, will override
486 the new feature). The check is done for either cause of synch, and forces
487 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
488 connection. The new facility is not tunable at this time.
490 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
491 library live data was being used, so the results became garbage. Make
492 copies while it is still usable.
494 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
495 on delivery deferred (==) and failed (**) lines (if a delivery was
496 attemtped). Previously it was only on completion (=>) lines.
498 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
499 for the expansion of the server_scram_iter and server_scram_salt options.
501 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
502 are now specifically given a NO_DATA response without hitting the system
503 resolver. The library goes on to do the now-standard TXT lookup.
504 Use of dnsdb lookups is not affected.
506 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
507 only retrieve the errormessage once. Previously two calls to dlerror()
508 were used, and the second one (for mainlog/paniclog) retrieved null
511 JH/20 Taint checking: disallow use of tainted data for
512 - the appendfile transport file and directory options
513 - the pipe transport command
514 - the autoreply transport file, log and once options
515 - file names used by the redirect router (including filter files)
517 - paths used by single-key lookups
518 Previously this was permitted.
520 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
521 adjusted the size of a major service buffer; this failed because the
522 buffer was in use at the time. Change to a compile-time increase in the
523 buffer size, when this authenticator is compiled into exim.
525 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
526 previous fast-mode was untenable in the face of glibs using mmap to
527 support larger malloc requests.
529 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
530 New values supported, if defined on system where compiled:
531 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
532 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
534 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
535 running a limited number of proceses in parallel, a benefit is gained. The
536 amount varies with the platform hardware and load. The use of the option
537 queue_run_in_order means we cannot do this, as ordering becomes
540 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
541 had introduced a string-copy (for ensuring NUL-termination) which was not
542 appropriate for that case, which can include embedded NUL bytes in the
543 block of data. Investigation showed the copy to actually be needless, the
544 data being length-specified.
546 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
547 done during a receiving connection, and both used TLS, global info was
548 used rather than per-connection info for tracking the state of data
549 queued for transmission. This could result in a connection hang.
551 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
552 Previously, when delivering serveral messages down a single connection
553 only the first would provide a SIZE. This was due to the size information
554 not being properly tracked.
556 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
557 TAI (at 37 seconds currently), pretend to be in UTC for time-related
558 expansion and logging. Previously, spurious values such as a future
559 minute could be seen.
561 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
562 it could crash from a null-deref. This could also affect the
563 ${addresses: } operator and ${readsock } item.
565 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
566 message following a mime one, the variable was not reset.
568 JH/30 When an pipelined-connect fails at the first response, assume incorrect
569 cached capability (perhaps the peer reneged?) and immediately retry in
572 JH/31 Fix spurious detection of timeout while writing to transport filter.
574 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
575 an attempt to copy the string was made before checking it.
577 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
578 taint of the lookup key was maintained; we now regard the presence in the
579 filesystem as sufficient validation.
581 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
582 argument is supplied.
584 JH/35 The dsearch lookup now requires that the directory is an absolute path.
585 Previously this was not checked, and nonempty relative paths made an
586 access under Exim's current working directory.
588 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
589 Previously no event was raised.
591 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
592 parameter supplied by the sender MAIL FROM command. Previously it was
593 ignored, and only the check_spool_space option value for the required
596 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
597 the size of the signing public-key. Previously it was instead giving
598 the size of the signature hash.
600 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
601 the default. See the (new) dkim_verify_min_keysizes option.
603 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
604 and an ACL use a lookup for checking either the local_part or domain,
605 stale data could be accessed. Ensure that variable references are
606 dropped between messages.
608 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
609 by the client was not checked as pointing within response data before
610 being used. A malicious client could thus cause an out-of-bounds read and
611 possibly gain authentication. Fix by adding the check.
613 JH/42 Internationalisation: change the default for downconversion in the smtp
614 transport to be "if needed". Previously it was "as previously set" for
615 the message, which usually meant "if needed" for message-submission but
616 "no" for everything else. However, MTAs have been seen using SMTPUTF8
617 even when the envelope addresses did not need it, resulting in forwarding
618 failures to non-supporting MTAs. A downconvert in such cases will be
619 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
620 transport. The change does mean that addresses needing conversion will
621 be converted when previously a delivery failure would occur.
623 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
624 response was received it would be used unchecked in a fail-DSN, violating
625 standards on line-length limits. Truncate if needed.
627 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
628 doesn't work. (Additionally add a new main config option to configure the
635 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
636 side, to decode a TLS 1.3 packet capture.
638 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
639 Previously the default library behaviour applied, sending two, each in
642 JH/03 Debug output for ACL now gives the config file name and line number for
645 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
647 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
649 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
650 buffer overrun for (non-chunking) other transports.
652 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
653 TLS1.3, means that a server rejecting a client certificate is not visible
654 to the client until the first read of encrypted data (typically the
655 response to EHLO). Add detection for that case and treat it as a failed
656 TLS connection attempt, so that the normal retry-in-clear can work (if
657 suitably configured).
659 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
660 and/or domain. Found and fixed by Jason Betts.
662 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
663 configuration). If a CNAME target was not a wellformed name pattern, a
666 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
667 the OS reports them interleaved with other addresses.
669 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
670 used both for input and for a verify callout, both encrypted, SMTP
671 responses being sent by the server could be lost. This resulted in
672 dropped connections and sometimes bounces generated by a peer sending
675 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
676 string. Previously it was possible to cause undefined behaviour in a
677 library routine (usually a crash). Found by "zerons".
679 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
682 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
683 API was removed, so update to use the newer ones.
685 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
686 any timeout set, is taking a long time. Previously we would hang on to a
687 rotated logfile "forever" if the input was arriving with long gaps
688 (a previous attempt to fix addressed lack, for a long time, of initial
691 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
692 shared (NFS) environment. The length of the tempfile name is now
693 4 + 16 ("hdr.$message_exim_id") which might break on file
694 systems which restrict the file name length to lower values.
697 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
698 shared (NFS) environment.
700 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
701 did for all versions <4.90). Notably -M, -m, --invert, -I may be
704 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
705 on some platforms for bit 31.
707 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
708 to changes apparently associated with TLS1.3 handling some of the APIs
709 previously used were either nonfunctional or inappropriate. Strings
710 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
711 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
712 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
713 This affects log line X= elements, the $tls_{in,out}_cipher variables,
714 and the use of specific cipher names in the encrypted= ACL condition.
716 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
718 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
719 verification result was not updated unless hosts_require_ocsp applied.
721 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
722 queue_list_requires_admin set to false, non-admin users were denied the
725 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
726 directory-of-certs mode. Previously they were advertised despite the
729 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
730 A single TCP connection by a client will now hold a TLS connection open
731 for multiple message deliveries, by default. Previously the default was to
734 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
735 default. If built with the facility, DANE will be used. The facility
736 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
738 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
739 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
740 must be defined and you must still, unless you define DISABLE_TLS, manage
741 the the include-dir and library-file requirements that go with that
742 choice. Non-TLS builds are still supported.
744 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
747 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
748 default. If the platform supports and has the facility enabled, it will
749 be requested on all coneections.
751 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
752 controlled by the build-time option SUPPORT_PIPE_CONNECT.
754 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
756 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
757 success-DSN messages. Previously the From: header was always the default
758 one for these; the option was ignored.
760 JH/28 Fix the timeout on smtp response to apply to the whole response.
761 Previously it was reset for every read, so a teergrubing peer sending
762 single bytes within the time limit could extend the connection for a
763 long time. Credit to Qualsys Security Advisory Team for the discovery.
765 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
766 delivery address, which leaked information of the results of local
767 forwarding. Change to the original envelope recipient address, per
770 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
771 requested. Previously not bounce was generated and a log entry of
772 error ignored was made.
774 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
776 JH/32 Introduce a general tainting mechanism for values read from the input
777 channel, and values derived from them. Refuse to expand any tainted
778 values, to catch one form of exploit.
780 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
781 was unused and the unexpanded text used for the test. Found and
782 fixed by Ruben Jenster.
784 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
785 an attempt to use a TLS library read routine dereffed a nul pointer,
788 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
789 them in our smtp response.
791 JH/36 Have the general router option retry_use_local_part default to true when
792 any of the restrictive preconditions are set (to anything). Previously it
793 was only for check_local user. The change removes one item of manual
794 configuration which is required for proper retries when a remote router
795 handles a subset of addresses for a domain.
797 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
798 link count into consideration.
800 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
801 caused the extension of big_buffer, the following lines were ignored.
803 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
804 accordance with RFC 2308. Previously there was no expiry, so a longlived
805 receive process (eg. due to ACL delays) versus a short SOA value could
808 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
810 JH/39 Promote DMARC support to mainline.
812 JH/40 Bug 2452: Add a References: header to DSNs.
814 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
815 parameters. The relevant library call is documented as "Deprecated: This
816 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
817 3.6.0, DH parameters are negotiated following RFC7919."
819 HS/06 Change the default of dnssec_request_domains to "*"
821 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
822 carried on and emitted a BDAT command, even when PIPELINING was not
825 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
826 buffer was used for the filename, resulting in a trap when tainted
827 arguments (eg. $domain) were used.
829 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
830 recommended to avoid a possible server-load attack. The feature can be
831 re-enabled via the openssl_options main cofiguration option.
833 JH/45 local_scan API: documented the current smtp_printf() call. This changed
834 for version 4.90 - adding a "more data" boolean to the arguments.
835 Bumped the ABI version number also, this having been missed previously;
836 release versions 4.90 to 4.92.3 inclusive were effectively broken in
837 respect of usage of smtp_printf() by either local_scan code or libraries
838 accessed via the ${dlfunc } expansion item. Both will need coding
839 adjustment for any calls to smtp_printf() to match the new function
840 signature; a FALSE value for the new argument is always safe.
842 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
843 the file-offset (which the Linux syscall does, and exim expects); this
844 resulted in an indefinite loop.
846 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
847 to do ARC verification. The Authentication-Results: header line added
848 by the configuration then had no ARC item.
850 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
851 an incorrect mode was used for reading the result, resulting in it being
858 JH/01 Remove code calling the customisable local_scan function, unless a new
859 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
861 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
862 non-signal-safe functions being used.
864 JH/03 Bug 2269: When presented with a received message having a stupidly large
865 number of DKIM-Signature headers, disable DKIM verification to avoid
866 a resource-consumption attack. The limit is set at twenty.
868 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
869 report of oldest_pass in ${authres } in consequence, and separate out
870 some descriptions of reasons for verification fail.
872 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
873 files in the spool were present and unlocked. A queue-runner could spot
874 them, resulting in a duplicate delivery. Fix that by doing the unlock
875 after the unlink. Investigation by Tim Stewart. Take the opportunity to
876 add more error-checking on spoolfile handling while that code is being
879 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
880 No known attacks, no CVE, this is defensive hardening.
882 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
883 a queue-runner could start a delivery while other operations were ongoing.
884 Cutthrough delivery was a common victim, resulting in duplicate delivery.
885 Found and investigated by Tim Stewart. Fix by using the open message data
886 file handle rather than opening another, and not locally closing it (which
887 releases a lock) for that case, while creating the temporary .eml format
888 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
890 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
891 $sender_verify_failure/$recipient_verify_failure to "random".
893 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
896 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
897 Previously this would segfault.
899 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
902 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
903 like zero, since the resolver should be doing this for us, But we need one
904 as a CNAME but no MX presence gets the CNAME returned; we need to check
905 that doesn't point to an MX to declare it "no MX returned" rather than
906 "error, loop". A new main option is added so the older capability of
907 following some limited number of chain links is maintained.
909 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
911 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
912 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
913 now-deprecated earlier definitions used only specified the range up to TLS
914 1.2 (in the older-version library docs).
916 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
918 JH/15 Rework TLS client-side context management. Stop using a global, and
919 explicitly pass a context around. This enables future use of TLS for
920 connections to service-daemons (eg. malware scanning) while a client smtp
921 connection is using TLS; with cutthrough connections this is quite likely.
923 JH/16 Fix ARC verification to do AS checks in reverse order.
925 JH/17 Support a "tls" option on the ${readsocket } expansion item.
927 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
928 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
929 Previously the "utf8" would be re-prepended for every additional message.
931 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
932 Previously thery were accepted, resulting in issues when attempting to
933 forward messages to a non-supporting MTA.
935 PP/02 Let -n work with printing macros too, not just options.
937 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
938 one parent address was copied, and bogus data was used at delivery-logging
939 time. Either a crash (after delivery) or bogus log data could result.
940 Discovery and analysis by Tim Stewart.
942 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
943 Previously if the string ended mid-character, we did not insert the
944 promised '?' replacement.
946 PP/04 Documentation: current string operators work on bytes, not codepoints.
948 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
949 should pack well giving a smaller memory footprint so better caching and
950 therefore performance. Group the declarations where this can't be done so
951 that the byte-sized flag variables are not interspersed among pointer
952 variables, giving a better chance of good packing by the compiler.
954 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
955 non-null, to avoid issues with sites running BATV. Previously reports were
956 sent with an empty envelope sender so looked like bounces.
958 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
959 The ignore_error flag wasn't being returned from the filter subprocess so
960 was not set for later routers. Investigation and fix by Matthias Kurz.
962 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
963 and a msg:complete for the whole, when a message is manually removed using
964 -Mrm. Developement by Matthias Kurz, hacked on by JH.
966 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
967 a "Gnu special" function, asprintf() in the DB utility binary builds; I
968 hope that is portable enough.
970 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
971 requiring a known-CA anchor certificate; make it now rely entirely on the
972 TLSA as an anchor. Checking the name on the leaf cert against the name
973 on the A-record for the host is still done for TA (but not for EE mode).
975 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
976 included in delivery lines for non-proxied connections, when compiled with
977 SUPPORT_SOCKS and running with proxy logging enabled.
979 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
980 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
981 move the existing event to fire before the normal logging of message
982 failure so that custom logging is bracketed by normal logging.
984 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
985 msg:fail:internal event. Developement by Matthias Kurz.
987 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
988 far too small for todays use of crypto signatures stored there. Go all
989 the way to the max DNS message size of 64kB, even though this might be
990 overmuch for IOT constrained device use.
992 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
993 copy a string over itself. The library routine is documented as not
994 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
996 JH/32 For main options check_spool_space and check_inode_space, where the
997 platform supports 64b integers, support more than the previous 2^31 kB
998 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
999 the previous G, M, k.
1001 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1002 $authenticated_fail_id variable on authentication failure. Previously
1005 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1006 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1007 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1008 GNUTLS_SEC_PARAM_MEDIUM.
1010 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1011 side. Previously we would continue as if no SNI had been received.
1013 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1014 "<" character, which should be a list-separator specification, we walked
1015 off past the nul-terimation.
1017 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1018 causes) even when the retry time is not yet met. Previously they were
1019 not, meaning that when (say) an account was over-quota and temp-rejecting,
1020 and multiple senders' messages were queued, only one sender would get
1021 notified on each configured delay_warning cycle.
1023 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1025 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1026 cipher-suites, an error can be left on the stack even for a succeeding
1027 accept; this results in impossible error messages when a later operation
1030 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1031 return error codes indicating retry. Under TLS1.3 this becomes required.
1033 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1034 it only wrote the new authenticators, resulting in a lack of tracking of
1035 peer changes of ESMTP extensions until the next cache flush.
1037 JH/41 Fix the loop reading a message header line to check for integer overflow,
1038 and more-often against header_maxsize. Previously a crafted message could
1039 induce a crash of the recive process; now the message is cleanly rejected.
1041 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1042 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1048 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1049 When redis_servers is set to a list of > 1 element, and the Redis servers
1050 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1051 case of MOVED into a DEFER case instead, thus moving the query onto the
1052 next server in the list. For a cluster of N elements, all N servers must
1053 be defined in redis_servers.
1055 GF/02 Catch and remove uninitialized value warning in exiqsumm
1056 Check for existence of @ARGV before looking at $ARGV[0]
1058 JH/01 Replace the store_release() internal interface with store_newblock(),
1059 which internalises the check required to safely use the old one, plus
1060 the allocate and data copy operations duplicated in both (!) of the
1061 extant use locations.
1063 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1064 modifier. This matches the restriction on the commandline.
1066 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1067 Previously only the last row was returned.
1069 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1070 we assumed that tags in the header were well-formed, and parsed the
1071 element content after inspecting only the first char of the tag.
1072 Assumptions at that stage could crash the receive process on malformed
1075 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1076 While running the DKIM ACL we operate on the Permanent memory pool so that
1077 variables created with "set" persist to the DATA ACL. Also (at any time)
1078 DNS lookups that fail create cache records using the Permanent pool. But
1079 expansions release any allocations made on the current pool - so a dnsdb
1080 lookup expansion done in the DKIM ACL releases the memory used for the
1081 DNS negative-cache, and bad things result. Solution is to switch to the
1082 Main pool for expansions.
1083 While we're in that code, add checks on the DNS cache during store_reset,
1084 active in the testsuite.
1085 Problem spotted, and debugging aided, by Wolfgang Breyha.
1087 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1088 When none of the hosts presented to a transport match an already-open
1089 connection, close it and proceed with the list. Previously we would
1090 queue the message. Spotted by Lena with Yahoo, probably involving
1093 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1094 Previously a spurious "250 OK id=" response was appended to the proper
1097 JH/08 The "support for" informational output now, which built with Content
1098 Scanning support, has a line for the malware scanner interfaces compiled
1099 in. Interface can be individually included or not at build time.
1101 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1102 by the template makefile "src/EDITME". The "STREAM" support for an older
1103 ClamAV interface method is removed.
1105 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1106 rows affected is given instead).
1108 JH/11 The runtime Berkeley DB library version is now additionally output by
1109 "exim -d -bV". Previously only the compile-time version was shown.
1111 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1112 SMTP connection. Previously, when one had more recipients than the
1113 first, an abortive onward connection was made. Move to full support for
1114 multiple onward connections in sequence, handling cutthrough connection
1115 for all multi-message initiating connections.
1117 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1118 routers. Previously, a multi-recipient message would fail to match the
1119 onward-connection opened for the first recipient, and cause its closure.
1121 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1122 a timeout on read on a GnuTLS initiating connection, resulting in the
1123 initiating connection being dropped. This mattered most when the callout
1124 was marked defer_ok. Fix to keep the two timeout-detection methods
1127 JH/15 Relax results from ACL control request to enable cutthrough, in
1128 unsupported situations, from error to silently (except under debug)
1129 ignoring. This covers use with PRDR, frozen messages, queue-only and
1132 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1134 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1135 metadata, resulting in a crash in free().
1137 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1138 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1139 Broken also in d185889f4, with init system revamp.
1141 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1142 Previously we abruptly closed the connection after reading a malware-
1143 found indication; now we go on to read the "scan ok" response line,
1146 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1147 ACL. Previously, a crash would result.
1149 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1150 macro text after a replacement (previously it was only once per line) and
1151 by skipping builtin macros when searching for an uppercase lead character.
1153 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1154 for the build is renamed.
1156 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1157 was allocated for every new TLS startup, meaning one per message. Fix
1158 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1160 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1161 reported the original. Fix to report (as far as possible) the ACL
1162 result replacing the original.
1164 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1165 OpenSSL. Certificate information is loaded for every new TLS startup,
1166 and the resources needed to be freed.
1168 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1170 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1173 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1174 DATA response info to the (existing) per-recipient response info for
1175 the "C=" log element. It can have useful tracking info from the
1176 destination system. Patch from Simon Arlott.
1178 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1179 length value. Previously this would segfault.
1181 HS/02 Support Avast multiline protoocol, this allows passing flags to
1182 newer versions of the scanner.
1184 JH/28 Ensure that variables possibly set during message acceptance are marked
1185 dead before release of memory in the daemon loop. This stops complaints
1186 about them when the debug_store option is enabled. Discovered specifically
1187 for sender_rate_period, but applies to a whole set of variables.
1188 Do the same for the queue-runner and queue-list loops, for variables set
1189 from spool message files. Do the same for the SMTP per-message loop, for
1190 certain variables indirectly set in ACL operations.
1192 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1193 as a multi-recipient message from a mailinglist manager). The coding had
1194 an arbitrary cutoff number of characters while checking for more input;
1195 enforced by writing a NUL into the buffer. This corrupted long / fast
1196 input. The problem was exposed more widely when more pipelineing of SMTP
1197 responses was introduced, and one Exim system was feeding another.
1198 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1199 receiving system, and refused recipients seen by the sending system
1200 (propating to people being dropped from mailing lists).
1201 Discovered and pinpointed by David Carter.
1203 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1204 replaced by the ${authresults } expansion.
1206 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1208 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1209 allows proper process termination in container environments.
1211 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1212 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1214 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1215 and "err_perm", deprecated since 4.83 when the RFC-defined words
1216 "temperror" and "permerror" were introduced.
1218 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1219 transport-filters or DKIM-signing. The restriction was lost in the
1220 consolidation of verify-callout and delivery SMTP handling.
1221 Extend the restriction to also cover ARC-signing.
1223 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1224 in defer=pass mode supply a 450 to the initiator. Previously the message
1227 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1228 tls_require_ciphers is used as before.
1230 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1231 "pass_unscanned". Only tmpfails from the scanner are written to
1232 the paniclog, as they may require admin intervention (permission
1233 denied, license issues). Other scanner errors (like decompression
1234 bombs) do not cause a paniclog entry.
1236 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1237 Previously it was possible to log spurious information in receive log
1240 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1241 triggered odd behaviour from Outlook Express clients.
1243 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1246 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1247 since the IETF WG has not yet settled on that versus the original
1248 "bare" representation.
1250 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1251 Previously the millisecond value corrupted the output.
1252 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1253 corrupted the output.
1259 JH/01 Rework error string handling in TLS interface so that the caller in
1260 more cases is responsible for logging. This permits library-sourced
1261 string to be attached to addresses during delivery, and collapses
1262 pairs of long lines into single ones.
1264 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1265 during configuration. Wildcards are allowed and expanded.
1267 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1268 This permits better logging.
1270 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1271 do not close it down and have the child transport start it up again on
1272 the passed-on TCP connection. Instead, proxy the child (and any
1273 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1274 affected: the continued delivery log lines do not have any DNSSEC, TLS
1275 Certificate or OCSP information. TLS cipher information is still logged.
1277 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1278 identical IP addresses on different listening ports. Will also affect
1281 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1282 add noisy ifdef guards to special-case this sillyness.
1283 Patch from Bernd Kuhls.
1285 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1286 than 255 are no longer allowed.
1288 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1289 Disable the session-cache too, which might reduce our load. Since we
1290 currrectly use a new context for every connection, both as server and
1291 client, there is no benefit for these.
1292 GnuTLS appears to not support tickets server-side by default (we don't
1293 call gnutls_session_ticket_enable_server()) but client side is enabled
1294 by default on recent versions (3.1.3 +) unless the PFS priority string
1297 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1298 <https://reproducible-builds.org/specs/source-date-epoch/>.
1300 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1301 the check for any unsuccessful recipients did not notice the limit, and
1302 erroneously found still-pending ones.
1304 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1305 MSG_MORE. Only in-clear (not on TLS connections).
1307 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1308 transport-filter is involved we can buffer the headers in memory for
1309 creating the signature, and read the spool data file once for the
1310 signature and again for transmission.
1312 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1313 4.77 as the kernel support then wasn't solid, having issues in 64bit
1314 mode. Now, it's been long enough. Add support for FreeBSD also.
1316 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1317 case where the routing stage had gathered several addresses to send to
1318 a host before calling the transport for the first, we previously failed
1319 to close down TLS in the old transport process before passing the TCP
1320 connection to the new process. The new one sent a STARTTLS command
1321 which naturally failed, giving a failed delivery and bloating the retry
1322 database. Investigation and fix prototype from Wolfgang Breyha.
1324 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1325 false-negatives in the check that the sender had not preempted a response
1326 or prompt from Exim (running as a server), due to that code's lack of
1327 awareness of the SMTP input buffering.
1329 PP/04 Add commandline_checks_require_admin option.
1330 Exim drops privileges sanely, various checks such as -be aren't a
1331 security problem, as long as you trust local users with access to their
1332 own account. When invoked by services which pass untrusted data to
1333 Exim, this might be an issue. Set this option in main configuration
1334 AND make fixes to the calling application, such as using `--` to stop
1337 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1338 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1339 responses to those, into a single TLS record each way (this usually means
1340 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1343 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1344 affects you only if you're dancing at the edge of the param size limits.
1345 If you are, and this message makes sense to you, then: raise the
1346 configured limit or use OpenSSL 1.1. Nothing we can do for older
1349 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1350 cmdline element to get the documented default one. Previously it was
1353 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1354 are unsuable for non-message-specific reasons - eg. connection timeout,
1355 banner-time rejection.
1357 JH/16 Fix logging of delivery remote port, when specified by router, under
1360 PP/06 Repair manualroute's ability to take options in any order, even if one
1361 is the name of a transport.
1364 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1366 JH/17 Change the list-building routines interface to use the expanding-string
1367 triplet model, for better allocation and copying behaviour.
1369 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1370 Previously it was constructed the first time a possibly-matching string
1371 was met in the configuration file input during startup; now it is done
1374 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1375 compatible one, to avoid the (poorly documented) possibility of a config
1376 file in the working directory redirecting the DB files, possibly correpting
1377 some existing file. CVE-2017-10140 assigned for BDB.
1379 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1380 cache-hot. Previously, although the result was properly cached, the
1381 initial verify call returned a defer.
1383 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1384 the main verify for receipient in uncached-mode.
1386 JH/22 Retire historical build files to an "unsupported" subdir. These are
1387 defined as "ones for which we have no current evidence of testing".
1389 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1390 if present. Previously it was ignored.
1392 JH/24 Start using specified-initialisers in C structure init coding. This is
1393 a C99 feature (it's 2017, so now considered safe).
1395 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1396 if was a fixed-sized field and bitmask ops via macros; it is now more
1399 PP/07 GitHub PR 56: Apply MariaDB build fix.
1400 Patch provided by Jaroslav Škarvada.
1402 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1403 during Coverity cleanups [4.87 JH/47]
1404 Diagnosis and fix provided by Michael Fischer v. Mollard.
1406 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1407 the right size to place the terminating semicolon on its own folded
1408 line, the header hash was calculated to an incorrect value thanks to
1409 the (relaxed) space the fold became.
1411 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1412 and confused the parent.
1414 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1415 which could crash as a result. This could lead to undeliverable messages.
1417 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1420 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1421 matters on fast-turnover and PID-randomising systems, which were getting
1422 out-of-order delivery.
1424 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1425 a possibly-overlapping copy. The symptom was that "Remote host closed
1426 connection in response to HELO" was logged instead of the actual 4xx
1429 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1430 Previously only that bufferd was discarded, resulting in SYMTP command
1433 JH/32 DKIM: when a message has multiple signatures matching an identity given
1434 in dkim_verify_signers, run the dkim acl once for each. Previously only
1435 one run was done. Bug 2189.
1437 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1438 "panic the current process" to "deliberately defer". The panic log is
1439 still written with the problem list name; the mail and reject logs now
1440 get a temp-reject line for the message that was being handled, saying
1441 something like "domains check lookup or other defer". The SMTP 451
1442 message is still "Temporary local problem".
1444 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1445 A crafted sequence of BDAT commands could result in in-use memory beeing
1446 freed. CVE-2017-16943.
1448 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1449 from SMTP input. Previously it was always done; now only done for DATA
1450 and not BDAT commands. CVE-2017-16944.
1452 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1453 to the message (such as an overlong header line). Previously this was
1454 not done and we did not exit BDAT mode. Followon from the previous item
1455 though a different problem.
1461 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1462 than -2003 did; needs libidn2 in addition to libidn.
1464 JH/02 The path option on a pipe transport is now expanded before use.
1466 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1467 Patch provided by "Björn", documentation fix added too.
1469 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1470 missing a wire-to-host endian conversion.
1472 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1473 close after a BDAT command line could be taken as a following command,
1474 giving a synch failure. Fix by only checking for synch immediately
1475 before acknowledging the chunk.
1477 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1478 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1479 macro. Patches provided by Josh Soref.
1481 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1482 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1483 by IANA as a service extension. However, John Klensin suggests that we
1486 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1487 the dkim code may be unix-mode line endings rather than smtp wire-format
1488 CRLF, so prepend a CR to any bare LF.
1490 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1491 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1493 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1494 the first were themselves being wrongly included in the feed into dkim
1495 processing; with most chunk sizes in use this resulted in an incorrect
1496 body hash calculated value.
1498 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1499 DKIM signature block, for verification. Although advised against by
1500 standards it is specifically not ruled illegal.
1502 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1504 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1505 missing a body hash (the bh= tag).
1507 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1508 It seems that HAProxy sends the Proxy Protocol information in clear and
1509 only then does a TLS startup, so do the same.
1511 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1512 TCP connections (such as for Spamd) unless the daemon successfully set
1513 Fast Open mode on its listening sockets. This fixes breakage seen on
1514 too-old kernels or those not configured for Fast Open, at the cost of
1515 requiring both directions being enabled for TFO, and TFO never being used
1516 by non-daemon-related Exim processes.
1518 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1519 endings, at least on the first header line. Try to canonify any that get
1520 past that check, despite the cost.
1522 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1523 now limited to an arbitrary five deep, while parsing addresses with the
1524 strip_excess_angle_brackets option enabled.
1526 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1527 instead leave the unprompted TLS handshake in socket buffer for the
1528 TLS library to consume.
1530 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1532 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1534 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1535 region in one operation, for speed, and this leaves assigned pointers
1536 dangling. Add checks run only under the testsuite which checks all
1537 variables at a store-reset and panics on a dangling pointer; add code
1538 explicitly nulling out all the variables discovered. Fixes one known
1539 bug: a transport crash, where a dangling pointer for $sending_ip_address
1540 originally assigned in a verify callout, is re-used.
1542 PP/06 Drop '.' from @INC in various Perl scripts.
1544 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1546 PP/08 Reduce a number of compilation warnings under clang; building with
1547 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1548 should be warning-free.
1550 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1552 HS/01 Fix portability problems introduced by PP/08 for platforms where
1553 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1555 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1556 chunk. This allows us to accept broken chunked messages. We need a more
1557 general solution here.
1559 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1560 already-broken messages in the queue.
1562 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1564 JH/19 Fix reference counting bug in routing-generated-address tracking.
1570 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1571 supports it and a size is available (ie. the sending peer gave us one).
1573 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1574 years of being deprecated). The replacements are the ACLs
1575 acl_smtp_mime and acl_not_smtp_mime.
1577 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1578 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1579 or even an in-clear connection were permitted. Now, if the host lookup
1580 was dnssec and dane was requested then the host is only used if the
1581 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1582 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1583 if one fails this test.
1584 This means that a poorly-configured remote DNS will make it incommunicado;
1585 but it protects against a DNS-interception attack on it.
1587 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1588 non-noisy when a race steals the message being considered.
1590 JH/05 If main configuration option tls_certificate is unset, generate a
1591 self-signed certificate for inbound TLS connections.
1593 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1594 in rewrites and routers.
1596 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1597 and logged a warning sing 4.83; now they are a configuration file error.
1599 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1600 (lacking @domain). Apply the same qualification processing as RCPT.
1602 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1604 JH/10 Support ${sha256:} applied to a string (as well as the previous
1607 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1608 a cutthrough deliver is pending, as we always want to make a connection.
1609 This also avoids re-routing the message when later placing the cutthrough
1610 connection after a verify cache hit.
1611 Do not update it with the verify result either.
1613 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1614 when routing results in more than one destination address.
1616 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1617 signing (which inhibits the cutthrough capability). Previously only
1618 the presence of an option was tested; now an expansion evaluating as
1619 empty is permissible (obviously it should depend only on data available
1620 when the cutthrough connection is made).
1622 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1623 the relevant preceding SMTP command did not note the pipelining mode.
1625 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1626 Previously they were not counted.
1628 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1629 as one having no matching records. Previously we deferred the message
1630 that needed the lookup.
1632 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1633 distinguished as "(=".
1635 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1636 for missing MX records. Previously it only worked for missing A records.
1638 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1640 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1641 after the data-go-ahead and data-ack. Patch from Jason Betts.
1643 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1644 even for a "none" policy. Patch from Tony Meyer.
1646 JH/22 Fix continued use of a connection for further deliveries. If a port was
1647 specified by a router, it must also match for the delivery to be
1650 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1651 When one of the recipients of a message was unsuitable for the connection
1652 (has no matching addresses), we lost track of needing to mark it
1653 deferred. As a result mail would be lost.
1655 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1657 JH/25 Decoding ACL controls is now done using a binary search; the source code
1658 takes up less space and should be simpler to maintain. Merge the ACL
1659 condition decode tables also, with similar effect.
1661 JH/26 Fix problem with one_time used on a redirect router which returned the
1662 parent address unchanged. A retry would see the parent address marked as
1663 delivered, so not attempt the (identical) child. As a result mail would
1666 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1667 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1668 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1671 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1672 Main config options check_{log,spool}_{inodes,space} are now
1673 100 inodes, 10MB unless set otherwise in the configuration.
1675 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1676 Previously it only applied to the main-section connection policy
1679 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1681 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1682 by me. Added RFC7919 DH primes as an alternative.
1684 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1685 are not in the system include path.
1687 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1688 GnuTLS, when a session startup failed (eg because the client disconnected)
1689 Exim did stdio operations after fclose. This was exposed by a recent
1690 change which nulled out the file handle after the fclose.
1692 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1693 signed directly by the cert-signing cert, rather than an intermediate
1694 OCSP-signing cert. This is the model used by LetsEncrypt.
1696 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1698 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1699 an incoming connection.
1701 HS/02 Bug 1802: Do not half-close the connection after sending a request
1704 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1705 fallback to "prime256v1".
1707 JH/34 SECURITY: Use proper copy of DATA command in error message.
1708 Could leak key material. Remotely exploitable. CVE-2016-9963.
1714 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1715 and 3.4.4 - once the server is enabled to respond to an OCSP request
1716 it does even when not requested, resulting in a stapling non-aware
1717 client dropping the TLS connection.
1719 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1720 support variable-length bit vectors. No functional change.
1722 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1723 The I= interface field on outgoing lines is now after the H= remote
1724 host field, same as incoming lines. There is a separate
1725 outgoing_interface log selector which allows you to disable the
1728 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1729 If not running log_selector +smtp_connection the mainlog would be held
1730 open indefinitely after a "too many connections" event, including to a
1731 deleted file after a log rotate. Leave the per net connection logging
1732 leaving it open for efficiency as that will be quickly detected by the
1733 check on the next write.
1735 HS/01 Bug 1671: Fix post transport crash.
1736 Processing the wait-<transport> messages could crash the delivery
1737 process if the message IDs didn't exist for some reason. When
1738 using 'split_spool_directory=yes' the construction of the spool
1739 file name failed already, exposing the same netto behaviour.
1741 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1742 mime_regex ACL conditions.
1744 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1745 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1746 to HELO, local diagnostic string.
1748 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1749 log line to debug. Even when configured with a tls authenticator many
1750 client connections are expected to not authenticate in this way, so
1751 an authenticate fail is not an error.
1753 HS/02 Add the Exim version string to the process info. This way exiwhat
1754 gives some more detail about the running daemon.
1756 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1757 matter for fast-change records such as DNSBLs.
1759 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1760 retry record, even if constant. There may be multiple transports with
1761 different interface settings and the retry behaviour needs to be kept
1764 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1766 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1768 JH/10 Bug 840: fix log_defer_output option of pipe transport
1770 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1771 hosts_randomize. This matters a lot when combined with mua_wrapper.
1773 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1774 ${quote_pgsql:<string>} operator.
1776 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1778 JH/14 Change header file name for freeradius-client. Relevant if compiling
1779 with Radius support; from the Gentoo tree and checked under Fedora.
1781 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1783 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1784 option transport_filter, meaning no filtering.
1786 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1788 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1789 defaults to "*" (all hosts). The variable is now available when not built
1790 with TLS, default unset, mainly to enable keeping the testsuite sane.
1791 If a server certificate is not supplied (via tls_certificate) an error is
1792 logged, and clients will find TLS connections fail on startup. Presumably
1793 they will retry in-clear.
1794 Packagers of Exim are strongly encouraged to create a server certificate
1795 at installation time.
1797 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1798 with the $config_file variable.
1800 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1801 in transport context, after the attempt, and per-recipient. The latter type
1802 is per host attempted. The event data is the error message, and the errno
1803 information encodes the lookup type (A vs. MX) used for the (first) host,
1804 and the trailing two digits of the smtp 4xx response.
1806 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1807 to write to mainlog (or rejectlog, paniclog) in the window between file
1808 creation and permissions/ownership being changed. Particularly affects
1809 installations where exicyclog is run as root, rather than exim user;
1810 result is that the running daemon panics and dies.
1812 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1814 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1815 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1816 "pri" and "weight". Note that the previous implicit priority given by the
1817 list order is no longer honoured.
1819 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1820 for DKIM processing.
1822 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1823 by defining SUPPORT_SOCKS.
1825 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1826 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1827 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1828 variables are renamed to proxy_{local,external}_{address,port}.
1830 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1831 for a build by defining SUPPORT_I18N
1833 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1834 of the query string, and make ${quote_redis:} do that quoting.
1836 JH/29 Move Events support from Experimental to mainline, enabled by default
1837 and removable for a build by defining DISABLE_EVENT.
1839 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1841 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1842 cached by the daemon.
1844 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1845 by defining LOOKUP_REDIS. The libhiredis library is required.
1847 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1848 keys are given for lookup.
1850 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1851 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1852 only supported when built with TLS support. The PolarSSL SHA routines
1853 are still used when the TLS library is too old for convenient support.
1855 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1856 openssl_options), for security. OpenSSL forces this from version 1.1.0
1857 server-side so match that on older versions.
1859 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1860 allocation for $value could be released as the expansion processing
1861 concluded, but leaving the global pointer active for it.
1863 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1864 and to use the domains and local_parts ACL conditions.
1866 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1867 incorrectly not doubled on cutthrough transmission, hence seen as a
1868 body-termination at the receiving system - resulting in truncated mails.
1869 Commonly the sender saw a TCP-level error, and retransmitted the message
1870 via the normal store-and-forward channel. This could result in duplicates
1871 received - but deduplicating mailstores were liable to retain only the
1872 initial truncated version.
1874 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1876 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1878 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1879 we're in there, support oversigning also; bug 1309.
1881 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1883 HS/04 Add support for keep_environment and add_environment options.
1885 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1886 either intentional arithmetic overflow during PRNG, or testing config-
1889 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1890 delivery resulted in actual delivery. Cancel cutthrough before DATA
1893 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1894 rejects a recipient: pass the reject to the originator.
1896 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1897 Many were false-positives and ignorable, but it's worth fixing the
1900 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1901 for the new environment-manipulation done at startup. Move the routines
1902 from being local to tls.c to being global via the os.c file.
1904 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1905 an extract embedded as result-arg for a map, the first arg for extract
1906 is unavailable so we cannot tell if this is a numbered or keyed
1907 extraction. Accept either.
1913 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1916 JH/02 The smtp transport option "multi_domain" is now expanded.
1918 JH/03 The smtp transport now requests PRDR by default, if the server offers
1921 JH/04 Certificate name checking on server certificates, when exim is a client,
1922 is now done by default. The transport option tls_verify_cert_hostnames
1923 can be used to disable this per-host. The build option
1924 EXPERIMENTAL_CERTNAMES is withdrawn.
1926 JH/05 The value of the tls_verify_certificates smtp transport and main options
1927 default to the word "system" to access the system default CA bundle.
1928 For GnuTLS, only version 3.0.20 or later.
1930 JH/06 Verification of the server certificate for a TLS connection is now tried
1931 (but not required) by default. The verification status is now logged by
1932 default, for both outbound TLS and client-certificate supplying inbound
1935 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1938 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1939 Status Notification (bounce) messages are now MIME format per RFC 3464.
1940 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1941 under the control of the dsn_advertise_hosts option, and routers may
1942 have a dsn_lasthop option.
1944 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1945 default, modifiable by a malware= option. The list separator for
1946 the options can now be changed in the usual way. Bug 68.
1948 JH/10 The smtp_receive_timeout main option is now expanded before use.
1950 JH/11 The incoming_interface log option now also enables logging of the
1951 local interface on delivery outgoing connections.
1953 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1954 if the interface and destination host and port all match.
1956 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1959 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1960 Patch from Andrew Lewis.
1962 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1963 now supports optional time-restrictions, weighting, and priority
1964 modifiers per server. Patch originally by <rommer@active.by>.
1966 JH/16 The spamd_address main option now supports a mixed list of local
1967 and remote servers. Remote servers can be IPv6 addresses, and
1968 specify a port-range.
1970 JH/17 Bug 68: The spamd_address main option now supports an optional
1971 timeout value per server.
1973 JH/18 Bug 1581: Router and transport options headers_add/remove can
1974 now have the list separator specified.
1976 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1979 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1982 JH/21 Support for the A6 type of dns record is withdrawn.
1984 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1985 rather than the verbs used.
1987 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1988 from 255 to 1024 chars.
1990 JH/24 Verification callouts now attempt to use TLS by default.
1992 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1993 are generic router options now. The defaults didn't change.
1995 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1996 Original patch from Alexander Shikoff, worked over by JH.
1998 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1999 files if ZCAT_COMMAND is not executable.
2001 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2003 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2005 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2006 Normally benign, it bites when the pair was led to by a CNAME;
2007 modern usage is to not canonicalize the domain to a CNAME target
2008 (and we were inconsistent anyway for A-only vs AAAA+A).
2010 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2012 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2013 when evaluating $sender_host_dnssec.
2015 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2016 $sender_helo_dnssec variable.
2018 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2020 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2022 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2024 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2025 documented as working, but never had. Support all but $spam_report.
2027 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2028 added for tls authenticator.
2030 HS/03 Add perl_taintmode main config option
2036 TL/01 When running the test suite, the README says that variables such as
2037 no_msglog_check are global and can be placed anywhere in a specific
2038 test's script, however it was observed that placement needed to be near
2039 the beginning for it to behave that way. Changed the runtest perl
2040 script to read through the entire script once to detect and set these
2041 variables, reset to the beginning of the script, and then run through
2042 the script parsing/test process like normal.
2044 TL/02 The BSD's have an arc4random API. One of the functions to induce
2045 adding randomness was arc4random_stir(), but it has been removed in
2046 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2047 function when detected.
2049 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2050 cause callback expansion.
2052 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2053 syntax errors in an expansion can be treated as a string instead of
2054 logging or causing an error, due to the internal use of bool_lax
2055 instead of bool when processing it.
2057 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2058 server certificates when making smtp deliveries.
2060 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2062 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2064 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2066 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2067 Merged patch from Sebastian Wiedenroth.
2069 JH/05 Fix results-pipe from transport process. Several recipients, combined
2070 with certificate use, exposed issues where response data items split
2071 over buffer boundaries were not parsed properly. This eventually
2072 resulted in duplicates being sent. This issue only became common enough
2073 to notice due to the introduction of connection certificate information,
2074 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2076 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2077 size buffer was used, resulting in syntax errors when an expansion
2080 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2081 version 3.3.6 or later.
2083 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2084 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2085 both become "event_action", the variables become $event_name, $event_data
2086 and $event_defer_errno. There is a new variable $verify_mode, usable in
2087 routers, transports and related events. The tls:cert event is now also
2088 raised for inbound connections, if the main configuration event_action
2091 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2092 early OCSP support, but no stapling (appears to be less than 1.0.0).
2094 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2095 server certificate names available under the smtp transport option
2096 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2099 JH/10 Time-related extraction expansions from certificates now use the main
2100 option "timezone" setting for output formatting, and are consistent
2101 between OpenSSL and GnuTLS compilations. Bug 1541.
2103 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2104 encoded parameter in the incoming message. Bug 1558.
2106 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2107 include certificate info, eximon was claiming there were spoolfile
2110 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2112 JH/14 Log delivery-related information more consistently, using the sequence
2113 "H=<name> [<ip>]" wherever possible.
2115 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2116 are problematic for Debian distribution, omit them from the release
2119 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2121 JH/16 Fix string representation of time values on 64bit time_t architectures.
2124 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2125 output list separator was used.
2130 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2131 checkers that were complaining about end of non-void function with no
2134 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2135 This was a regression introduced in 4.83 by another bugfix.
2137 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2139 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2140 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2146 TF/01 Correctly close the server side of TLS when forking for delivery.
2148 When a message was received over SMTP with TLS, Exim failed to clear up
2149 the incoming connection properly after forking off the child process to
2150 deliver the message. In some situations the subsequent outgoing
2151 delivery connection happened to have the same fd number as the incoming
2152 connection previously had. Exim would try to use TLS and fail, logging
2153 a "Bad file descriptor" error.
2155 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2156 utilities have not been installed.
2158 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2159 temporary space as the ACL may create new global variables.
2161 TL/01 LDAP support uses per connection or global context settings, depending
2162 upon the detected version of the libraries at build time.
2164 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2165 to extract and use the src ip:port in logging and expansions as if it
2166 were a direct connection from the outside internet. PPv2 support was
2167 updated based on HAProxy spec change in May 2014.
2169 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2171 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2172 Properly escape header and check for NULL return.
2174 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2177 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2179 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2180 characters in header names, implemented as a verify condition.
2181 Contributed by Michael Fischer v. Mollard.
2183 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2184 results permerror and temperror. Previous values are deprecated but
2185 still accepted. In a future release, err_perm and err_temp will be
2186 completely removed, which will be a backward incompatibility if the
2187 ACL tests for either of these two old results. Patch contributed by
2188 user bes-internal on the mailing list.
2190 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2192 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2193 selectors, in both main and reject logs.
2195 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2198 JH/07 Add malware type "sock" for talking to simple daemon.
2200 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2202 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2203 routers/transports under cutthrough routing.
2205 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2206 numbers. Touch up "bool" conditional to keep the same definition.
2208 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2210 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2212 JH/12 Expand items in router/transport headers_add or headers_remove lists
2213 individually rather than the list as a whole. Bug 1452.
2215 Required for reasonable handling of multiple headers_ options when
2216 they may be empty; requires that headers_remove items with embedded
2217 colons must have them doubled (or the list-separator changed).
2219 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2220 view the policy declared in the DMARC record. Currently, $dmarc_status
2221 is a combined value of both the record presence and the result of the
2224 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2226 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2227 dnslookup router and the smtp transport (applying to the forward
2230 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2231 of ldap servers used for a specific lookup. Patch provided by Heiko
2234 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2235 New variable $lookup_dnssec_authenticated for observability.
2237 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2238 Patch submitted by Lars Timman.
2240 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2242 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2243 Requires trusted mode and valid format message id, aborts otherwise.
2244 Patch contributed by Heiko Schlichting.
2246 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2247 certextract with support for various fields. Bug 1358.
2249 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2250 is requested by default, modifiable by smtp transport option
2253 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2254 operate on certificate variables to give certificate fingerprints
2255 Also new ${sha256:cert_variable}.
2257 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2259 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2260 Christian Aistleitner.
2262 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2264 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2265 file. Patch from Wolfgang Breyha.
2267 JH/25 Expand the coverage of the delivery $host and $host_address to
2268 client authenticators run in verify callout. Bug 1476.
2270 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2271 align with daemon_smtp_ports. Bug 72.
2273 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2274 support and error reporting did not work properly.
2276 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2277 and is readable. Patch from Andrew Colin Kissa.
2279 TL/14 Enhance documentation of ${run expansion and how it parses the
2280 commandline after expansion, particularly in the case when an
2281 unquoted variable expansion results in an empty value.
2283 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2285 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2286 by renaming away from it. Addresses GH issue 15, reported by
2289 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2291 TL/15 SECURITY: prevent double expansion in math comparison functions
2292 (can expand unsanitized data). Not remotely exploitable.
2299 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2301 PP/02 Make -n do something, by making it not do something.
2302 When combined with -bP, the name of an option is not output.
2304 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2307 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2308 $sender_host_name and config options to manage this, and basic check
2311 PP/05 DSCP support for outbound connections and control modifier for inbound.
2313 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2314 (Only plugin which currently uses this is kerberos4, which nobody should
2315 be using, but we should make it available and other future plugins might
2316 conceivably use it, even though it would break NAT; stuff *should* be
2317 using channel bindings instead).
2319 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2320 name; added for Sendmail compatibility; requires admin caller.
2321 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2322 just ignore it); requires trusted caller.
2323 Also parse but ignore: -Ac -Am -X<logfile>
2326 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2328 TL/02 Add +smtp_confirmation as a default logging option.
2330 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2331 Patch by Magnus Holmgren from 2007-02-20.
2333 TL/04 Bugzilla 1281 - Spec typo.
2334 Bugzilla 1283 - Spec typo.
2335 Bugzilla 1290 - Spec grammar fixes.
2337 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2339 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2341 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2342 mailing list by by Dmitry Isaikin.
2344 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2346 JH/02 Support "G" suffix to numbers in ${if comparisons.
2348 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2350 NM/01 Bugzilla 1197 - Spec typo
2351 Bugzilla 1196 - Spec examples corrections
2353 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2355 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2356 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2359 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2360 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2362 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2363 "acl {{name}{arg}...}", and optional args on acl condition
2366 JH/05 Permit multiple router/transport headers_add/remove lines.
2368 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2370 JH/07 Avoid using a waiting database for a single-message-only transport.
2371 Performance patch from Paul Fisher. Bugzilla 1262.
2373 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2376 JH/09 Add $headers_added variable, with content from use of ACL modifier
2377 add_header (but not yet added to the message). Bugzilla 199.
2379 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2380 Pulled from Bugzilla 817 by Wolfgang Breyha.
2382 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2384 (nb: this is the same fix as in Exim 4.80.1)
2386 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2389 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2392 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2394 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2395 advertise SMTP AUTH mechanism to us, instead of a generic
2396 protocol violation error. Also, make Exim more robust to bad
2397 data from the Dovecot auth socket.
2399 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2401 When a queue runner is handling a message, Exim first routes the
2402 recipient addresses, during which it prunes them based on the retry
2403 hints database. After that it attempts to deliver the message to
2404 any remaining recipients. It then updates the hints database using
2407 So if a recipient address works intermittently, it can get repeatedly
2408 deferred at routing time. The retry hints record remains fresh so the
2409 address never reaches the final cutoff time.
2411 This is a fairly common occurrence when a user is bumping up against
2412 their storage quota. Exim had some logic in its local delivery code
2413 to deal with this. However it did not apply to per-recipient defers
2414 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2416 This change adds a proper retry rule check during routing so that the
2417 final cutoff time is checked against the message's age. We only do
2418 this check if there is an address retry record and there is not a
2419 domain retry record; this implies that previous attempts to handle
2420 the address had the retry_use_local_parts option turned on. We use
2421 this as an approximation for the destination being like a local
2422 delivery, as in LMTP.
2424 I suspect this new check makes the old local delivery cutoff check
2425 redundant, but I have not verified this so I left the code in place.
2427 TF/02 Correct gecos expansion when From: is a prefix of the username.
2429 Test 0254 submits a message to Exim with the header
2433 When I ran the test suite under the user fanf2, Exim expanded
2434 the header to contain my full name, whereas it should have added
2435 a Resent-Sender: header. It erroneously treats any prefix of the
2436 username as equal to the username.
2438 This change corrects that bug.
2440 GF/01 DCC debug and logging tidyup
2441 Error conditions log to paniclog rather than rejectlog.
2442 Debug lines prefixed by "DCC: " to remove any ambiguity.
2444 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2446 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2447 Bug spotted by Jeremy Harris; was flawed since initial commit.
2448 Would have resulted in OCSP responses post-SNI triggering an Exim
2449 NULL dereference and crash.
2451 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2453 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2454 Bug detection, analysis and fix by Samuel Thibault.
2455 Bugzilla 1331, Debian bug #698092.
2457 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2459 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2460 Server implementation by Todd Lyons, client by JH.
2461 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2462 config variable "prdr_enable" controls whether the server
2463 advertises the facility. If the client requests PRDR a new
2464 acl_data_smtp_prdr ACL is called once for each recipient, after
2465 the body content is received and before the acl_smtp_data ACL.
2466 The client is controlled by both of: a hosts_try_prdr option
2467 on the smtp transport, and the server advertisement.
2468 Default client logging of deliveries and rejections involving
2469 PRDR are flagged with the string "PRDR".
2471 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2472 fclose(). Diagnosis by Todd Lyons.
2474 PP/17 Update configure.default to handle IPv6 localhost better.
2475 Patch by Alain Williams (plus minor tweaks).
2478 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2479 This is now consistent with GnuTLS, and is now documented: the
2480 previous undocumented portable approach to treating the option as
2481 unset was to force an expansion failure. That still works, and
2482 an empty string is now equivalent.
2484 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2485 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2486 not performing validation itself.
2488 PP/20 Added force_command boolean option to pipe transport.
2489 Patch from Nick Koston, of cPanel Inc.
2491 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2494 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2496 PP/21 Fix eximon continuous updating with timestamped log-files.
2497 Broken in a format-string cleanup in 4.80, missed when I repaired the
2498 other false fix of the same issue.
2499 Report and fix from Heiko Schlichting.
2502 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2503 Report from Prashanth Katuri.
2505 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2506 It's SecureTransport, so affects any MacOS clients which use the
2507 system-integrated TLS libraries, including email clients.
2509 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2510 using a MIME ACL for non-SMTP local injection.
2511 Report and assistance in diagnosis by Warren Baker.
2513 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2515 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2517 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2518 last id that failed so it may be referenced in subsequent ACL's.
2520 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2523 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2524 ldap library initialization, allowing self-signed CA's to be
2525 used. Also properly sets require_cert option later in code by
2526 using NULL (global ldap config) instead of ldap handle (per
2527 session). Bug diagnosis and testing by alxgomz.
2529 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2530 the src/util/ subdirectory.
2532 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2533 renamed to Transport Post Delivery Action by Jeremy Harris, as
2536 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2537 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2538 redis_servers = needs to be configured which will be used by the redis
2539 lookup. Patch from Warren Baker, of The Packet Hub.
2541 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2543 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2544 hostname or reverse DNS when processing a host list. Used suggestions
2545 from multiple comments on this bug.
2547 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2549 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2550 Missed a few lines, added it to make the runtest require no keyboard
2553 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2554 contains upper case chars. Make router use caseful_local_part.
2556 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2557 support when GnuTLS has been built with p11-kit.
2563 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2565 This, or similar/improved, will also be change PP/11 of 4.82.
2571 PP/01 Handle short writes when writing local log-files.
2572 In practice, only affects FreeBSD (8 onwards).
2573 Bugzilla 1053, with thanks to Dmitry Isaikin.
2575 NM/01 Bugzilla 949 - Documentation tweak
2577 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2580 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2582 PP/02 Implemented gsasl authenticator.
2584 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2586 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2587 `pkg-config foo` for cflags/libs.
2589 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2590 with rest of GSASL and with heimdal_gssapi.
2592 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2593 `pkg-config foo` for cflags/libs for the TLS implementation.
2595 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2596 properties get this fed in as external SSF. A number of robustness
2597 and debugging improvements to the cyrus_sasl authenticator.
2599 PP/08 cyrus_sasl server now expands the server_realm option.
2601 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2602 Patch by Jeremy Harris.
2604 PP/10 Added dbmjz lookup type.
2606 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2608 PP/12 MAIL args handles TAB as well as SP, for better interop with
2609 non-compliant senders.
2610 Analysis and variant patch by Todd Lyons.
2612 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2613 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2614 Patch from Dirk Mueller <dmueller@suse.com>
2616 PP/13 tls_peerdn now print-escaped for spool files.
2617 Observed some $tls_peerdn in wild which contained \n, which resulted
2618 in spool file corruption.
2620 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2621 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2622 or write after TLS renegotiation, which otherwise led to messages
2625 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2626 as a tracking header (ie: a signed header comes before the signature).
2627 Patch from Wolfgang Breyha.
2629 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2630 comma-sep list; embedded commas doubled.
2632 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2634 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2636 Report and patch from Dmitry Banschikov.
2638 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2639 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2640 failures. We appear to now support TLS1.1+ with Exim.
2642 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2643 lets Exim select keys and certificates based upon TLS SNI from client.
2644 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2645 before an outbound SMTP session. New log_selector, +tls_sni.
2647 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2648 NULL dereference. Report and patch from Alun Jones.
2650 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2651 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2652 Not seeing resolver debug output on NetBSD, but suspect this is a
2653 resolver implementation change.
2655 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2656 Left warnings. Added "eximon gdb" invocation mode.
2658 PP/21 Defaulting "accept_8bitmime" to true, not false.
2660 PP/22 Added -bw for inetd wait mode support.
2662 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2663 locate the relevant includes and libraries. Made this the default.
2665 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2666 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2668 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2669 This may cause build issues on older platforms.
2671 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2672 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2673 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2674 Added SNI support via GnuTLS too.
2675 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2677 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2679 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2680 Applied second patch from Janne, implementing suggestion to default
2681 multiple-strings-in-record handling to match SPF spec.
2683 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2685 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2686 read-only, out of scope).
2687 Patch from Wolfgang Breyha, report from Stuart Northfield.
2689 PP/29 Fix three issues highlighted by clang analyser static analysis.
2690 Only crash-plausible issue would require the Cambridge-specific
2691 iplookup router and a misconfiguration.
2692 Report from Marcin Mirosław.
2694 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2696 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2697 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2698 As part of this, removing so much warning spew let me fix some minor
2699 real issues in debug logging.
2701 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2702 assignment on my part. Fixed.
2704 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2705 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2706 Janne Snabb (who went above and beyond: thank you).
2708 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2709 string otherwise requires a connection and a bunch more work and it's
2710 relatively easy to get wrong. Should also expose TLS library linkage
2713 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2714 64-bit ${eval} (JH/03).
2716 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2717 GNU libc to support some of the 64-bit stuff, should not lead to
2718 conflicts. Defined before os.h is pulled in, so if a given platform
2719 needs to override this, it can.
2721 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2722 protection layer was required, which is not implemented.
2723 Bugzilla 1254, patch from Wolfgang Breyha.
2725 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2726 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2727 tls_dhparam take prime identifiers. Also unbreak combination of
2728 OpenSSL+DH_params+TLSSNI.
2730 PP/39 Disable SSLv2 by default in OpenSSL support.
2736 PP/01 Solaris build fix for Oracle's LDAP libraries.
2737 Bugzilla 1109, patch from Stephen Usher.
2739 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2741 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2744 TF/02 Fix a couple more cases where we did not log the error message
2745 when unlink() failed. See also change 4.74-TF/03.
2747 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2748 lock up or crash if it happened to be inside a call to libc when it
2749 got a SIGUSR1 from exiwhat.
2751 The SIGUSR1 handler appends the current process status to the process
2752 log which is later printed by exiwhat. It used to use the general
2753 purpose logging code to do this, but several functions it calls are
2754 not safe for signals.
2756 The new output code in the SIGUSR1 handler is specific to the process
2757 log, and simple enough that it's easy to inspect for signal safety.
2758 Removing some special cases also simplifies the general logging code.
2759 Removing the spurious timestamps from the process log simplifies
2762 TF/04 Improved ratelimit ACL condition.
2764 The /noupdate option has been deprecated in favour of /readonly which
2765 has clearer semantics. The /leaky, /strict, and /readonly update modes
2766 are mutually exclusive. The update mode is no longer included in the
2767 database key; it just determines when the database is updated. (This
2768 means that when you upgrade Exim will forget old rate measurements.)
2770 Exim now checks that the per_* options are used with an update mode that
2771 makes sense for the current ACL. For example, when Exim is processing a
2772 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2773 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2774 must specify per_mail/readonly. If you omit the update mode it defaults to
2775 /leaky where that makes sense (as before) or /readonly where required.
2777 The /noupdate option is now undocumented but still supported for
2778 backwards compatibility. It is equivalent to /readonly except that in
2779 ACLs where /readonly is required you may specify /leaky/noupdate or
2780 /strict/noupdate which are treated the same as /readonly.
2782 A useful new feature is the /count= option. This is a generalization
2783 of the per_byte option, so that you can measure the throughput of other
2784 aggregate values. For example, the per_byte option is now equivalent
2785 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2787 The per_rcpt option has been generalized using the /count= mechanism
2788 (though it's more complicated than the per_byte equivalence). When it is
2789 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2790 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2791 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2792 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2793 Note that using per_rcpt with a non-readonly update mode in more than
2794 one ACL will cause the recipients to be double-counted. (The per_mail
2795 and per_byte options don't have this problem.)
2797 The handling of very low rates has changed slightly. If the computed rate
2798 is less than the event's count (usually one) then this event is the first
2799 after a long gap. In this case the rate is set to the same as this event's
2800 count, so that the first message of a spam run is counted properly.
2802 The major new feature is a mechanism for counting the rate of unique
2803 events. The new per_addr option counts the number of different
2804 recipients that someone has sent messages to in the last time period. It
2805 behaves like per_rcpt if all the recipient addresses are different, but
2806 duplicate recipient addresses do not increase the measured rate. Like
2807 the /count= option this is a general mechanism, so the per_addr option
2808 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2809 example, measure the rate that a client uses different sender addresses
2810 with the options per_mail/unique=$sender_address. There are further
2811 details in the main documentation.
2813 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2815 TF/06 Removed a few PCRE remnants.
2817 TF/07 Automatically extract Exim's version number from tags in the git
2818 repository when doing development or release builds.
2820 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2821 Bugzilla 879. Patch from Paul Fisher.
2823 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2824 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2827 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2829 PP/05 Variable $av_failed, true if the AV scanner deferred.
2830 Bugzilla 1078. Patch from John Horne.
2832 PP/06 Stop make process more reliably on build failure.
2833 Bugzilla 1087. Patch from Heiko Schlittermann.
2835 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2836 Bugzilla 1089. Patch from Heiko Schlittermann.
2838 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2839 Bugzilla 1131. Patch from Holger Weiß.
2841 PP/09 Handle IPv6 addresses with SPF.
2842 Bugzilla 860. Patch from Wolfgang Breyha.
2844 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2846 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2849 PP/11 match_* no longer expand right-hand-side by default.
2850 New compile-time build option, EXPAND_LISTMATCH_RHS.
2851 New expansion conditions, "inlist", "inlisti".
2853 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2855 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2857 PP/14 fix log_write() format string regression from TF/03.
2858 Bugzilla 1152. Patch from Dmitry Isaikin.
2864 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2866 PP/02 Harmonised TLS library version reporting; only show if debugging.
2867 Layout now matches that introduced for other libraries in 4.74 PP/03.
2869 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2871 PP/04 New "dns_use_edns0" global option.
2873 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2876 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2877 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2879 TK/01 Updated PolarSSL code to 0.14.2.
2880 Bugzilla 1097. Patch from Andreas Metzler.
2882 PP/07 Catch divide-by-zero in ${eval:...}.
2883 Fixes bugzilla 1102.
2885 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2888 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2889 format-string attack -- SECURITY: remote arbitrary code execution.
2891 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2892 time unintentionally subject to list matching rules, letting the header
2893 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2894 arbitrary string expansion). This allowed for information disclosure.
2896 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2897 INT_MIN/-1 -- value coerced to INT_MAX.
2903 NM/01 Workaround for PCRE version dependency in version reporting
2906 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2907 This fixes portability to compilers other than gcc, notably
2908 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2910 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2911 makefiles for portability to HP-UX and POSIX correctness.
2913 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2914 Also via indented variable definition in the Makefile.
2915 (Debugging by Oliver Heesakkers).
2917 PP/02 Restore caching of spamd results with expanded spamd_address.
2918 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2920 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2921 Improves build reliability. Fix from: Frank Elsner
2923 NM/02 Fix wide character breakage in the rfc2047 coding
2924 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2926 NM/03 Allow underscore in dnslist lookups
2927 Fixes bug 1026. Patch from Graeme Fowler
2929 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2930 Code patches from Adam Ciarcinski of NetBSD.
2932 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2935 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2936 is logged, to avoid truncation. Patch from John Horne.
2938 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2939 Patch from Jakob Hirsch.
2941 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2942 SQL string expansion failure details.
2943 Patch from Andrey Oktyabrski.
2945 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2946 Patch from Simon Arlott.
2948 PP/09 New lookups functionality failed to compile on old gcc which rejects
2949 extern declarations in function scope.
2950 Patch from Oliver Fleischmann
2952 PP/10 Use sig_atomic_t for flags set from signal handlers.
2953 Check getgroups() return and improve debugging.
2954 Fixed developed for diagnosis in bug 927 (which turned out to be
2957 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2958 Patch from Mark Zealey.
2960 PP/12 Bugzilla 1056: Improved spamd server selection.
2961 Patch from Mark Zealey.
2963 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2964 Based on patch from Heiko Schlittermann.
2966 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2967 Patch from Uwe Doering, sign-off by Michael Haardt.
2969 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2970 variable declaration deep within a block. Bug and patch from
2973 PP/15 lookups-Makefile IRIX compatibility coercion.
2975 PP/16 Make DISABLE_DKIM build knob functional.
2977 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2978 Patch by Simon Arlott
2980 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2981 variable argument macros. Our copy now differs from upstream.
2987 TF/01 Failure to get a lock on a hints database can have serious
2988 consequences so log it to the panic log.
2990 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2991 controlled using the smtp_confirmation log selector.
2993 TF/03 Include the error message when we fail to unlink a spool file.
2995 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2996 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2997 for maintaining out-of-tree patches for some time.
2999 PP/01 Bugzilla 139: Documentation and portability issues.
3000 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3001 Handle per-OS dynamic-module compilation flags.
3003 PP/02 Let /dev/null have normal permissions.
3004 The 4.73 fixes were a little too stringent and complained about the
3005 permissions on /dev/null. Exempt it from some checks.
3006 Reported by Andreas M. Kirchwitz.
3008 PP/03 Report version information for many libraries, including
3009 Exim version information for dynamically loaded libraries. Created
3010 version.h, now support a version extension string for distributors
3011 who patch heavily. Dynamic module ABI change.
3013 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3014 privilege escalation vulnerability whereby the Exim run-time user
3015 can cause root to append content of the attacker's choosing to
3018 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3021 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3022 If dropping privileges for untrusted macros, we disabled normal logging
3023 on the basis that it would fail; for the Exim run-time user, this is not
3024 the case, and it resulted in successful deliveries going unlogged.
3025 Fixed. Reported by Andreas Metzler.
3031 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3032 only prepend for the Resent-* case. Fixes regression introduced in
3033 Exim 4.70 by NM/22 for Bugzilla 607.
3035 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3036 increasing numbers of administrators be bitten by this.
3038 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3040 PP/03 Bugzilla 994: added openssl_options main configuration option.
3042 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3044 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3046 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3048 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3049 without a peer certificate, leading to a segfault because of an
3050 assumption that peers always have certificates. Be a little more
3051 paranoid. Problem reported by Martin Tscholak.
3053 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3054 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3055 NB: ClamAV planning to remove STREAM in "middle of 2010".
3056 CL also introduces -bmalware, various -d+acl logging additions and
3057 more caution in buffer sizes.
3059 PP/09 Implemented reverse_ip expansion operator.
3061 PP/10 Bugzilla 937: provide a "debug" ACL control.
3063 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3065 PP/12 Bugzilla 973: Implement --version.
3067 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3069 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3071 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3073 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3074 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3075 ignore trailing whitespace.
3077 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3079 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3082 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3083 Notification from Dr Andrew Aitchison.
3085 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3086 ExtendedDetectionInfo response format.
3087 Notification from John Horne.
3089 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3092 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3093 XSL and documented dependency on system catalogs, with examples of how
3096 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3099 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3100 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3101 configuration file which is writeable by the Exim user or group.
3103 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3104 of configuration files to cover files specified with the -C option if
3105 they are going to be used with root privileges, not just the default
3108 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3109 option (effectively making it always true).
3111 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3112 files to be used while preserving root privileges.
3114 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3115 that rogue child processes cannot use them.
3117 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3118 run-time user, instead of root.
3120 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3121 Exim run-time user without dropping privileges.
3123 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3124 result string, instead of calling string_vformat() twice with the same
3127 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3128 for other users. Others should always drop root privileges if they use
3129 -C on the command line, even for a whitelisted configure file.
3131 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3133 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3139 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3140 $data_path, and $header_path variables; fixed documentation bugs and
3143 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3144 exipick to access non-standard spools, including the "frozen" queue
3147 NM/01 Bugzilla 965: Support mysql stored procedures.
3148 Patch from Alain Williams
3150 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3152 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3153 Patch from Andreas Metzler
3155 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3156 Patch from Kirill Miazine
3158 NM/05 Bugzilla 671: Added umask to procmail example.
3160 JJ/03 installed exipick 20100323.0, fixing doc bug
3162 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3163 directory. Notification and patch from Dan Rosenberg.
3165 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3167 TK/02 Improve log output when DKIM signing operation fails.
3169 MH/01 Treat the transport option dkim_domain as a colon separated
3170 list, not as a single string, and sign the message with each element,
3171 omitting multiple occurences of the same signer.
3173 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3174 Bugzilla 985, 986. Patch by Simon Arlott
3176 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3177 Patch by Simon Arlott
3179 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3180 MBX locking. Notification from Dan Rosenberg.
3186 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3188 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3190 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3192 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3194 NM/04 Bugzilla 915: Flag broken perl installation during build.
3200 TK/01 Added patch by Johannes Berg that expands the main option
3201 "spamd_address" if it starts with a dollar sign.
3203 TK/02 Write list of recipients to X-Envelope-Sender header when building
3204 the mbox-format spool file for content scanning (suggested by Jakob
3207 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3208 (http://www.dcc-servers.net/) support via dccifd. Activated by
3209 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3211 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3212 by Mark Daniel Reidel <mr@df.eu>.
3214 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3215 When building exim an external PCRE library is now needed -
3216 PCRE is a system library on the majority of modern systems.
3217 See entry on PCRE_LIBS in EDITME file.
3219 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3220 conversation. Added nologin parameter to request.
3221 Patch contributed by Kirill Miazine.
3223 TF/01 Do not log submission mode rewrites if they do not change the address.
3225 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3227 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3228 log files in place. Contributed by Roberto Lima.
3230 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3232 TF/03 Bugzilla 615: When checking the local_parts router precondition
3233 after a local_part_suffix or local_part_prefix option, Exim now
3234 does not use the address's named list lookup cache, since this
3235 contains cached lookups for the whole local part.
3237 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3238 Robert Millan. Documentation is in experimental-spec.txt.
3240 TF/04 Bugzilla 668: Fix parallel build (make -j).
3242 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3244 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3245 Patch provided by Jan Srzednicki.
3247 TF/05 Leading white space used to be stripped from $spam_report which
3248 wrecked the formatting. Now it is preserved.
3250 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3251 that they are available at delivery time.
3253 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3255 TF/08 TLS error reporting now respects the incoming_interface and
3256 incoming_port log selectors.
3258 TF/09 Produce a more useful error message if an SMTP transport's hosts
3259 setting expands to an empty string.
3261 NM/06 Bugzilla 744: EXPN did not work under TLS.
3262 Patch provided by Phil Pennock.
3264 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3265 Patch provided by Richard Godbee.
3267 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3268 acl_smtp_notquit, added index entry.
3270 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3271 Patch provided by Eugene Bujak.
3273 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3274 accept(). Patch provided by Maxim Dounin.
3276 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3277 Patch provided by Phil Pennock.
3279 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3281 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3282 Patch provided by Brad "anomie" Jorsch.
3284 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3285 Patch provided by Dean Brooks.
3287 TK/05 Add native DKIM support (does not depend on external libraries).
3289 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3290 Patch provided by Graeme Fowler.
3292 NM/16 Bugzilla 851: Documentation example syntax fix.
3294 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3296 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3299 NM/19 Bugzilla 745: TLS version reporting.
3300 Patch provided by Phil Pennock.
3302 NM/20 Bugzilla 167: bool: condition support.
3303 Patch provided by Phil Pennock.
3305 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3306 clients. Patch provided by Phil Pennock.
3308 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3309 Patch provided by Brad "anomie" Jorsch.
3311 NM/23 Bugzilla 687: Fix misparses in eximstats.
3312 Patch provided by Heiko Schlittermann.
3314 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3315 Patch provided by Heiko Schlittermann.
3317 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3318 plus update to original patch.
3320 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3322 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3323 Patch provided by David Brownlee.
3325 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3327 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3329 NM/30 Bugzilla 888: TLS documentation bugfixes.
3331 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3333 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3334 Unlike the original bugzilla I have changed all shell scripts in src tree.
3336 NM/33 Bugzilla 898: Transport filter timeout fix.
3337 Patch by Todd Rinaldo.
3339 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3340 Patch by Serge Demonchaux.
3342 NM/35 Bugzilla 39: Base64 decode bug fixes.
3343 Patch by Jakob Hirsch.
3345 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3347 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3349 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3351 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3357 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3358 ALT-N's libdkim that I have put here:
3359 http://duncanthrax.net/exim-experimental/
3361 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3362 were called 'true' and it seems that C99 defines that as a reserved
3363 keyword to be used with 'bool' variable types. That means you could
3364 not include C99-style headers which use bools without triggering
3365 build errors in sieve.c.
3367 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3368 as mailq or other aliases. Changed the --help handling significantly
3369 to do whats expected. exim_usage() emits usage/help information.
3371 SC/01 Added the -bylocaldomain option to eximstats.
3373 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3375 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3377 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3383 PH/01 Another patch from the Sieve maintainer.
3385 PH/02 When an IPv6 address is converted to a string for single-key lookup
3386 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3387 dots are used instead of colons so that keys in lsearch files need not
3388 contain colons. This was done some time before quoting was made available
3389 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3390 (notated using the quote facility) so as to distinguish them from IPv4
3391 keys. This meant that lookups for IP addresses in host lists did not work
3392 for iplsearch lookups.
3394 This has been fixed by arranging for IPv6 addresses to be expressed with
3395 colons if the lookup type is iplsearch. This is not incompatible, because
3396 previously such lookups could never work.
3398 The situation is now rather anomalous, since one *can* have colons in
3399 ordinary lsearch keys. However, making the change in all cases is
3400 incompatible and would probably break a number of configurations.
3402 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3405 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3406 conversion specification without a maximum field width, thereby enabling
3407 a rogue spamd server to cause a buffer overflow. While nobody in their
3408 right mind would setup Exim to query an untrusted spamd server, an
3409 attacker that gains access to a server running spamd could potentially
3410 exploit this vulnerability to run arbitrary code as the Exim user.
3412 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3413 $primary_hostname instead of what libspf2 thinks the hosts name is.
3415 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3416 a directory entry by the name of the lookup key. Previously, if a
3417 symlink pointed to a non-existing file or a file in a directory that
3418 Exim lacked permissions to read, a lookup for a key matching that
3419 symlink would fail. Now it is enough that a matching directory entry
3420 exists, symlink or not. (Bugzilla 503.)
3422 PH/03 The body_linecount and body_zerocount variables are now exported in the
3425 PH/04 Added the $dnslist_matched variable.
3427 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3428 This means they are set thereafter only if the connection becomes
3431 PH/06 Added the client_condition to authenticators so that some can be skipped
3432 by clients under certain conditions.
3434 PH/07 The error message for a badly-placed control=no_multiline_responses left
3435 "_responses" off the end of the name.
3437 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3439 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3440 (without spaces) instead of just copying the configuration text.
3442 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3444 PH/11 Added $max_received_linelength.
3446 PH/12 Added +ignore_defer and +include_defer to host lists.
3448 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3449 way in which PCRE > 7.0 is built.
3451 PH/14 Implemented queue_only_load_latch.
3453 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3454 MAIL command. The effect was to mangle the value on 64-bit systems.
3456 PH/16 Another patch from the Sieve maintainer.
3458 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3460 PH/18 If a system quota error occurred while trying to create the file for
3461 a maildir delivery, the message "Mailbox is full" was not appended to the
3462 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3463 only to a quota excession during the actual writing of the file.
3465 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3466 characters?) which causes problems in log lines. The DN values are now
3467 passed through string_printing() before being added to log lines.
3469 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3470 and InterBase are left for another time.)
3472 PH/21 Added message_body_newlines option.
3474 PH/22 Guard against possible overflow in moan_check_errorcopy().
3476 PH/23 POSIX allows open() to be a macro; guard against that.
3478 PH/24 If the recipient of an error message contained an @ in the local part
3479 (suitably quoted, of course), incorrect values were put in $domain and
3480 $local_part during the evaluation of errors_copy.
3486 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3487 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3490 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3491 issue a MAIL command.
3493 PH/02 In an ACL statement such as
3495 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3497 if a client was not listed at all, or was listed with a value other than
3498 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3499 the condition was not true (as it should be), so access was not denied.
3500 The bug was that the ! inversion was incorrectly passed on to the second
3501 item. This has been fixed.
3503 PH/03 Added additional dnslists conditions == and =& which are different from
3504 = and & when the dns lookup returns more than one IP address.
3506 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3507 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3509 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3510 FSYNC, which compiles an option called disable_fsync that allows for
3511 bypassing fsync(). The documentation is heavily laced with warnings.
3513 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3515 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3516 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3517 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3518 including adding "make clean"; (3) Added -fPIC when compiling the test
3519 dynamically loaded module, to get rid of a warning.
3521 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3522 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3523 = 0s. The bug is otherwise harmless.
3525 PH/07 There was a bug in the dovecot authenticator such that the value of
3526 $auth1 could be overwritten, and so not correctly preserved, after a
3527 successful authentication. This usually meant that the value preserved by
3528 the server_setid option was incorrect.
3530 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3532 PH/09 Installed PCRE release 7.0.
3534 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3535 run for batched SMTP input. It is now run at the start of every message
3536 in the batch. While fixing this I discovered that the process information
3537 (output by running exiwhat) was not always getting set for -bs and -bS
3538 input. This is fixed, and it now also says "batched" for BSMTP.
3540 PH/11 Added control=no_pipelining.
3542 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3543 patch, slightly modified), and move the expansion of helo_data till after
3544 the connection is made in the smtp transport (so it can use these
3547 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3549 PH/14 Added log_selector = +pid.
3551 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3553 PH/16 Add ${if forany and ${if forall.
3555 PH/17 Added dsn_from option to vary the From: line in DSNs.
3557 PH/18 Flush SMTP output before performing a callout, unless control =
3558 no_callout_flush is set.
3560 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3561 was true (the default) a successful delivery failed to delete the retry
3562 item, thus causing premature timeout of the address. The bug is now
3565 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3567 PH/21 Long custom messages for fakedefer and fakereject are now split up
3568 into multiline responses in the same way that messages for "deny" and
3569 other ACL rejections are.
3571 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3572 with slight modification.
3574 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3575 draft, changing the syntax and factoring some duplicate code".
3577 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3578 for deliveries of the second and subsequent messages over the same SMTP
3581 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3582 ${reduce, with only minor "tidies".
3584 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3586 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3587 expansion side effects.
3589 PH/27 When a message times out after an over-quota error from an Exim-imposed
3590 quota, the bounce message says "mailbox is full". This message was not
3591 being given when it was a system quota that was exceeded. It now should
3594 MH/03 Made $recipients available in local_scan(). local_scan() already has
3595 better access to the recipient list through recipients_list[], but
3596 $recipients can be useful in postmaster-provided expansion strings.
3598 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3599 in the case of a MAIL command with additional options following the
3600 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3601 were accidentally chopped off.
3603 PH/29 SMTP synchronization checks are implemented when a command is read -
3604 there is a check that no more input is waiting when there shouldn't be
3605 any. However, for some commands, a delay in an ACL can mean that it is
3606 some time before the response is written. In this time, more input might
3607 arrive, invalidly. So now there are extra checks after an ACL has run for
3608 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3609 pipelining has not been advertised.
3611 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3613 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3614 returned in the SMTP response when smtp_return_error_details was false.
3615 This has been fixed.
3617 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3618 instead of the C I/O that was originally supplied, because problems were
3619 reported on Solaris.
3621 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3622 Exim which did not show up earlier: it was assuming that a call to
3623 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3624 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3625 was a macro that became an assignment, so it seemed to work. This has
3626 changed to a proper function call with a void return, hence the compile
3627 error. Exim's code has been fixed.
3629 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3632 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3634 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3636 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3637 because it turns out to be too big or there is a timeout) there is no
3638 "Completed" line in the log. When some messages of this type were
3639 selected by exigrep, they were listed as "not completed". Others were
3640 picked up by some special patterns. I have improved the selection
3641 criteria to be more general.
3643 PH/38 The host_find_failed option in the manualroute router can now be set
3644 to "ignore", to completely ignore a host whose IP address cannot be
3645 found. If all hosts are ignored, the behaviour is controlled by the new
3646 host_all_ignored option.
3648 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3649 homing or because of multiple MX records with /mx) generated more than
3650 one IP address, and the following item turned out to be the local host,
3651 all the secondary addresses of the first item were incorrectly removed
3652 from the list, along with the local host and any following hosts (which
3653 is what is supposed to happen).
3655 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3656 whoever called Exim into the -H file. In the case of the daemon it was
3657 behaving confusingly. When first started, it used values for whoever
3658 started the daemon, but after a SIGHUP it used the Exim user (because it
3659 calls itself on a restart). I have changed the code so that it now always
3662 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3663 message are rejected with the same error (e.g. no authentication or bad
3664 sender address), and a DATA command is nevertheless sent (as can happen
3665 with PIPELINING or a stupid MUA), the error message that was given to the
3666 RCPT commands is included in the rejection of the DATA command. This is
3667 intended to be helpful for MUAs that show only the final error to their
3670 PH/42 Another patch from the Sieve maintainer.
3672 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3673 Eximstats - Fixed some broken HTML links and added missing column headers
3675 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3676 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3678 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3680 PH/43 Yet another patch from the Sieve maintainer.
3682 PH/44 I found a way to check for a TCP/IP connection going away before sending
3683 the response to the final '.' that terminates a message, but only in the
3684 case where the client has not sent further data following the '.'
3685 (unfortunately, this is allowed). However, in many cases there won't be
3686 any further data because there won't be any more messages to send. A call
3687 to select() can be used: if it shows that the input is "ready", there is
3688 either input waiting, or the socket has been closed. An attempt to read
3689 the next input character can distinguish the two cases. Previously, Exim
3690 would have sent an OK response which the client would never have see.
3691 This could lead to message repetition. This fix should cure that, at
3692 least in a lot of common cases.
3694 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3695 advertised in response to EHLO.
3701 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3702 fixed by 4.65/MH/01 (is this a record?) are fixed:
3704 (i) An empty string was always treated as zero by the numeric comparison
3705 operators. This behaviour has been restored.
3707 (ii) It is documented that the numeric comparison operators always treat
3708 their arguments as decimal numbers. This was broken in that numbers
3709 starting with 0 were being interpreted as octal.
3711 While fixing these problems I realized that there was another issue that
3712 hadn't been noticed. Values of message_size_limit (both the global option
3713 and the transport option) were treated as octal if they started with 0.
3714 The documentation was vague. These values are now always treated as
3715 decimal, and I will make that clear in the documentation.
3721 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3722 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3725 MH/01 Don't check that the operands of numeric comparison operators are
3726 integers when their expansion is in "skipping" mode (fixes bug
3727 introduced by 4.64-PH/07).
3729 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3730 child addresses, Exim now panics and dies. Previously, because the count
3731 is held in a short int, deliveries were likely to be lost. As such a
3732 large number of recipients for a single message is ridiculous
3733 (performance will be very, very poor), I have chosen to impose a limit
3734 rather than extend the field.
3740 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3741 leftover -K file (the existence of which was triggered by #402).
3742 While we were at it, introduced process PID as part of the -K
3743 filename. This should rule out race conditions when creating
3746 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3747 processing considerably. Previous code took too long for large mails,
3748 triggering a timeout which in turn triggers #401.
3750 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3751 in the DK code in transports.c. sendfile() is not really portable,
3752 hence the _LINUX specificness.
3754 TF/01 In the add_headers option to the mail command in an Exim filter,
3755 there was a bug that Exim would claim a syntax error in any
3756 header after the first one which had an odd number of characters
3759 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3760 callout verification, Exim cached a "reject" for the entire domain. This
3761 is correct for most verifications, but it is not correct for a recipient
3762 verification with use_sender or use_postmaster set, because in that case
3763 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3764 case of MAIL FROM:<> rejection from other early rejections (e.g.
3765 rejection of HELO). When verifying a recipient using a non-null MAIL
3766 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3767 Whatever the result of the callout, the value of the domain cache is
3768 left unchanged (for any other kind of callout, getting as far as trying
3769 RCPT means that the domain itself is ok).
3771 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3774 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3775 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3776 session, the daemon ignores ECONNECT errors and logs others; it now
3777 ignores EPIPE as well.
3779 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3780 (quoted-printable decoding).
3782 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3783 later the small subsequent patch to fix an introduced bug.
3785 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3787 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3789 PH/08 An error is now given if message_size_limit is specified negative.
3791 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3792 to be given (somewhat) arbitrary names.
3794 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3797 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3798 miscellaneous code fixes
3800 PH/10 Added the log_reject_target ACL modifier to specify where to log
3803 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3804 hostname. This is wrong, because it relates to the incoming message (and
3805 probably the interface on which it is arriving) and not to the outgoing
3806 callout (which could be using a different interface). This has been
3807 changed to use the value of the helo_data option from the smtp transport
3808 instead - this is what is used when a message is actually being sent. If
3809 there is no remote transport (possible with a router that sets up host
3810 addresses), $smtp_active_hostname is used.
3812 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3813 tweaks were necessary in order to get it to work (see also 21 below):
3814 (a) The code assumed that strncpy() returns a negative number on buffer
3815 overflow, which isn't the case. Replaced with Exim's string_format()
3817 (b) There were several signed/unsigned issues. I just did the minimum
3818 hacking in of casts. There is scope for a larger refactoring.
3819 (c) The code used strcasecmp() which is not a standard C function.
3820 Replaced with Exim's strcmpic() function.
3821 (d) The code set only $1; it now sets $auth1 as well.
3822 (e) A simple test gave the error "authentication client didn't specify
3823 service in request". It would seem that Dovecot has changed its
3824 interface. Fortunately there's a specification; I followed it and
3825 changed what the client sends and it appears to be working now.
3827 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3830 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3831 address A is aliased to B and C, where B exists and C does not. Without
3832 -v the output is "A verified" because verification stops after a
3833 successful redirection if more than one address is generated. However,
3834 with -v the child addresses are also verified. Exim was outputting "A
3835 failed to verify" and then showing the successful verification for C,
3836 with its parentage. It now outputs "B failed to verify", showing B's
3837 parentage before showing the successful verification of C.
3839 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3840 look up a TXT record in a specific list after matching in a combined
3843 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3844 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3845 they consult the DNS. I had assumed they would set it the way they
3846 wanted; and indeed my experiments on Linux seem to show that in some
3847 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3848 To be on the safe side, however, I have now made the interface to
3849 host_find_byname() similar to host_find_bydns(), with an argument
3850 containing the DNS resolver options. The host_find_byname() function now
3851 sets these options at its start, just as host_find_bydns() does. The smtp
3852 transport options dns_qualify_single and dns_search_parents are passed to
3853 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3854 of host_find_byname() use the default settings of RES_DEFNAMES
3855 (qualify_single) but not RES_DNSRCH (search_parents).
3857 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3858 spool_read_header() do less string testing, by means of a preliminary
3859 switch on the second character of optional "-foo" lines. (This is
3860 overdue, caused by the large number of possibilities that now exist.
3861 Originally there were few.) While I was there, I also converted the
3862 str(n)cmp tests so they don't re-test the leading "-" and the first
3863 character, in the hope this might squeeze out yet more improvement.
3865 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3866 flag allowing group syntax was set by the header_syntax check but not
3867 turned off, possible causing trouble later; (2) The flag was not being
3868 set at all for the header_verify test, causing "group"-style headers to
3869 be rejected. I have now set it in this case, and also caused header_
3870 verify to ignore an empty address taken from a group. While doing this, I
3871 came across some other cases where the code for allowing group syntax
3872 while scanning a header line wasn't quite right (mostly, not resetting
3873 the flag correctly in the right place). These bugs could have caused
3874 trouble for malformed header lines. I hope it is now all correct.
3876 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3877 with the "reply" argument non-NULL. The code, however (which originally
3878 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3879 but it didn't always do it. This confused somebody who was copying the
3880 code for some other use. I have removed all the tests.
3882 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3883 feature that was used to support insecure browsers during the U.S. crypto
3884 embargo. It requires special client support, and Exim is probably the
3885 only MTA that supported it -- and would never use it because real RSA is
3886 always available. This code has been removed, because it had the bad
3887 effect of slowing Exim down by computing (never used) parameters for the
3888 RSA_EXPORT functionality.
3890 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3891 authenticator to fail if there's a tab character in the incoming data
3892 (there should never be unless someone is messing about, as it's supposed
3893 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3894 if the connection is using TLS or if the remote IP is the same as the
3895 local IP, and the "valid-client-cert option" if a client certificate has
3898 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3899 authenticators. This can be used for authorization after authentication
3900 succeeds. (In the case of plaintext, it servers for both authentication
3903 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3904 if any retry times were supplied.
3906 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3907 connection, where there is no client IP address to check. In this
3908 situation, the verify now always succeeds.
3910 PH/25 Applied John Jetmore's -Mset patch.
3912 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3914 PH/27 In a string expansion for a processed (not raw) header when multiple
3915 headers of the same name were present, leading whitespace was being
3916 removed from all of them, but trailing whitespace was being removed only
3917 from the last one. Now trailing whitespace is removed from each header
3918 before concatenation. Completely empty headers in a concatenation (as
3919 before) are ignored.
3921 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3922 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3924 PH/29 [Removed. This was a change that I later backed out, and forgot to
3925 correct the ChangeLog entry (that I had efficiently created) before
3926 committing the later change.]
3928 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3929 address errors (4xx response to RCPT) over the same connection as other
3930 messages routed to the same hosts. Such deliveries are always "forced",
3931 so retry times are not inspected. This resulted in far too many retries
3932 for the affected addresses. The effect occurred only when there were more
3933 hosts than the hosts_max_try setting in the smtp transport when it had
3934 the 4xx errors. Those hosts that it had tried were not added to the list
3935 of hosts for which the message was waiting, so if all were tried, there
3936 was no problem. Two fixes have been applied:
3938 (i) If there are any address or message errors in an SMTP delivery, none
3939 of the hosts (tried or untried) are now added to the list of hosts
3940 for which the message is waiting, so the message should not be a
3941 candidate for sending over the same connection that was used for a
3942 successful delivery of some other message. This seems entirely
3943 reasonable: after all the message is NOT "waiting for some host".
3944 This is so "obvious" that I'm not sure why it wasn't done
3945 previously. Hope I haven't missed anything, but it can't do any
3946 harm, as the worst effect is to miss an optimization.
3948 (ii) If, despite (i), such a delivery is accidentally attempted, the
3949 routing retry time is respected, so at least it doesn't keep
3950 hammering the server.
3952 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3953 in ${readsocket because some servers need this prod.
3955 PH/32 Added some extra debug output when updating a wait-xxx database.
3957 PH/33 The hint "could be header name not terminated by colon", which has been
3958 given for certain expansion errors for a long time, was not being given
3959 for the ${if def:h_colon_omitted{... case.
3961 PH/34 The spec says: "With one important exception, whenever a domain list is
3962 being scanned, $domain contains the subject domain." There was at least
3963 one case where this was not true.
3965 PH/35 The error "getsockname() failed: connection reset by peer" was being
3966 written to the panic log as well as the main log, but it isn't really
3967 panic-worthy as it just means the connection died rather early on. I have
3968 removed the panic log writing for the ECONNRESET error when getsockname()
3971 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3972 runs only) independently of the message's sender address. This meant
3973 that, if the 4xx error was in fact related to the sender, a different
3974 message to the same recipient with a different sender could confuse
3975 things. In particular, this can happen when sending to a greylisting
3976 server, but other circumstances could also provoke similar problems.
3977 I have changed the default so that the retry time for these errors is now
3978 based a combination of the sender and recipient addresses. This change
3979 can be overridden by setting address_retry_include_sender=false in the
3982 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3983 remote server are returned as part of bounce messages. This was not
3984 happening for LMTP over a pipe (the lmtp transport), but now it is the
3985 same for both kinds of LMTP.
3987 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3988 in header lines that were in fact CNAMEs. This is no longer the case.
3990 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3991 and queue runs started by the daemon processed all messages. This has
3992 been fixed so that -R and -S can now usefully be given with -q<time>.
3994 PH/40 Import PCRE release 6.7 (fixes some bugs).
3996 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3998 PH/42 Give an error if -q is specified more than once.
4000 PH/43 Renamed the variables $interface_address and $interface_port as
4001 $received_ip_address and $received_port, to make it clear that these
4002 values apply to message reception, and not to the outgoing interface when
4003 a message is delivered. (The old names remain recognized, of course.)
4005 PH/44 There was no timeout on the connect() call when using a Unix domain
4006 socket in the ${readsocket expansion. There now is.
4008 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4009 be meaningful with "accept".
4011 SC/01 Eximstats V1.43
4012 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4014 SC/02 Eximstats V1.44
4015 Use a glob alias rather than an array ref in the generated
4016 parser. This improves both readability and performance.
4018 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4019 Collect SpamAssassin and rejection statistics.
4020 Don't display local sender or destination tables unless
4021 there is data to show.
4022 Added average volumes into the top table text output.
4024 SC/04 Eximstats V1.46
4025 Collect data on the number of addresses (recipients)
4026 as well as the number of messages.
4028 SC/05 Eximstats V1.47
4029 Added 'Message too big' to the list of mail rejection
4030 reasons (thanks to Marco Gaiarin).
4032 SC/06 Eximstats V1.48
4033 Mainlog lines which have GMT offsets and are too short to
4034 have a flag are now skipped.
4036 SC/07 Eximstats V1.49 (Alain Williams)
4037 Added the -emptyok flag.
4039 SC/08 Eximstats V1.50
4040 Fixes for obtaining the IP address from reject messages.
4042 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4043 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4044 whitespace changes from 4.64-PH/27
4046 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4049 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4050 are found, allow negative numbers in numeric criteria)
4052 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4054 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4055 to match changes made in 4.64-PH/43
4057 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4059 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4060 "domain" is actually an IP address, and force a failure. This locks out
4061 those revolvers/nameservers that support "A-for-A" lookups, in
4062 contravention of the specifications.
4064 PH/48 When a host name was looked up from an IP address, and the subsequent
4065 forward lookup of the name timed out, the host name was left in
4066 $sender_host_name, contrary to the specification.
4068 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4069 restricted to single-key lookups, Exim was not diagnosing an error if
4070 * or *@ was used with a query-style lookup.
4072 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4074 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4075 long ago, but noone interested enough thought of it. Let's just say that
4076 the "1.1" means that there are some new functions that weren't there at
4077 some point in the past.
4079 PH/51 Error processing for expansion failure of helo_data from an smtp
4080 transport during callout processing was broken.
4082 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4083 tested/used via the -bh/-bhc/-bs options.
4085 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4086 bug, fixed in subsequent PCRE releases).
4088 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4089 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4091 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4097 SC/01 Use a glob alias rather than an array ref in eximstats generated
4098 parser. This improves both readability and performance.
4100 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4101 Don't display local sender or destination tables in eximstats unless
4102 there is data to show.
4103 Added average volumes into the eximstats top table text output.
4105 SC/03 Collect data on the number of addresses (recipients) as well
4106 as the number of messages in eximstats.
4108 TF/01 Correct an error in the documentation for the redirect router. Exim
4109 does (usually) call initgroups() when daemonizing.
4111 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4112 with consistent privilege compared to when running as a daemon.
4114 TF/03 Note in the spec that $authenticated_id is not set for local
4115 submissions from trusted users.
4117 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4118 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4120 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4121 by adding some example configuration directives to the default
4122 configuration file. A little bit of work is required to uncomment the
4123 directives and define how usernames and passwords are checked, but
4124 there is now a framework to start from.
4126 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4127 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4128 without this. I don't know how relevant this is to other LDAP libraries.
4130 PH/02 Add the verb name to the "unknown ACL verb" error.
4132 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4134 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4136 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4137 directory not expanded when it should be if an expanded home directory
4138 was set for the address (which is overridden by the transport).
4140 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4143 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4144 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4145 because it is too late at that time, and has no effect.
4147 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4148 security issue with \' (bugzilla #107). I could not use the
4149 PQescapeStringConn() function, because it needs a PGconn value as one of
4152 PH/08 When testing addresses using -bt, indicate those final addresses that
4153 are duplicates that would not cause an additional delivery. At least one
4154 person was confused, thinking that -bt output corresponded to deliveries.
4155 (Suppressing duplicates isn't a good idea as you lose the information
4156 about possibly different redirections that led to the duplicates.)
4158 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4159 systems where poll() doesn't work, in particular OS X.
4161 PH/10 Added more information to debugging output for retry time not reached.
4163 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4164 operations in malware.c.
4166 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4169 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4170 syslog_duplication set false, log lines that would normally be written
4171 both the the main log and to the reject log were not written to syslog at
4174 PH/14 In the default configuration, change the use of "message" in ACL warn
4175 statements to "add_header".
4177 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4178 not followed by a command (e.g. "seen endif").
4180 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4181 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4184 PH/17 Added extra conditions to the default value of delay_warning_condition
4188 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4189 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4190 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4193 The Auto-Submitted: and various List- headers are standardised, whereas I
4194 don't think Precedence: ever was.
4196 PH/18 Refactored debugging code in route_finduser() to show more information,
4197 in particular, the error code if getpwnam() issues one.
4199 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4200 This is apparently needed in addition to the PH/07 change above to avoid
4201 any possible encoding problems.
4203 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4204 but not after initializing Perl.
4206 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4207 output them only if debugging. By default they are written stderr,
4208 apparently, which is not desirable.
4210 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4213 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4216 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4218 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4219 authenticated or an ident call has been made. Suppress the default
4220 values for $authenticated_id and $authenticated_sender (but permit -oMai
4221 and -oMas) when testing with -bh.
4223 PH/24 Re-jigged the order of the tests in the default configuration so that the
4224 tests for valid domains and recipients precede the DNS black list and CSA
4225 tests, on the grounds that those ones are more expensive.
4227 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4228 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4229 This bug exists in every version of Exim that I still have, right back to
4232 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4233 However, an attempt to turn on case-sensitivity in a regex key by
4234 including (?-i) didn't work because the subject string was already
4235 lowercased, and the effects were non-intuitive. It turns out that a
4236 one-line patch can be used to allow (?-i) to work as expected.
4242 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4243 other effects) broke the use of negated acl sub-conditions.
4245 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4248 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4249 "Deny" causes Exim to reject the incoming connection with a 554 error.
4250 Unfortunately, if there is a major crisis, such as a disk failure,
4251 tcp-wrappers gives "deny", whereas what one would like would be some
4252 kind of temporary error. A kludge has been added to help with this.
4253 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4254 554 error is used if errno is still zero or contains ENOENT (which occurs
4255 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4258 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4260 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4261 errors. Otherwise a message that provokes a temporary error (when other
4262 messages do not) can cause a whole host to time out.
4264 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4265 addresses were routed directly to files or pipes from a redirect router.
4266 File deliveries just didn't batch; pipe deliveries might have suffered
4269 PH/06 A failure to get a lock for a hints database would erroneously always say
4270 "Failed to get write lock", even when it was really a read lock.
4272 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4273 of 0600. This has been changed to use the value of the lockfile_mode
4274 option (which defaults to 0600).
4276 PH/08 Applied small patch from the Sieve maintainer.
4278 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4279 folder from quota calculations, a direct delivery into this folder messed
4280 up the contents of the maildirsize file. This was because the regex was
4281 used only to exclude .Trash (or whatever) when the size of the mailbox
4282 was calculated. There was no check that a delivery was happening into an
4283 excluded directory. This bug has been fixed by ignoring all quota
4284 processing for deliveries into excluded directories.
4286 PH/10 Added the maildirfolder_create_regex option to appendfile.
4292 PH/01 The code for finding all the local interface addresses on a FreeBSD
4293 system running IPv6 was broken. This may well have applied to all BSD
4294 systems, as well as to others that have similar system calls. The broken
4295 code found IPv4 interfaces correctly, but gave incorrect values for the
4296 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4297 that it would not match correctly against @[] and not recognize the IPv6
4300 PH/02 The ipliteral router was not recognizing addresses of the form user@
4301 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4303 PH/03 Added disable_ipv6.
4305 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4306 decoded form, because it is most often used to construct To: headers
4307 lines in autoreplies, and the decoded form may well be syntactically
4308 invalid. However, $reply_address has leading white space removed, and all
4309 newlines turned into spaces so that the autoreply transport does not
4312 PH/05 If group was specified without a user on a router, and no group or user
4313 was specified on a transport, the group from the router was ignored.
4315 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4316 for visible compile-time settings that can be used to change these
4317 numbers, for those that want even more. Backwards compatibility with old
4318 spool files has been maintained. However, going back to a previous Exim
4319 release will lost any variables that are in spool files.
4321 PH/07 Two small changes when running in the test harness: increase delay when
4322 passing a TCP/IP connection to a new process, in case the original
4323 process has to generate a bounce, and remove special handling of
4324 127.0.0.2 (sic), which is no longer necessary.
4326 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4327 be the same on different OS.
4329 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4332 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4333 whether --show-vars was specified or not
4335 JJ/02 exipick: Added support for new ACL variable spool format introduced
4338 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4339 syntactically invalid From: or Reply-to: line, and a filter used this to
4340 generate an autoreply, and therefore failed to obtain an address for the
4341 autoreply, Exim could try to deliver to a non-existent relative file
4342 name, causing unrelated and misleading errors. What now happens is that
4343 it logs this as a hard delivery error, but does not attempt to create a
4346 PH/11 The exinext utility has a -C option for testing purposes, but although
4347 the given file was scanned by exinext itself; it wasn't being passed on
4348 when Exim was called.
4350 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4351 an end-of-file indication when reading a command response.
4353 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4354 compiled. In many other places in Exim, IPv6 addresses are always
4355 recognized, so I have changed this. It also means that IPv4 domain
4356 literals of the form [IPV4:n.n.n.n] are now always recognized.
4358 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4359 used if the router is not running as root, for example, when verifying at
4360 ACL time, or when using -bh. The debugging output from this situation was
4361 non-existent - all you got was a failure to exec. I have made two
4364 (a) Failures to set uid/gid, the current directory, or a process leader
4365 in a subprocess such as that created by queryprogram now generate
4366 suitable debugging output when -d is set.
4368 (b) The queryprogram router detects when it is not running as root,
4369 outputs suitable debugging information if -d is set, and then runs
4370 the subprocess without attempting to change uid/gid.
4372 PH/15 Minor change to Makefile for building test_host (undocumented testing
4375 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4376 additional section of a DNS packet that returns MX or SRV records.
4377 Instead, it always explicitly searches for A/AAAA records. This avoids
4378 major problems that occur when a DNS server includes only records of one
4379 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4380 fixed another bug: if SRV records were looked up and the corresponding
4381 address records were *not* found in the additional section, the port
4382 values from the SRV records were lost.
4384 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4385 using the correct key (the original address) when searching the retry
4386 rules in order to find which one to use for generating the retry hint.
4388 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4389 adding the default one. Similarly, if it contains a Reply-To: header, the
4390 errors_reply_to option, if set, is not used.
4392 PH/19 When calculating a retry time, Exim used to measure the "time since
4393 failure" by looking at the "first failed" field in the retry record. Now
4394 it does not use this if it is later than than the arrival time of the
4395 message. Instead it uses the arrival time. This makes for better
4396 behaviour in cases where some deliveries succeed, thus re-setting the
4397 "first failed" field. An example is a quota failure for a huge message
4398 when small messages continue to be delivered. Without this change, the
4399 "time since failure" will always be short, possible causing more frequent
4400 delivery attempts for the huge message than are intended.
4401 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4403 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4404 $1, $2, $3) because the numerical variables can be reset during some
4405 expansion items (e.g. "match"), thereby losing the authentication data.
4407 PH/21 Make -bV show the size of off_t variables so that the test suite can
4408 decide whether to run tests for quotas > 2G.
4410 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4411 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4412 filecount value is greater than 2G or if a quota value is greater than 2G
4413 on a system where the size of off_t is not greater than 4, a panic error
4416 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4417 never match. The debug and -bh output now contains an explicit error
4418 message indicating a malformed IPv4 address or mask.
4420 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4421 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4422 PH/23 above applies.
4424 PH/25 Do not write to syslog when running in the test harness. The only
4425 occasion when this arises is a failure to open the main or panic logs
4426 (for which there is an explicit test).
4428 PH/26 Added the /no_tell option to "control=freeze".
4430 PH/27 If a host name lookup failed very early in a connection, for example, if
4431 the IP address matched host_lookup and the reverse lookup yielded a name
4432 that did not have a forward lookup, an error message of the form "no IP
4433 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4434 could be logged. Now it outputs the IP address instead of "NULL".
4436 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4437 allows the sender and the authenticated sender to be set when
4438 submitting a message from within Exim. Since child_open_exim() is
4439 documented for local_scan(), the new function should be too.
4441 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4442 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4443 results in an empty string is now treated as unset.
4445 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4447 PH/31 Added sender_verify_fail logging option.
4449 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4450 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4451 tidied the source and removed it altogether.
4453 PH/33 When a queue run was abandoned because the load average was too high, a
4454 log line was always written; now it is written only if the queue_run log
4455 selector is set. In addition, the log line for abandonment now contains
4456 information about the queue run such as the pid. This is always present
4457 in "start" and "stop" lines but was omitted from the "abandon" line.
4459 PH/34 Omit spaces between a header name and the colon in the error message that
4460 is given when verify = headers_syntax fails (if there are lots of them,
4461 the message gets confusing).
4463 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4464 names, as there are now some PTR records that contain slashes. This check
4465 is only to protect against broken name servers that fall over on strange
4466 characters, so the fact that it applies to all lookups doesn't matter.
4468 PH/36 Now that the new test suite is complete, we can remove some of the
4469 special code in Exim that was needed for the old test suite. For example,
4470 sorting DNS records because real resolvers return them in an arbitrary
4471 order. The new test suite's fake resolver always returns records in the
4474 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4475 bounces) except when queue_only is set, to avoid logging races between
4476 the different processes.
4478 PH/38 Panic-die if .include specifies a non-absolute path.
4480 PH/39 A tweak to the "H" retry rule from its user.
4482 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4483 a label. They prevented compilation on older perls.
4485 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4486 a warning to be raised on newish perls.
4488 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4489 on queue. Changes to match documented behaviour of showing count of
4490 messages matching specified criteria.
4492 PH/40 Changed the default ident timeout from 30s to 5s.
4494 PH/41 Added support for the use of login_cap features, on those BSD systems
4495 that have them, for controlling the resources used by pipe deliveries.
4497 PH/42 The content-scanning code uses fopen() to create files in which to put
4498 message data. Previously it was not paying any attention to the mode of
4499 the files. Exim runs with umask(0) because the rest of the code creates
4500 files with open(), and sets the required mode explicitly. Thus, these
4501 files were ending up world-writeable. This was not a big issue, because,
4502 being within the spool directory, they were not world-accessible. I have
4503 created a function called modefopen, which takes an additional mode
4504 argument. It sets umask(777), creates the file, chmods it to the required
4505 mode, then resets the umask. All the relevant calls to fopen() in the
4506 content scanning code have been changed to use this function.
4508 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4509 to 24 hours. This avoids potential overflow problems when processing G
4510 and H retry rules. I suspect nobody ever tinkers with this value.
4512 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4514 PH/45 When the plaintext authenticator is running as a client, the server's
4515 challenges are checked to ensure they are valid base64 strings. By
4516 default, the authentication attempt is cancelled if an invalid string is
4517 received. Setting client_ignore_invalid_base64 true ignores these errors.
4518 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4519 they are received. Thus, the responses can be made to depend on the
4520 challenges. If an invalid string is ignored, an empty string is placed in
4523 PH/46 Messages that are created by the autoreply transport now contains a
4524 References: header, in accordance with RFCs 2822 and 3834.
4526 PH/47 Added authenticated_sender_force to the smtp transport.
4528 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4530 PH/49 Installed latest patch from the Sieve maintainer.
4532 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4533 was also set, the appendfile transport was unnecessarily scanning a
4534 directory of message files (e.g. for maildir delivery) to find the count
4535 of files (along with the size), even though it did not need this
4536 information. It now does the scan only if it needs to find either the
4537 size of the count of files.
4539 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4541 PH/52 Two bugs concerned with error handling when the smtp transport is
4544 (i) Exim was not creating retry information for temporary errors given
4545 for individual recipients after the DATA command when the smtp transport
4546 was used in LMTP mode. This meant that they could be retried too
4547 frequently, and not timed out correctly.
4549 (ii) Exim was setting the flag that allows error details to be returned
4550 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4551 recipients that were returned after the DATA command.
4553 PH/53 This is related to PH/52, but is more general: for any failing address,
4554 when detailed error information was permitted to be returned to the
4555 sender, but the error was temporary, then after the final timeout, only
4556 "retry timeout exceeded" was returned. Now it returns the full error as
4557 well as "retry timeout exceeded".
4559 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4560 do this, and (what is worse) MTAs that accept it.
4562 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4563 will now be deprecated.
4565 PH/56 New os.c-cygwin from the Cygwin maintainer.
4567 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4568 formats (previously only available in exim formats via -bpr, -bpru,
4569 and -bpra. Now also available in native and exiqgrep formats)
4571 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4572 with very large, slow to parse queues
4574 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4576 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4578 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4579 responses to authentication challenges, though it was showing the
4580 challenges; (ii) I've removed the CR characters from the debug output for
4583 PH/58 Allow for the insertion of a newline as well as a space when a string
4584 is turned into more than one encoded-word during RFC 2047 encoding. The
4585 Sieve code now uses this.
4587 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4588 data_4xx, lost_connection, tls_required.
4590 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4591 message was being sent as an SMTP response.
4593 PH/61 Add -l and -k options to exicyclog.
4595 PH/62 When verifying, if an address was redirected to one new address, so that
4596 verification continued, and the new address failed or deferred after
4597 having set something in $address_data, the value of $address_data was not
4598 passed back to the ACL. This was different to the case when no
4599 redirection occurred. The value is now passed back in both cases.
4601 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4602 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4603 use login_cap.h, so on its own it isn't the distinguishing feature. The
4604 new name refers directly to the setclassresources() function.
4606 PH/65 Added configuration files for NetBSD3.
4608 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4610 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4611 is preferred over IPv4.
4613 PH/68 The bounce_return_message and bounce_return_body options were not being
4614 honoured for bounces generated during the reception of non-SMTP messages.
4615 In particular, this applied to messages rejected by the ACL. This bug has
4616 been fixed. However, if bounce_return_message is true and bounce_return_
4617 body is false, the headers that are returned for a non-SMTP message
4618 include only those that have been read before the error was detected.
4619 (In the case of an ACL rejection, they have all been read.)
4621 PH/69 The HTML version of the specification is now built in a directory called
4622 spec_html instead of spec.html, because the latter looks like a path with
4623 a MIME-type, and this confuses some software.
4625 PH/70 Catch two compiler warnings in sieve.c.
4627 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4628 function verify_get_ident() calls ip_connect() to connect a socket, but
4629 if the "connect()" function timed out, ip_connect() used to close the
4630 socket. However, verify_get_ident() also closes the socket later, and in
4631 between Exim writes to the log, which may get opened at this point. When
4632 the socket was closed in ip_connect(), the log could get the same file
4633 descriptor number as the socket. This naturally causes chaos. The fix is
4634 not to close the socket in ip_connect(); the socket should be closed by
4635 the function that creates it. There was only one place in the code where
4636 this was missing, in the iplookup router, which I don't think anybody now
4637 uses, but I've fixed it anyway.
4639 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4640 well as to direct DNS lookups. Otherwise the handling of names in host
4641 lists is inconsistent and therefore confusing.
4647 PH/01 Two changes to the default runtime configuration:
4649 (1) Move the checks for relay_from_hosts and authenticated clients from
4650 after to before the (commented out) DNS black list checks.
4652 (2) Add control=submission to the relay_from_hosts and authenticated
4653 clients checks, on the grounds that messages accepted by these
4654 statements are most likely to be submissions.
4656 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4658 (1) Generate an error if the third argument for the ${prvs expansion is
4661 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4664 (3) Reset the variables that are obtained from the first argument of
4665 ${prvscheck and used in the second argument before leaving the code,
4666 because their memory is reclaimed, so using them afterwards may do
4669 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4670 one (it's much easier than Tom thought :-).
4672 (5) Because of (4), we can now allow for the use of $prvscheck_result
4673 inside the third argument.
4675 PH/03 For some reason, the default setting of PATH when running a command from
4676 a pipe transport was just "/usr/bin". I have changed it to
4679 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4680 anything to be listed in the output from -bV.
4682 PH/05 When a filter generated an autoreply, the entire To: header line was
4683 quoted in the delivery log line, like this:
4685 => >A.N.Other <ano@some.domain> <original@ddress> ...
4687 This has been changed so that it extracts the operative address. There
4688 may be more than one such address. If so, they are comma-separated, like
4691 => >ano@some.domain,ona@other.domain <original@ddress> ...
4693 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4694 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4695 not being looked up in the rDNS to get a host name, Exim was showing the
4696 IP address twice in Received: lines, even though the IP addresses were
4697 identical. For example:
4699 Received: from [1.2.3.4] (helo=[1.2.3.4])
4701 However, if the real host name was known, it was omitting the HELO data
4702 if it matched the actual IP address. This has been tidied up so that it
4703 doesn't show the same IP address twice.
4705 PH/07 When both +timestamp and +memory debugging was on, the value given by
4706 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4707 called by the debug printing, thereby overwriting the timestamp buffer.
4708 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4710 PH/08 When the original message was included in an autoreply transport, it
4711 always said "this is a copy of the message, including all the headers",
4712 even if body_only or headers_only was set. It now gives an appropriate
4715 PH/09 Applied a patch from the Sieve maintainer which:
4717 o fixes some comments
4718 o adds the (disabled) notify extension core
4719 o adds some debug output for the result of if/elsif tests
4720 o points to the current vacation draft in the documentation
4721 and documents the missing references header update
4725 o fixes a bug in processing the envelope test (when testing
4726 multiple envelope elements, the last element determined the
4729 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4730 Electronic Mail") by including:
4732 Auto-submitted: auto-generated
4734 in the messages that it generates (bounce messages and others, such as
4735 warnings). In the case of bounce messages for non-SMTP messages, there was
4736 also a typo: it was using "Auto_submitted" (underscore instead of
4737 hyphen). Since every message generated by Exim is necessarily in response
4738 to another message, thes have all been changed to:
4740 Auto-Submitted: auto-replied
4742 in accordance with these statements in the RFC:
4744 The auto-replied keyword:
4746 - SHOULD be used on messages sent in direct response to another
4747 message by an automatic process,
4749 - MUST NOT be used on manually-generated messages,
4751 - MAY be used on Delivery Status Notifications (DSNs) and Message
4752 Disposition Notifications (MDNs),
4754 - MUST NOT be used on messages generated by automatic or periodic
4755 processes, except for messages which are automatic responses to
4758 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4759 to the default Received: header definition.
4761 PH/12 Added log selector acl_warn_skipped (default on).
4763 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4764 variables because (a) they are in the wrong storage pool and (b) even if
4765 they were copied, it wouldn't work properly because of the caching.
4767 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4768 checking when decoding. Apparently there are clients that generate
4769 overlong encoded strings. Why am I not surprised?
4771 PH/15 If the first argument of "${if match_address" was not empty, but did not
4772 contain an "@" character, Exim crashed. Now it writes a panic log message
4773 and treats the condition as false.
4775 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4777 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4778 extension "envelope-auth". The code is finished and in agreement with
4779 other implementations, but there is no documentation so far and in fact,
4780 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4781 not changing the active code.
4783 Print executed "if" and "elsif" statements when debugging is used. This
4784 helps a great deal to understand what a filter does.
4786 Document more things not specified clearly in RFC3028. I had all this
4787 sorted out, when out of a sudden new issues came to my mind. Oops."
4789 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4792 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4793 canonical form (as documented). However, after a host name lookup from
4794 the IP address, check_host() was doing a simple string comparison with
4795 addresses acquired from the DNS when checking that the found name did
4796 have the original IP as one of its addresses. Since any found IPv6
4797 addresses are likely to be in abbreviated form, the comparison could
4798 fail. Luckily, there already exists a function for doing the comparison
4799 by converting both addresses to binary, so now that is used instead of
4800 the text comparison.
4802 PH/20 There was another similar case to PH/19, when a complete host name was
4803 given in a host list; looking up its IP address could give an abbreviated
4804 form, whereas the current host's name might or might not be abbreviated.
4805 The same fix has been applied.
4811 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4812 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4815 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4816 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4818 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4820 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4821 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4822 Also, for "Auto-Submitted", treat anything other than "no" as
4823 non-personal, in accordance with RFC 3834. (Previously it treated
4824 anything starting "auto-" as non-personal.)
4826 TF/01 The control=submission/name=... option had a problem with syntax
4827 errors if the name included a slash character. The /name= option
4828 now slurps the rest of the string, so it can include any characters
4829 but it must come last in the list of options (after /sender_retain
4832 PH/05 Some modifications to the interface to the fake nameserver for the new
4840 TK/01 Added the "success_on_redirect" address verification option. See
4841 NewStuff for rationale and an example.
4843 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4845 PH/02 Patch to exigrep to allow it to work on syslog lines.
4847 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4848 fread() to skip over the body file's header line, because in Cygwin the
4849 header line is locked and is inaccessible.
4851 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4852 co-exist for some time) to make it clear that it is the Exim ID that is
4853 referenced, not the Message-ID: header line.
4855 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4856 string_format() function, because snprintf() does not exist on all
4859 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4860 ${stat: expansion item.
4862 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4863 protocol synchronization error", to keep the pedants happy.
4865 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4866 well as for IRIX systems, when gcc is being used. See the host.c source
4869 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4871 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4874 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4875 transport and to the smtp transport in LMTP mode.
4877 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4879 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4880 run a filter in a subprocess. This could lead to confusion in subsequent
4881 lookups in the parent process. There should also be a search_tidyup() at
4882 the end of the subprocess.
4884 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4885 only if the host matched helo_try_verify_hosts, which caused the
4886 verification to occur when the EHLO/HELO command was issued. The ACL just
4887 tested the remembered result. Now, if a previous verification attempt has
4888 not happened, "verify = helo" does it there and then.
4890 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4892 TK/03 Fix log output including CR from clamd.
4894 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4895 exist provoked a memory error which could cause a segfault.
4897 PH/15 Installed PCRE 6.2
4899 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4901 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4902 of the problem. Specifically, suggested +O2 rather than +O1 for the
4905 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4907 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4908 record, the port was not correctly logged when the outgoing_port log
4909 selector was set (it logged the transort's default port).
4911 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4912 fallback_hosts, and "hosts" in the smtp transport.
4914 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4915 host errors such as "Connection refused".
4917 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4918 authentication with radiusclient 0.4.9:
4920 - Error returned from rc_read_config was caught wrongly
4921 - Username/password not passed on to radius server due to wrong length.
4923 The presumption is that some radiusclient API changes for 4.51/PH/17
4924 were not taken care of correctly. The code is still untested by me (my
4925 Linux distribution still has 0.3.2 of radiusclient), but it was
4926 contributed by a Radius user.
4928 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4929 expanding the "port" option of the smtp transport.
4931 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4932 while reading a MIME header. Thanks to Tom Hughes for a patch.
4934 PH/24 Include config.h inside local_scan.h so that configuration settings are
4937 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4938 that in an ACL for RCPT (for example), you can examine exactly what was
4941 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4942 commands, but it was not correctly comparing the address with the actual
4943 client host address. Thus, it would show the EHLO address in Received:
4944 header lines when this was not necessary.
4946 PH/27 Added the % operator to ${eval:}.
4948 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4949 it should be ignoring failures (because with -C, for example, it has lost
4950 privilege). It wasn't ignoring creation failures other than "already
4953 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4956 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4957 because an input line was too long, either on its own, or by virtue of
4958 too many continuations, the temporary file was not being removed, and the
4959 return code was incorrect.
4961 PH/31 Missing "BOOL" in function definition in filtertest.c.
4963 PH/32 Applied Sieve patches from the maintainer.
4965 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4967 PH/33 Added "verify = not_blind".
4969 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4970 Local/Makefile (with some defaults set). These are used in built scripts
4971 such as exicyclog, but they have never been used in the exim_install
4972 script (though there are many overriding facilities there). I have
4973 arranged that the exim_install script now takes note of these two
4976 PH/35 Installed configuration files for Dragonfly.
4978 PH/36 When a locally submitted message by a trusted user did not contain a
4979 From: header, and the sender address was obtained from -f or from an SMTP
4980 MAIL command, and the trusted user did not use -F to supply a sender
4981 name, $originator_name was incorrectly used when constructing a From:
4982 header. Furthermore, $originator_name was used for submission mode
4983 messages from external hosts without From: headers in a similar way,
4984 which is clearly wrong.
4986 PH/37 Added control=suppress_local_fixups.
4988 PH/38 When log_selector = +received_sender was set, and the addition of the
4989 sender made the log line's construction buffer exactly full, or one byte
4990 less than full, an overflow happened when the terminating "\n" was
4993 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4994 when the result of a list match is failure because a DNS lookup failed.
4996 PH/40 RM_COMMAND is now used in the building process.
4998 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4999 the "build-* directories that it finds.
5001 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5002 address in a domain literal was a prefix of an interface address.
5004 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5005 when verifying a sender address, unless rewrite_headers is false.
5007 PH/44 Wrote a long comment about why errors_to addresses are verified as
5008 recipients, not senders.
5010 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5011 the ratelimit ACL was added.
5013 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5015 PH/46 Added extra information about PostgreSQL errors to the error string.
5017 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5018 avoiding the need to install special zones in a real server. This is
5019 backwards compatible; if it can't find the fake resolver, it drops back.
5020 Thus, both old and new test suites can be run.
5022 TF/02 Added util/ratelimit.pl
5024 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5025 clock is set back in time.
5027 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5028 Candler <B.Candler@pobox.com>.
5030 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5031 OK for addresses that are the result of redirections.
5033 PH/48 A number of further additions for the benefit of the new test suite,
5034 including a fake gethostbyname() that interfaces to the fake DNS resolver
5037 TF/06 The fix for widen_domains has also been applied to qualify_single and
5038 search_parents which are the other dnslookup options that can cause
5041 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5044 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5046 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5047 service" if it encountered an MX record with an empty target hostname.
5048 The message is now "an MX or SRV record indicated no SMTP service".
5050 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5051 if the DNS of the sending site is misconfigured. This is quite a
5052 common situation. This change restores the behaviour of treating a
5053 helo verification defer as a failure.
5055 PH/51 If self=fail was set on a router, the bounce message did not include the
5056 actual error message.
5062 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5064 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5065 command itself ended in error, the underlying message about the transport
5066 filter timeout was being overwritten with the pipe command error. Now the
5067 underlying error message should be appended to the second error message.
5069 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5071 PH/02 Reduce the amount of output that "make" produces by default. Full output
5072 can still be requested.
5074 PH/03 The warning log line about a condition test deferring for a "warn" verb
5075 was being output only once per connection, rather than after each
5076 occurrence (because it was using the same function as for successful
5077 "warn" verbs). This seems wrong, so I have changed it.
5079 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5080 it should not have, which might have caused a crash in the right
5081 circumstances, but probably never did.
5083 PH/04 Installed a modified version of Tony Finch's patch to make submission
5084 mode fix the return path as well as the Sender: header line, and to
5085 add a /name= option so that you can make the user's friendly name appear
5088 TF/03 Added the control = fakedefer ACL modifier.
5090 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5091 Mark Lowes for thorough testing.
5093 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5095 TK/03 Merged latest SRS patch from Miles Wilton.
5097 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5098 with the definition in sysexits.h (which is #included earlier).
5099 Fortunately, Exim does not actually use EX_OK. The code used to try to
5100 preserve the sysexits.h value, by assuming that macro definitions were
5101 scanned for macro replacements. I have been disabused of this notion,
5102 so now the code just undefines EX_OK before #including unistd.h.
5104 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5105 in the smtp transport. When a block could not be written in a single
5106 write() function, the timeout was being re-applied to each part-write.
5107 This seems wrong - if the receiver was accepting one byte at a time it
5108 would take for ever. The timeout is now adjusted when this happens. It
5109 doesn't have to be particularly precise.
5111 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5112 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5114 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5115 without a domain if the check to <postmaster@domain> fails.
5117 SC/01 Eximstats: added -xls and the ability to specify output files
5118 (patch written by Frank Heydlauf).
5120 SC/02 Eximstats: use FileHandles for outputting results.
5122 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5124 SC/04 Eximstats: fixed display of large numbers with -nvr option
5126 SC/05 Eximstats: fixed merging of reports with empty tables.
5128 SC/06 Eximstats: added the -include_original_destination flag
5130 SC/07 Eximstats: removed tabs and trailing whitespace.
5132 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5134 TK/06 MBOX spool code: Add real "From " MBOX separator line
5135 so the .eml file is really in mbox format (even though
5136 most programs do not really care). Patch from Alex Miller.
5138 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5139 The latter is generated from $received_to and is only set if the
5140 message has one envelope recipient. SA can use these headers,
5141 obviously out-of-the-box. Patch from Alex Miller.
5143 PH/08 The ${def test on a variable was returning false if the variable's
5144 value was "0", contrary to what the specification has always said!
5145 The result should be true unless the variable is empty.
5147 PH/09 The syntax error of a character other than { following "${if
5148 def:variable_name" (after optional whitespace) was not being diagnosed.
5149 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5150 accidental colon was present, for example, could give incorrect results.
5152 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5153 result is used (not including appendfile, where other changes are about
5156 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5157 This involved changing a lot of size variables from int to off_t. It
5158 should work with maildirs and everything.
5160 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5161 spamd dying while we are connected to it.
5163 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5166 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5167 function for BDB 4.3.
5169 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5171 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5172 the sources for PCRE have been re-arranged and more files are now
5175 PH/15 The code I had for printing potentially long long variables in PH/11
5176 above was not the best (it lost precision). The length of off_t variables
5177 is now inspected at build time, and an appropriate printing format (%ld
5178 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5179 to be "long long int" or "long int". This is needed for the internal
5180 formatting function string_vformat().
5182 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5183 the configuration file to be ":syslog", then the script "guesses" where
5184 the logs files are, rather than using the compiled in default. In our
5185 case the guess is not the same as the compiled default, so the script
5186 suddenly stopped working when I started to use syslog. The patch checks
5187 to see if log_file_path is "". If so, it attempts to read it from exim
5188 with no configuration file to get the compiled in version, before it
5189 falls back to the previous guessing code."
5191 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5192 implementing BATV in an Exim configuration. See NewStuff for the gory
5195 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5196 Makefile that are specific to HP-UX.
5198 PH/18 If the "use_postmaster" option was set for a recipient callout together
5199 with the "random" option, the postmaster address was used as the MAIL
5200 FROM address for the random test, but not for the subsequent recipient
5201 test. It is now used for both.
5203 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5204 patch removes a few documentation additions to RFC 3028, because the
5205 latest draft now contains them. It adds the new en;ascii-case comparator
5206 and a new error check for 8bit text in MIME parts. Comparator and
5207 require names are now matched exactly. I enabled the subaddress
5208 extension, but it is not well tested yet (read: it works for me)."
5210 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5211 rework some of the code of TK/09 above to avoid the hardwired use of
5212 "%lld" and "long long". Replaced the call to snprintf() with a call to
5215 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5216 records point to non-existent hosts", "retry timeout exceeded", and
5217 "retry time not reached for any host after a long failure period".
5219 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5220 experimental DomainKeys support:
5222 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5223 (2) On an error such as an illegally used "control", the wrong name for
5224 the control was given.
5226 These problems did NOT occur unless DomainKeys support was compiled.
5228 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5230 PH/24 Added ${if match_ip condition.
5232 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5233 databases so that it will be absolutely obvious if a crash occurs in the
5234 DB library. This is a regular occurrence (often caused by mis-matched
5237 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5238 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5239 picked up on a user's system that detects such things. There doesn't seem
5240 to be a gcc warning option for this - only an attribute that has to be
5241 put on the function's prototype. It seems that in Fedora Core 4 they have
5242 set this on a number of new functions. No doubt there will be more in due
5245 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5246 specify a transport. However, if an address that was verified by such a
5247 router was the subject of a callout, Exim crashed because it tried to
5248 read the rcpt_include_affixes from the non-existent transport. Now it
5249 just assumes that the setting of that option is false. This bug was
5250 introduced by 4.51/PH/31.
5252 PH/28 Changed -d+all to exclude +memory, because that information is very
5253 rarely of interest, but it makes the output a lot bigger. People tend to
5254 do -d+all out of habit.
5256 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5257 code in os-type was giving problems when libc.so lives in lib64, like on
5260 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5261 aren't the modern standard, and it seems that some systems' include files
5262 don't always have them. Exim was already checking for some of the newer
5263 ones like T_AAAA, and defining it itself. I've added checks for all the
5264 record types that Exim uses.
5266 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5267 not automatically generating a new one, as it is supposed to. This
5268 prevented TLS from working. If the file did exist, but contained invalid
5269 data, a new version was generated, as expected. It was only the case of a
5270 non-existent file that was broken.
5272 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5273 with a change in libdomainkeys > 0.64.
5275 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5276 from DNS. If the selector record carries the flag, it now has
5277 precedence over the domain-wide flag.
5279 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5281 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5282 the use of a port name that isn't defined in /etc/services) occurred, the
5283 message was deferred as in a normal delivery, and thus remained on the
5284 spool, instead of being failed because of the mua_wrapper setting. This
5285 is now fixed, and I tidied up some of the mua_wrapper messages at the
5288 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5289 the messages in a hash of arrays rather than using individual hashes.
5290 This is a bit cleaner and results in dramatic memory savings, albeit
5291 at a slight CPU cost.
5293 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5294 as requested by Marc Sherman.
5296 SC/10 Eximstats: added histograms for user specified patterns as requested
5299 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5301 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5302 fopen() in the content-scanning modules that did not already have it.
5308 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5309 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5311 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5313 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5315 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5316 to test Sieve filters that use "vacation".
5318 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5319 that changes the way the GnuTLS parameters are stored in the cache file.
5320 The new format can be generated externally. For backward compatibility,
5321 if the data in the cache doesn't make sense, Exim assumes it has read an
5322 old-format file, and it generates new data and writes a new file. This
5323 means that you can't go back to an older release without removing the
5326 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5327 work if there are any delivery delays because "one_time" forces the
5328 parent to be marked "delivered", so its unseen clone is never tried
5329 again. For this reason, Exim now forbids the simultaneous setting of
5332 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5333 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5334 entry if you want to know the details. The fix, however, neglected to
5335 consider the case when local delivery batching is involved. The test for
5336 "previously delivered" was not happening when checking to see if an
5337 address could be batched with a previous (undelivered) one; under
5338 certain circumstances this could lead to multiple deliveries to the same
5341 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5342 in its include files, and this causes problems building Exim.
5344 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5345 header_syntax) but Exim was just ignoring anything given after a slash.
5346 In particular, this caused confusion with an attempt to use "verify =
5347 reverse_host_lookup/defer_ok". An error is now given when options are
5348 supplied for verify items that do not have them. (Maybe reverse_host_
5349 lookup should have a defer_ok option, but that's a different point.)
5351 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5352 defined by RFC 821) to 2048, because there were problems with some AUTH
5353 commands, and RFC 1869 says the size should be increased for extended
5354 SMTP commands that take arguments.
5356 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5359 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5360 "unknown" error; now it says that the functionality isn't in the binary.
5362 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5363 an address' error message when a string expansion fails (syntax or
5364 whatever). Otherwise the password may appear in the log. Following change
5365 PH/42 below, there is no longer a chance of it appearing in a bounce
5368 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5370 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5371 of its addresses was ever tried. (Bugzilla bug #2.)
5373 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5374 the result incorrectly in the debug output. (It correctly added a newline
5375 to what was transported.)
5377 TF/01 Added $received_time.
5379 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5380 commented out examples of how to interface to a virus scanner and to
5381 SpamAssassin. Also added commented examples of av_scanner and
5382 spamd_address settings.
5384 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5385 and controls are allowed in which ACLs. There were a couple of minor
5386 errors. Some of the entries in the conditions table (which is a table of
5387 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5388 negation of where the condition IS allowed.
5390 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5392 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5393 header file does not have a version number, so I've had to invent a new
5394 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5395 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5396 radiusclient), but it was contributed by a Radius user.
5398 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5399 files or directories, for OpenSSL.
5401 PH/19 When an Exim process that is running as root has to create an Exim log
5402 file, it does so in a subprocess that runs as exim:exim so as to get the
5403 ownership right at creation (otherwise, other Exim processes might see
5404 the file with the wrong ownership). There was no test for failure of this
5405 fork() call, which would lead to the process getting stuck as it waited
5406 for a non-existent subprocess. Forks do occasionally fail when resources
5407 run out. I reviewed all the other calls to fork(); they all seem to check
5410 PH/20 When checking for unexpected SMTP input at connect time (before writing
5411 the banner), Exim was not dealing correctly with a non-positive return
5412 from the read() function. If the client had disconnected by this time,
5413 the result was a log entry for a synchronization error with an empty
5414 string after "input=" when read() returned zero. If read() returned -1
5415 (an event I could not check), uninitialized data bytes were printed.
5416 There were reports of junk text (parts of files, etc) appearing after
5419 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5421 PH/22 Added support for macro redefinition, and (re)definition in between
5422 driver and ACL definitions.
5424 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5425 forgetting to use the resulting value; it was using the unexpanded value.
5427 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5428 hadn't been configured. The fix is from Juergen Kreileder, who
5429 understands it better than I do:
5431 "Here's what I see happening with three configured cyrus_sasl
5432 authenticators configured (plain, login, cram-md5):
5434 On startup auth_cyrus_sasl_init() gets called for each of these.
5435 This means three calls to sasl_listmech() without a specified mech_list.
5436 => SASL tests which mechs of all available mechs actually work
5437 => three warnings about OTP not working
5438 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5440 With the patch, sasl_listmech() also gets called three times. But now
5441 SASL's mech_list option is set to the server_mech specified in the the
5442 authenticator. Or in other words, the answer from sasl_listmech()
5443 gets limited to just the mech you're testing for (which is different
5445 => the return list contains just 'plain' or 'login', 'cram-md5' or
5446 nothing depending on the value of ob->server_mech.
5448 I've just tested the patch: Authentication still works fine,
5449 unavailable mechs specified in the exim configuration are still
5450 caught, and the auth.log warnings about OTP are gone."
5452 PH/25 When debugging is enabled, the contents of the command line are added
5453 to the debugging output, even when log_selector=+arguments is not
5456 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5457 answer is "GNU", and only if the return is "GNU/something" is the answer
5460 PH/27 $acl_verify_message is now set immediately after the failure of a
5461 verification in an ACL, and so is available in subsequent modifiers. In
5462 particular, the message can be preserved by coding like this:
5464 warn !verify = sender
5465 set acl_m0 = $acl_verify_message
5467 Previously, $acl_verify_message was set only while expanding "message"
5468 and "log_message" when a very denied access.
5470 PH/28 Modified OS/os.c-Linux with
5472 -#ifndef OS_LOAD_AVERAGE
5473 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5475 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5476 nomenclature these days.)
5478 PH/29 Installed patch from the Sieve maintainer that adds the options
5479 sieve_useraddress and sieve_subaddress to the redirect router.
5481 PH/30 In these circumstances:
5482 . Two addresses routed to the same list of hosts;
5483 . First host does not offer TLS;
5484 . First host accepts first address;
5485 . First host gives temporary error to second address;
5486 . Second host offers TLS and a TLS session is established;
5487 . Second host accepts second address.
5488 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5489 and peerdn, if requested) that were in fact used only for the second
5492 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5493 attention to any local part prefix or suffix that was matched by the
5494 router that accepted the address. It now behaves in the same way as it
5495 does for delivery: the affixes are removed from the local part unless
5496 rcpt_include_affixes is set on the transport.
5498 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5499 timeout during the DATA phase of an incoming message.
5501 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5502 applied a patch sanctioned by the Sieve maintainer.
5504 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5505 the uid or gid is negative. A case of a negative gid caused this to be
5506 noticed. The fix allows for either to be negative.
5508 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5509 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5512 PH/36 The change PH/12 above was broken. Fixed it.
5514 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5515 the grounds that routing the same address twice would always produce the
5516 same answer. This might have been true once, but it is certainly no
5517 longer true now. Routing a child address may depend on the previous
5518 routing that produced that child. Some complicated redirection strategies
5519 went wrong when messages had multiple recipients, and made Exim's
5520 behaviour dependent on the order in which the addresses were given.
5522 I have moved the duplicate checking until after the routing is complete.
5523 Exim scans the addresses that are assigned to local and remote
5524 transports, and removes any duplicates. This means that more work will be
5525 done, as duplicates will always all be routed, but duplicates are
5526 presumably rare, so I don't expect this is of any significance.
5528 For deliveries to pipes, files, and autoreplies, the duplicate checking
5529 still happens during the routing process, since they are not going to be
5532 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5533 It corrects a timeout issue with spamd. This is Ian's comment: "The
5534 background is that sometimes spamd either never reads data from a
5535 connection it has accepted, or it never writes response data. The exiscan
5536 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5537 blindly assumes that writes won't block so it may never time out."
5539 PH/39 Allow G after quota size as well as K and M.
5541 PH/40 The value set for $authenticated_id in an authenticator may not contain
5542 binary zeroes or newlines because the value is written to log lines and
5543 to spool files. There was no check on this. Now the value is run through
5544 the string_printing() function so that such characters are converted to
5545 printable escape sequences.
5547 PH/41 $message_linecount is a new variable that contains the total number of
5548 lines in the message. Compare $body_linecount, which is the count for the
5551 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5552 bounce and delay warning messages, except in certain special cases, which
5555 (a) An SMTP error message from a remote host;
5556 (b) A message specified in a :fail: redirection;
5557 (c) A message specified in a "fail" command in a system filter;
5558 (d) A message specified in a FAIL return from the queryprogram router;
5559 (e) A message specified by the cannot_route_message router option.
5561 In these cases only, Exim does include the error details in bounce and
5562 warning messages. There are also a few cases where bland messages such
5563 as "unrouteable address" or "local delivery error" are given.
5565 PH/43 $value is now also set for the "else" part of a ${run expansion.
5567 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5568 being worked on, but at least Exim now implements the latest version to
5571 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5572 process to complete was treated as a delivery failure, a timeout while
5573 writing the message to the pipe was logged, but erroneously treated as a
5574 successful delivery. Such timeouts include transport filter timeouts. For
5575 consistency with the overall process timeout, these timeouts are now
5576 treated as errors, giving rise to delivery failures by default. However,
5577 there is now a new Boolean option for the pipe transport called
5578 timeout_defer, which, if set TRUE, converts the failures into defers for
5579 both kinds of timeout. A transport filter timeout is now identified in
5582 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5583 systems where "make" and "gmake" are different, calling "gmake" at top
5584 level broke things. I've arranged for the value of $(MAKE) to be passed
5585 from the Makefile to this script so that it can call the same version of
5589 A note about Exim versions 4.44 and 4.50
5590 ----------------------------------------
5592 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5593 changes of various kinds. As a consequence, a big documentation update was
5594 needed. This delayed the release for rather longer than seemed good, especially
5595 in the light of a couple of (minor) security issues. Therefore, the changes
5596 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5597 release. So 4.44 and 4.50 are in effect two different branches that both start
5600 I have left the 4.50 change log unchanged; it contains all the changes since
5601 4.43. The change log for 4.44 is below; many of its items are identical to
5602 those for 4.50. This seems to be the most sensible way to preserve the
5603 historical information.
5609 1. Minor wording change to the doc/README.SIEVE file.
5611 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5612 computation of the current number of files was incorrect.
5614 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5615 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5618 4. Give more explanation in the error message when the command for a transport
5619 filter fails to execute.
5621 5. There are several places where Exim runs a non-Exim command in a
5622 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5623 was being done only for the command run by the queryprogram router. It is
5624 now done for all such subprocesses. The other cases are: ${run, transport
5625 filters, and the commands run by the lmtp and pipe transports.
5627 6. Added CONFIGURE_GROUP build-time option.
5629 7. Some older OS have a limit of 256 on the maximum number of file
5630 descriptors. Exim was using setrlimit() to set 1000 as a large value
5631 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5632 systems. I've change it so that if it can't get 1000, it tries for 256.
5634 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5635 was an oversight, and furthermore, ever since the addition of extra
5636 controls (e.g. 4.43/32), the checks on when to allow different forms of
5637 "control" were broken. There should now be diagnostics for all cases when a
5638 control that does not make sense is encountered.
5640 9. Added the /retain_sender option to "control=submission".
5642 10. $recipients is now available in the predata ACL (oversight).
5644 11. Tidy the search cache before the fork to do a delivery from a message
5645 received from the command line. Otherwise the child will trigger a lookup
5646 failure and thereby defer the delivery if it tries to use (for example) a
5647 cached ldap connection that the parent has called unbind on.
5649 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5650 of $address_data from the recipient verification was clobbered by the
5651 sender verification.
5653 13. The value of address_data from a sender verification is now available in
5654 $sender_address_data in subsequent conditions in the ACL statement.
5656 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5658 15. Added a new option "connect=<time>" to callout options, to set a different
5661 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5662 was its contents. (It was OK if the option was not defined at all.)
5664 17. A "Completed" log line is now written for messages that are removed from
5665 the spool by the -Mrm option.
5667 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5668 information about exactly what failed.
5670 19. Added -dd to debug only the daemon process.
5672 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5673 handles timeouts, both on the server side and network timeouts. Renamed the
5674 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5676 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5677 It is now set to "smtps".
5679 22. $host_address is now set to the target address during the checking of
5680 ignore_target_hosts.
5682 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5683 being passed; this would have caused $sender_host_name to have been used if
5684 matching the list had actually called for a host name (not very likely,
5685 since this list is usually IP addresses). A host name is now passed as
5688 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5689 code that specifies a non-restarting handler (typically sigaction() in
5690 modern systems) in an attempt to fix a rare and obscure crash bug.
5692 25. Narrowed the window for a race in the daemon that could cause it to ignore
5693 SIGCHLD signals. This is not a major problem, because they are used only to
5694 wake it up if nothing else does.
5696 26. A malformed maildirsize file could cause Exim to calculate negative values
5697 for the mailbox size or file count. Odd effects could occur as a result.
5698 The maildirsize information is now recalculated if the size or filecount
5701 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5702 support for a long time. Removed HAVE_SYS_VFS_H.
5704 28. Installed the latest version of exipick from John Jetmore.
5706 29. In an address list, if the pattern was not a regular expression, an empty
5707 subject address (from a bounce message) matched only if the pattern was an
5708 empty string. Non-empty patterns were not even tested. This was the wrong
5709 because it is perfectly reasonable to use an empty address as part of a
5710 database query. An empty address is now tested by patterns that are
5711 lookups. However, all the other forms of pattern expect the subject to
5712 contain a local part and a domain, and therefore, for them, an empty
5713 address still always fails if the pattern is not itself empty.
5715 30. Exim went into a mad DNS loop when attempting to do a callout where the
5716 host was specified on an smtp transport, and looking it up yielded more
5717 than one IP address.
5719 31. Re-factored the code for checking spool and log partition space into a
5720 function that finds that data and another that does the check. The former
5721 is then used to implement four new variables: $spool_space, $log_space,
5722 $spool_inodes, and $log_inodes.
5724 32. The RFC2047 encoding function was originally intended for short strings
5725 such as real names; it was not keeping to the 75-character limit for
5726 encoded words that the RFC imposes. It now respects the limit, and
5727 generates multiple encoded words if necessary. To be on the safe side, I
5728 have increased the buffer size for the ${rfc2047: expansion operator from
5731 33. It is now permitted to omit both strings after an "if" condition; if the
5732 condition is true, the result is "true". As before, when the second string
5733 is omitted, a false condition yields an empty string. This makes it less
5734 cumbersome to write custom ACL and router conditions.
5736 34. Failure to deliver a bounce message always caused it to be frozen, even if
5737 there was an errors_to setting on the router. The errors_to setting is now
5740 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5741 canonical form (fully expanded) before being placed in
5742 $sender_host_address.
5744 36. The table in the code that translates DNS record types into text (T_A to
5745 "A" for instance) was missing entries for NS and CNAME. It is just possible
5746 that this could have caused confusion if both these types were looked up
5747 for the same domain, because the text type is used as part of Exim's
5748 per-process caching. But the chance of anyone hitting this buglet seems
5751 37. The dnsdb lookup has been extended in a number of ways.
5753 (1) There is a new type, "zns", which walks up the domain tree until it
5754 finds some nameserver records. It should be used with care.
5756 (2) There is a new type, "mxh", which is like "mx" except that it returns
5757 just the host names, not the priorities.
5759 (3) It is now possible to give a list of domains (or IP addresses) to be
5760 looked up. The behaviour when one of the lookups defers can be
5761 controlled by a keyword.
5763 (4) It is now possible to specify the separator character for use when
5764 multiple records are returned.
5766 38. The dnslists ACL condition has been extended: it is now possible to supply
5767 a list of IP addresses and/or domains to be looked up in a particular DNS
5770 39. Added log_selector=+queue_time_overall.
5772 40. When running the queue in the test harness, wait just a tad after forking a
5773 delivery process, to get repeatability of debugging output.
5775 41. Include certificate and key file names in error message when GnuTLS fails
5776 to set them up, because the GnuTLS error message doesn't include the name
5777 of the failing file when there is a problem reading it.
5779 42. Allow both -bf and -bF in the same test run.
5781 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5783 44. The "Exiscan patch" is now merged into the mainline Exim source.
5785 45. Sometimes the final signoff response after QUIT could fail to get
5786 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5787 < 0 before doing a fflush(). This bug looks as though it goes back to the
5788 introduction of TLS in release 3.20, but "sometimes" must have been rare
5789 because the tests only now provoked it.
5791 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5792 (this can affect the format of dates).
5794 47. exim_tidydb, when checking for the continued existence of a message for
5795 which it has found a message-specific retry record, was not finding
5796 messages that were in split spool directories. Consequently, it was
5797 deleting retry records that should have stayed in existence.
5799 48. Steve fixed some bugs in eximstats.
5801 49. The SPA authentication driver was not abandoning authentication and moving
5802 on to the next authenticator when an expansion was forced to fail,
5803 contradicting the general specification for all authenticators. Instead it
5804 was generating a temporary error. It now behaves as specified.
5806 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5807 (the order specifies the preference for clients). The order is now AES256,
5808 AES128, 3DES, ARCFOUR128.
5810 51. Small patch to Sieve code - explicitly set From: when generating an
5813 52. Exim crashed if a remote delivery caused a very long error message to be
5814 recorded - for instance if somebody sent an entire SpamAssassin report back
5815 as a large number of 550 error lines. This bug was coincidentally fixed by
5816 increasing the size of one of Exim's internal buffers (big_buffer) that
5817 happened as part of the Exiscan merge. However, to be on the safe side, I
5818 have made the code more robust (and fixed the comments that describe what
5821 53. Now that there can be additional text after "Completed" in log lines (if
5822 the queue_time_overall log selector is set), a one-byte patch to exigrep
5823 was needed to allow it to recognize "Completed" as not the last thing in
5826 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5827 patch that reportedly fixes this has been added. I am not expert enough to
5828 create a test for it. This is what the patch creator wrote:
5830 "I found a little strange behaviour of ldap code when working with
5831 Windows 2003 AD Domain, where users was placed in more than one
5832 Organization Units. When I tried to give exim partial DN, the exit code
5833 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5834 But simultaneously result of request was absolutely normal ldap result,
5835 so I produce this patch..."
5837 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5838 REFERENCE, so I have modified the code to exclude the patch when that macro
5841 55. Some experimental protocols are using DNS PTR records for new purposes. The
5842 keys for these records are domain names, not reversed IP addresses. The
5843 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5844 leaves it alone. Component reversal etc. now happens only for IP addresses.
5847 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5849 57. Double the size of the debug message buffer (to 2048) so that more of very
5850 long debug lines gets shown.
5852 58. The exicyclog utility now does better if the number of log files to keep
5853 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5855 59. Two changes related to the smtp_active_hostname option:
5857 (1) $smtp_active_hostname is now available as a variable.
5858 (2) The default for smtp_banner uses $smtp_active_hostname instead
5859 of $primary_hostname.
5861 60. The host_aton() function is supposed to be passed a string that is known
5862 to be a valid IP address. However, in the case of IPv6 addresses, it was
5863 not checking this. This is a hostage to fortune. Exim now panics and dies
5864 if the condition is not met. A case was found where this could be provoked
5865 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5866 components; fortuitously, this particular loophole had already been fixed
5867 by change 4.50/55 above.
5869 If there are any other similar loopholes, the new check in host_aton()
5870 itself should stop them being exploited. The report I received stated that
5871 data on the command line could provoke the exploit when Exim was running as
5872 exim, but did not say which command line option was involved. All I could
5873 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5874 running as the user.
5877 61. There was a buffer overflow vulnerability in the SPA authentication code
5878 (which came originally from the Samba project). I have added a test to the
5879 spa_base64_to_bits() function which I hope fixes it.
5882 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5883 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5885 63. The daemon start-up calls getloadavg() while still root for those OS that
5886 need the first call to be done as root, but it missed one case: when
5887 deliver_queue_load_max is set with deliver_drop_privilege. This is
5888 necessary for the benefit of the queue runner, because there is no re-exec
5889 when deliver_drop_privilege is set.
5891 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5892 This has been fixed.
5894 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5895 was in use, was not putting the data itself into the right store pool;
5896 consequently, it could be overwritten for a subsequent message in the same
5897 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5900 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5902 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5903 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5904 the function were treating the return as a boolean value, which happened to
5905 work because 0=false and not-0=true, but is not correct code.
5907 68. The host_aton() function was not handling scoped IPv6 addresses (those
5908 with, for example, "%eth0" on the end) correctly.
5910 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5911 negated items (that is, ~something) in unsigned ints. Some compilers
5912 apparently mutter when there is no cast.
5914 70. If an address verification called from an ACL failed, and did not produce a
5915 user-specific message (i.e. there was only a "system" message), nothing was
5916 put in $acl_verify_message. In this situation, it now puts the system
5919 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5920 session; change 4.31/43 added the unwanted input to the log line - except
5921 that it did not do this in the start of session case. It now does.
5923 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5924 This is wrong and can cause the other end to generate a synchronization
5925 error if it is another Exim or anything else that does the synchronization
5926 check. A QUIT command is no longer sent after a timeout.
5928 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5929 during host lookups.
5931 74. The defer_ok option of callout verification was not working if it was used
5932 when verifying addresses in header lines, that is, for this case:
5934 verify = header_sender/callout=defer_ok
5936 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5937 those file descriptors could be used for SMTP connections. If anything
5938 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5939 could be sent to the SMTP client, causing chaos. The daemon now opens
5940 stdin, stdout, and stderr to /dev/null when it puts itself into the
5943 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5944 log by default. The user can override this with suitable Perl magic.
5946 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5947 the log message when discard triggers, was not working for the DATA ACL or
5948 for the non-SMTP ACL.
5950 78. Error message wording change in sieve.c.
5952 79. If smtp_accept_max_per_host was set, the number of connections could be
5953 restricted to fewer than expected, because the daemon was trying to set up
5954 a new connection before checking whether the processes handling previous
5955 connections had finished. The check for completed processes is now done
5956 earlier. On busy systems, this bug wouldn't be noticed because something
5957 else would have woken the daemon, and it would have reaped the completed
5960 80. If a message was submitted locally by a user whose login name contained one
5961 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5962 It caused a spool format error. I have fixed the spool reading code. A
5963 related problem was that the "from" clause in the Received: line became
5964 illegal because of the space(s). It is now covered by ${quote_local_part.
5966 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5969 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5971 83. Patch from Sieve maintainer to fix unterminated string problem in
5972 "vacation" handling.
5974 84. Some minor changes to the Linux configuration files to help with other
5975 OS variants using glibc.
5977 85. One more patch for Sieve to update vacation handling to latest spec.
5980 ----------------------------------------------------
5981 See the note above about the 4.44 and 4.50 releases.
5982 ----------------------------------------------------
5988 1. Change 4.43/35 introduced a bug that caused file counts to be
5989 incorrectly computed when quota_filecount was set in an appendfile
5992 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5993 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5996 3. Give more explanation in the error message when the command for a transport
5997 filter fails to execute.
5999 4. There are several places where Exim runs a non-Exim command in a
6000 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6001 was being done only for the command run by the queryprogram router. It is
6002 now done for all such subprocesses. The other cases are: ${run, transport
6003 filters, and the commands run by the lmtp and pipe transports.
6005 5. Some older OS have a limit of 256 on the maximum number of file
6006 descriptors. Exim was using setrlimit() to set 1000 as a large value
6007 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6008 systems. I've change it so that if it can't get 1000, it tries for 256.
6010 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6011 was an oversight, and furthermore, ever since the addition of extra
6012 controls (e.g. 4.43/32), the checks on when to allow different forms of
6013 "control" were broken. There should now be diagnostics for all cases when a
6014 control that does not make sense is encountered.
6016 7. $recipients is now available in the predata ACL (oversight).
6018 8. Tidy the search cache before the fork to do a delivery from a message
6019 received from the command line. Otherwise the child will trigger a lookup
6020 failure and thereby defer the delivery if it tries to use (for example) a
6021 cached ldap connection that the parent has called unbind on.
6023 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6024 of $address_data from the recipient verification was clobbered by the
6025 sender verification.
6027 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6028 was its contents. (It was OK if the option was not defined at all.)
6030 11. A "Completed" log line is now written for messages that are removed from
6031 the spool by the -Mrm option.
6033 12. $host_address is now set to the target address during the checking of
6034 ignore_target_hosts.
6036 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6037 being passed; this would have caused $sender_host_name to have been used if
6038 matching the list had actually called for a host name (not very likely,
6039 since this list is usually IP addresses). A host name is now passed as
6042 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6043 code that specifies a non-restarting handler (typically sigaction() in
6044 modern systems) in an attempt to fix a rare and obscure crash bug.
6046 15. Narrowed the window for a race in the daemon that could cause it to ignore
6047 SIGCHLD signals. This is not a major problem, because they are used only to
6048 wake it up if nothing else does.
6050 16. A malformed maildirsize file could cause Exim to calculate negative values
6051 for the mailbox size or file count. Odd effects could occur as a result.
6052 The maildirsize information is now recalculated if the size or filecount
6055 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6056 support for a long time. Removed HAVE_SYS_VFS_H.
6058 18. Updated exipick to current release from John Jetmore.
6060 19. Allow an empty sender to be matched against a lookup in an address list.
6061 Previously the only cases considered were a regular expression, or an
6064 20. Exim went into a mad DNS lookup loop when doing a callout where the
6065 host was specified on the transport, if the DNS lookup yielded more than
6068 21. The RFC2047 encoding function was originally intended for short strings
6069 such as real names; it was not keeping to the 75-character limit for
6070 encoded words that the RFC imposes. It now respects the limit, and
6071 generates multiple encoded words if necessary. To be on the safe side, I
6072 have increased the buffer size for the ${rfc2047: expansion operator from
6075 22. Failure to deliver a bounce message always caused it to be frozen, even if
6076 there was an errors_to setting on the router. The errors_to setting is now
6079 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6080 canonical form (fully expanded) before being placed in
6081 $sender_host_address.
6083 24. Updated eximstats to version 1.33
6085 25. Include certificate and key file names in error message when GnuTLS fails
6086 to set them up, because the GnuTLS error message doesn't include the name
6087 of the failing file when there is a problem reading it.
6089 26. Expand error message when OpenSSL has problems setting up cert/key files.
6092 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6093 (this can affect the format of dates).
6095 28. exim_tidydb, when checking for the continued existence of a message for
6096 which it has found a message-specific retry record, was not finding
6097 messages that were in split spool directories. Consequently, it was
6098 deleting retry records that should have stayed in existence.
6100 29. eximstats updated to version 1.35
6101 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6102 1.35 - bugfix such that pie charts by volume are generated correctly
6104 30. The SPA authentication driver was not abandoning authentication and moving
6105 on to the next authenticator when an expansion was forced to fail,
6106 contradicting the general specification for all authenticators. Instead it
6107 was generating a temporary error. It now behaves as specified.
6109 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6110 (the order specifies the preference for clients). The order is now AES256,
6111 AES128, 3DES, ARCFOUR128.
6113 31. Small patch to Sieve code - explicitly set From: when generating an
6116 32. Exim crashed if a remote delivery caused a very long error message to be
6117 recorded - for instance if somebody sent an entire SpamAssassin report back
6118 as a large number of 550 error lines. This bug was coincidentally fixed by
6119 increasing the size of one of Exim's internal buffers (big_buffer) that
6120 happened as part of the Exiscan merge. However, to be on the safe side, I
6121 have made the code more robust (and fixed the comments that describe what
6124 33. Some experimental protocols are using DNS PTR records for new purposes. The
6125 keys for these records are domain names, not reversed IP addresses. The
6126 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6127 leaves it alone. Component reversal etc. now happens only for IP addresses.
6130 34. The host_aton() function is supposed to be passed a string that is known
6131 to be a valid IP address. However, in the case of IPv6 addresses, it was
6132 not checking this. This is a hostage to fortune. Exim now panics and dies
6133 if the condition is not met. A case was found where this could be provoked
6134 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6135 components; fortuitously, this particular loophole had already been fixed
6136 by change 4.50/55 or 4.44/33 above.
6138 If there are any other similar loopholes, the new check in host_aton()
6139 itself should stop them being exploited. The report I received stated that
6140 data on the command line could provoke the exploit when Exim was running as
6141 exim, but did not say which command line option was involved. All I could
6142 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6143 running as the user.
6146 35. There was a buffer overflow vulnerability in the SPA authentication code
6147 (which came originally from the Samba project). I have added a test to the
6148 spa_base64_to_bits() function which I hope fixes it.
6151 36. The daemon start-up calls getloadavg() while still root for those OS that
6152 need the first call to be done as root, but it missed one case: when
6153 deliver_queue_load_max is set with deliver_drop_privilege. This is
6154 necessary for the benefit of the queue runner, because there is no re-exec
6155 when deliver_drop_privilege is set.
6157 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6158 was in use, was not putting the data itself into the right store pool;
6159 consequently, it could be overwritten for a subsequent message in the same
6160 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6163 38. Sometimes the final signoff response after QUIT could fail to get
6164 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6165 < 0 before doing a fflush(). This bug looks as though it goes back to the
6166 introduction of TLS in release 3.20, but "sometimes" must have been rare
6167 because the tests only now provoked it.
6173 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6174 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6175 Now it returns an integer. A number of calls were still expecting a T/F
6176 return. Fortuitously, in all cases, the tests worked in OK situations,
6177 which is the norm. However, things would have gone wrong on any write
6178 failures on the smtp file descriptor. This function is used when sending
6179 messages over SMTP and also when doing verify callouts.
6181 2. When Exim is called to do synchronous delivery of a locally submitted
6182 message (the -odf or -odi options), it no longer closes stderr before doing
6185 3. Implemented the mua_wrapper option.
6187 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6189 5. Implemented the functions header_remove(), header_testname(),
6190 header_add_at_position(), and receive_remove_recipient(), and exported them
6193 6. If an ACL "warn" statement specified the addition of headers, Exim already
6194 inserted X-ACL-Warn: at the start if there was no header name. However, it
6195 was not making this test for the second and subsequent header lines if
6196 there were newlines in the string. This meant that an invalid header could
6197 be inserted if Exim was badly configured.
6199 7. Allow an ACL "warn" statement to add header lines at the start or after all
6200 the Received: headers, as well as at the end.
6202 8. Added the rcpt_4xx retry error code.
6204 9. Added postmaster_mailfrom=xxx to callout verification option.
6206 10. Added mailfrom=xxxx to the callout verification option, for verify=
6209 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6210 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6212 12. Inserted some casts to stop certain compilers warning when using pointer
6213 differences as field lengths or precisions in printf-type calls (mostly
6214 affecting debugging statements).
6216 13. Added optional readline() support for -be (dynamically loaded).
6218 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6219 same clock tick as a message's arrival, so that its received time was the
6220 same as the "first fail" time on the retry record, and that message
6221 remained on the queue past the ultimate address timeout, every queue runner
6222 would try a delivery (because it was past the ultimate address timeout) but
6223 after another failure, the ultimate address timeout, which should have then
6224 bounced the address, did not kick in. This was a "< instead of <=" error;
6225 in most cases the first failure would have been in the next clock tick
6226 after the received time, and all would be well.
6228 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6229 being recognized when the domain list was tested by the match_domain
6230 condition in an expansion string.
6232 16. Added the ${str2b64: operator.
6234 17. Exim was always calling setrlimit() to set a large limit for the number of
6235 processes, without checking whether the existing limit was already
6236 adequate. (It did check for the limit on file descriptors.) Furthermore,
6237 errors from getrlimit() and setrlimit() were being ignored. Now they are
6238 logged to the main and panic logs, but Exim does carry on, to try to do its
6239 job under whatever limits there are.
6241 18. Imported PCRE 5.0.
6243 19. Trivial typo in log message " temporarily refused connection" (the leading
6246 20. If the log selector return_path_on_delivery was set and an address was
6247 redirected to /dev/null, the delivery process crashed because it assumed
6248 that a return path would always be set for a "successful" delivery. In this
6249 case, the whole delivery is bypassed as an optimization, and therefore no
6252 21. Internal re-arrangement: the function for sending a challenge and reading
6253 a response while authentication was assuming a zero-terminated challenge
6254 string. It's now changed to take a pointer and a length, to allow for
6255 binary data in such strings.
6257 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6259 23. Exim was not respecting finduser_retries when seeking the login of the
6260 uid under which it was called; it was always trying 10 times. (The default
6261 setting of finduser_retries is zero.) Also, it was sleeping after the final
6262 failure, which is pointless.
6264 24. Implemented tls_on_connect_ports.
6266 25. Implemented acl_smtp_predata.
6268 26. If the domain in control=submission is set empty, Exim assumes that the
6269 authenticated id is a complete email address when it generates From: or
6270 Sender: header lines.
6272 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6273 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6274 chown and chgrp in /bin and hostname in /usr/bin.
6276 28. Exim was keeping the "process log" file open after each use, just as it
6277 does for the main log. This opens the possibility of it remaining open for
6278 long periods when the USR1 signal hits a daemon. Occasional processlog
6279 errors were reported, that could have been caused by this. Anyway, it seems
6280 much more sensible not to leave this file open at all, so that is what now
6283 29. The long-running daemon process does not normally write to the log once it
6284 has entered its main loop, and it closes the log before doing so. This is
6285 so that log files can straightforwardly be renamed and moved. However,
6286 there are a couple of unusual error situations where the daemon does write
6287 log entries, and I had neglected to close the log afterwards.
6289 30. The text of an SMTP error response that was received during a remote
6290 delivery was being truncated at 512 bytes. This is too short for some of
6291 the long messages that one sometimes sees. I've increased the limit to
6294 31. It is now possible to make retry rules that apply only when a message has a
6295 specific sender, in particular, an empty sender.
6297 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6298 it possible to be selective about when SMTP synchronization is enforced.
6300 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6302 32. Implemented hosts_connection_nolog.
6304 33. Added an ACL for QUIT.
6306 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6309 35. Added mailbox_size and mailbox_filecount to appendfile.
6311 36. Added control = no_multiline_responses to ACLs.
6313 37. There was a bug in the logic of the code that waits for the clock to tick
6314 in the case where the clock went backwards by a substantial amount such
6315 that the microsecond fraction of "now" was more than the microsecond
6316 fraction of "then" (but the whole seconds number was less).
6318 38. Added support for the libradius Radius client library this is found on
6319 FreeBSD (previously only the radiusclient library was supported).
6325 1. When certain lookups returned multiple values in the form name=value, the
6326 quoting of the values was not always being done properly. Specifically:
6327 (a) If the value started with a double quote, but contained no whitespace,
6329 (b) If the value contained whitespace other than a space character (i.e.
6330 tabs or newlines or carriage returns) it was not quoted.
6331 This fix has been applied to the mysql and pgsql lookups by writing a
6332 separate quoting function and calling it from the lookup code. The fix
6333 should probably also be applied to nisplus, ibase and oracle lookups, but
6334 since I cannot test any of those, I have not disturbed their existing code.
6336 2. A hit in the callout cache for a specific address caused a log line with no
6337 reason for rejecting RCPT. Now it says "Previous (cached) callout
6338 verification failure".
6340 3. There was an off-by-one bug in the queryprogram router. An over-long
6341 return line was truncated at 256 instead of 255 characters, thereby
6342 overflowing its buffer with the terminating zero. As well as fixing this, I
6343 have increased the buffer size to 1024 (and made a note to document this).
6345 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6346 when Exim is waiting for an SMTP response from a remote server, Exim
6347 restarts its select() call on the socket, thereby resetting its timeout.
6348 This is not a problem when such interrupts are rare. Somebody set up a cron
6349 job to run exiwhat every 2 minutes, which is less than the normal select()
6350 timeout (5 or 10 minutes). This meant that the select() timeout never
6351 kicked in because it was always reset. I have fixed this by comparing the
6352 time when an interrupt arrives with the time at the start of the first call
6353 to select(). If more time than the timeout has elapsed, the interrupt is
6354 treated as a timeout.
6356 5. Some internal re-factoring in preparation for the addition of Sieve
6357 extensions (by MH). In particular, the "personal" test is moved to a
6358 separate function, and given an option for scanning Cc: and Bcc: (which is
6359 not set for Exim filters).
6361 6. When Exim created an email address using the login of the caller as the
6362 local part (e.g. when creating a From: or Sender: header line), it was not
6363 quoting the local part when it contained special characters such as @.
6365 7. Installed new OpenBSD configuration files.
6367 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6368 try to make them clearer.
6370 9. Callout options, other than the timeout value, were being ignored when
6371 verifying sender addresses in header lines. For example, when using
6373 verify = header_sender/callout=no_cache
6375 the cache was (incorrectly) being used.
6377 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6378 only the Cygwin environment.
6380 11. When return_path_on_delivery was set as a log selector, if different remote
6381 addresses in the same message used different return paths and parallel
6382 remote delivery occurred, the wrong values would sometimes be logged.
6383 (Whenever a remote delivery process finished, the return path value from
6384 the most recently started remote delivery process was logged.)
6386 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6387 lines when AUTH and/or TLS are in use. This is the "received protocol"
6388 field. Exim used to use "asmtp" for authenticated SMTP, without any
6389 indication (in the protocol name) for TLS use. Now it follows the RFC and
6390 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6391 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6392 names appear in log lines as well as in Received: header lines.
6394 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6395 extensions, and comparison tests, and to fix some bugs.
6397 14. Changes to the "personal" filter test:
6399 (1) The test was buggy in that it was just doing the equivalent of
6400 "contains" tests on header lines. For example, if a user's address was
6401 anne@some.where, the "personal" test would incorrectly be true for
6403 To: susanne@some.where
6405 This test is now done by extracting each address from the header in turn,
6406 and checking the entire address. Other tests that are part of "personal"
6407 are now done using regular expressions (for example, to check local parts
6408 of addresses in From: header lines).
6410 (2) The list of non-personal local parts in From: addresses has been
6411 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6412 taken from the Sieve specification recommendations.
6414 (3) If the message contains any header line starting with "List-" it is
6415 treated as non-personal.
6417 (4) The test for "circular" in the Subject: header line has been removed
6418 because it now seems ill-conceived.
6420 15. Minor typos in src/EDITME comments corrected.
6422 16. Installed latest exipick from John Jetmore.
6424 17. If headers_add on a router specified a text string that was too long for
6425 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6426 of string_sprintf() is now avoided.
6428 18. $message_body_size was not set (it was always zero) when running the DATA
6429 ACL and the local_scan() function.
6431 19. For the "mail" command in an Exim filter, no default was being set for
6432 the once_repeat time, causing a random time value to be used if "once" was
6433 specified. (If the value happened to be <= 0, no repeat happened.) The
6434 default is now 0s, meaning "never repeat". The "vacation" command was OK
6435 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6436 (I found it when inspecting the code).
6438 20. There is now an overall timeout for performing a callout verification. It
6439 defaults to 4 times the callout timeout, which applies to individual SMTP
6440 commands during the callout. The overall timeout applies when there is more
6441 than one host that can be tried. The timeout is checked before trying the
6442 next host. This prevents very long delays if there are a large number of
6443 hosts and all are timing out (e.g. when the network connections are timing
6444 out). The value of the overall timeout can be changed by specifying an
6445 additional sub-option for "callout", called "maxwait". For example:
6447 verify = sender/callout=5s,maxwait=20s
6449 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6450 to the end before writing, but this should make it even safer).
6452 22. Exim was forgetting that it had advertised PIPELINING for the second and
6453 subsequent messages on an SMTP connection. It was also not resetting its
6454 memory on STARTTLS and an internal HELO.
6456 23. When Exim logs an SMTP synchronization error within a session, it now
6457 records whether PIPELINING has been advertised or not.
6459 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6460 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6461 rather than long int.
6463 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6465 26. Added the never_mail option to autoreply.
6471 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6472 crash if the getsockname() call failed; this can happen if a connection is
6473 closed very soon after it is established. The problem was simply in the
6474 order in which certain operations were done, causing Exim to try to write
6475 to the SMTP stream before it had set up the file descriptor. The bug has
6476 been fixed by making things happen in the correct order.
6482 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6483 before the connection was closed, thus losing the rejection response.
6485 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6486 some early Solaris releases, but causes trouble in current releases where
6487 socklen_t is defined.
6489 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6492 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6495 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6496 that releasing the top of it at the end releases what was used for sub-
6497 expansions (unless the block got too big). However, discard this block if
6498 the first thing is a variable or header, so that we can use its block when
6499 it is dynamic (useful for very large $message_headers, for example).
6501 6. Lookups now cache *every* query, not just the most recent. A new, separate
6502 store pool is used for this. It can be recovered when all lookup caches are
6503 flushed. Lookups now release memory at the end of their result strings.
6504 This has involved some general refactoring of the lookup sources.
6506 7. Some code has been added to the store_xxx() functions to reduce the amount
6507 of flapping under certain conditions.
6509 8. log_incoming_interface used to affect only the <= reception log lines. Now
6510 it causes the local interface and port to be added to several more SMTP log
6511 lines, for example "SMTP connection from", and rejection lines.
6513 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6515 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6517 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6518 could be overwritten at the end of the current message (or the start of a
6519 new message if it was set in a HELO ACL). The value is now preserved for
6520 the duration of the SMTP connection.
6522 12. If a transport had a headers_rewrite setting, and a matching header line
6523 contained an unqualified address, that address was qualified, even if it
6524 did not match any rewriting rules. The underlying bug was that the values
6525 of the flags that permit the existence of unqualified sender and recipient
6526 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6527 non-local messages, and by -bnq for local messages) were not being
6528 preserved with the message after it was received.
6530 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6531 "next input=" as part of the text comprising the host identity instead of
6532 the correct text. The code was using the same buffer for two different
6533 strings. However, depending on which order the printing function evaluated
6534 its arguments, the bug did not always show up. Under Linux, for example, my
6535 test suite worked just fine.
6537 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6538 doesn't work with some older versions of Perl. It has been changed to "my",
6539 which in any case is probably the better facility to use.
6541 15. A really picky compiler found some instances of statements for creating
6542 error messages that either had too many or two few arguments for the format
6545 16. The size of the buffer for calls to the DNS resolver has been increased
6546 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6547 for addresses that have a lot of PTR records. This alleviates a problem; it
6548 does not fully solve it.
6550 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6551 buffer now truncates the list and logs the incident, which is the same
6552 action as happens when Exim is looking up a host name and its aliases.
6553 Previously in this situation something unpredictable would happen;
6554 sometimes it was "internal error: store_reset failed".
6556 18. If a server dropped the connection unexpectedly when an Exim client was
6557 using GnuTLS and trying to read a response, the client delivery process
6558 crashed while trying to generate an error log message.
6560 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6561 string, for example:
6563 warn message = H1: something\nH2: something
6565 the text was added as a single header line from Exim's point of view
6566 though it ended up OK in the delivered message. However, searching for the
6567 second and subsequent header lines using $h_h2: did not work. This has been
6568 fixed. Similarly, if a system filter added multiple headers in this way,
6569 the routers could not see them.
6571 20. Expanded the error message when iplsearch is called with an invalid key to
6572 suggest using net-iplsearch in a host list.
6574 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6575 ACLs are no longer actually imposed (and a message to that effect is
6578 22. If a "gecos" field in a passwd entry contained escaped characters, in
6579 particular, if it contained a \" sequence, Exim got it wrong when building
6580 a From: or a Sender: header from that name. A second bug also caused
6581 incorrect handling when an unquoted " was present following a character
6582 that needed quoting.
6584 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6585 was not being matched caselessly.
6587 24. Arranged for all hyphens in the exim.8 source to be escaped with
6590 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6591 properly. Recipient callout cache records were still being keyed to include
6592 the sender, even when use_sender was set false. This led to far more
6593 callouts that were necessary. The sender is no longer included in the key
6594 when use_sender is false.
6596 26. Added "control = submission" modifier to ACLs.
6598 27. Added the ${base62d: operator to decode base 62 numbers.
6600 28. dnsdb lookups can now access SRV records.
6602 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6603 the configuration file.
6605 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6606 (-v) mode. This makes the output for a verbose queue run more intelligible.
6608 31. Added a use_postmaster feature to recipient callouts.
6610 32. Added the $body_zerocount variable, containing the number of binary zero
6611 bytes in the message body.
6613 33. The time of last modification of the "new" subdirectory is now used as the
6614 "mailbox time last read" when there is a quota error for a maildir
6617 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6619 35. Added +ignore_unknown as a special item in host lists.
6621 36. Code for decoding IPv6 addresses in host lists is now included, even if
6622 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6623 address was recognized as an IP address, but was then not correctly decoded
6624 into binary, causing unexpected and incorrect effects when compared with
6631 1. Very minor rewording of debugging text in manualroute to say "list of
6632 hosts" instead of "hostlist".
6634 2. If verify=header_syntax was set, and a header line with an unqualified
6635 address (no domain) and a large number of spaces between the end of the
6636 name and the colon was received, the reception process suffered a buffer
6637 overflow, and (when I tested it) crashed. This was caused by some obsolete
6638 code that should have been removed. The fix is to remove it!
6640 3. When running in the test harness, delay a bit after writing a bounce
6641 message to get a bit more predictability in the log output.
6643 4. Added a call to search_tidyup() just before forking a reception process. In
6644 theory, someone could use a lookup in the expansion of smtp_accept_max_
6645 per_host which, without the tidyup, could leave open a database connection.
6647 5. Added the variables $recipient_data and $sender_data which get set from a
6648 lookup success in an ACL "recipients" or "senders" condition, or a router
6649 "senders" option, similar to $domain_data and $local_part_data.
6651 6. Moved the writing of debug_print from before to after the "senders" test
6654 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6655 problems for message scanning, either using a data ACL, or using
6656 local_scan() because the Received: header was not generated till after they
6657 were called (in order to set the time as the time of reception completion).
6658 I have revised the way this works. The header is now generated after the
6659 body is received, but before the ACL or local_scan() are called. After they
6660 are run, the timestamp in the header is updated.
6666 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6667 before starting a queue runner without re-exec. This happened only when
6668 deliver_drop_privilege was set or when the Exim user was set to root. The
6669 effect of the bug was that timeouts during subsequent deliveries caused
6670 crashes instead of being properly handled. The handler is now left at its
6671 default (and expected) setting.
6673 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6674 message, again when deliver_drop_privilege is set or Exim is run as root.
6675 The bug described in (1) was not present in this case, but the tidying up
6676 of the other signals was missing. I have made the two cases consistent.
6678 3. The ignore_target_hosts setting on a manualroute router was being ignored
6679 for hosts that were looked up using the /MX notation.
6681 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6684 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6685 operated on the sender address. After changing the $sender_address to <>
6686 for the sender address verify, Exim was re-instated it as the original
6687 (before rewriting) address, but remembering that it had rewritten it, so it
6688 wasn't rewriting it again. This bug also had the effect of breaking the
6689 sender address verification caching when the sender address was rewritten.
6691 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6692 This has been changed so that if the ip literal address matches
6693 ignore_target_hosts, the router declines.
6695 7. Added expansion conditions match_domain, match_address, and match_local_
6696 part (NOT match_host).
6698 8. The placeholder for the Received: header didn't have a length field set.
6700 9. Added code to Exim itself and to exim_lock to test for a specific race
6701 condition that could lead to file corruption when using MBX delivery. The
6702 issue is with the lockfile that is created in /tmp. If this file is removed
6703 after a process has opened it but before that process has acquired a lock,
6704 there is the potential for a second process to recreate the file and also
6705 acquire a lock. This could lead to two Exim processes writing to the file
6706 at the same time. The added code performs the same test as UW imapd; it
6707 checks after acquiring the lock that its file descriptor still refers to
6708 the same named file.
6710 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6711 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6714 11. Added the smtp_active_hostname option. If used, this will typically be made
6715 to depend on the incoming interface address. Because $interface_address is
6716 not set up until the daemon has forked a reception process, error responses
6717 that can happen earlier (such as "too many connections") no longer contain
6720 12. If an expansion in a condition on a "warn" statement fails because a lookup
6721 defers, the "warn" statement is abandoned, and the next ACL statement is
6722 processed. Previously this caused the whole ACL to be aborted.
6724 13. Added the iplsearch lookup type.
6726 14. Added ident_timeout as a log selector.
6728 15. Added tls_certificate_verified as a log selector.
6730 16. Added a global option tls_require_ciphers (compare the smtp transport
6731 option of the same name). This controls incoming TLS connections.
6733 17. I finally figured out how to make tls_require_ciphers do a similar thing
6734 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6735 before starting the TLS session.
6737 18. Tabs are now shown as \t in -bP output.
6739 19. If the log selector return_path_on_delivery was set, Exim crashed when
6740 bouncing a message because it had too many Received: header lines.
6742 20. If two routers both had headers_remove settings, and the first one included
6743 a superfluous trailing colon, the final name in the first list and the
6744 first name in the second list were incorrectly joined into one item (with a
6745 colon in the middle).
6751 1. Added -C and -D options to the exinext utility, mainly to make it easier
6752 to include in the automated testing, but these could be helpful when
6753 multiple configurations are in use.
6755 2. The exinext utility was not formatting the output nicely when there was
6756 an alternate port involved in the retry record key, nor when there was a
6757 message id as well (for retries that were specific to a specific message
6758 and a specific host). It was also confused by IPv6 addresses, because of
6759 the additional colons they contain. I have fixed the IPv4 problem, and
6760 patched it up to do a reasonable job for IPv6.
6762 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6763 delivery, the log line now contains "pipelined" if PIPELINING was used.
6765 4. An SMTP transport process used to panic and die if the bind() call to set
6766 an explicit outgoing interface failed. This has been changed; it is now
6767 treated in the same way as a connect() failure.
6769 5. A reference to $sender_host_name in the part of a conditional expansion
6770 that was being skipped was still causing a DNS lookup. This no longer
6773 6. The def: expansion condition was not recognizing references to header lines
6774 that used bh_ and bheader_.
6776 7. Added the _cache feature to named lists.
6778 8. The code for checking quota_filecount in the appendfile transport was
6779 allowing one more file than it should have been.
6781 9. For compatibility with Sendmail, the command line option
6789 and sets the incoming protocol and host name (for trusted callers). The
6790 host name and its colon can be omitted when only the protocol is to be set.
6791 Note the Exim already has two private options, -pd and -ps, that refer to
6792 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6793 "s", but I don't think that's a major issue.
6795 10. A number of refactoring changes to the code, none of which should affect
6798 (a) The number of logging options was getting close to filling up the
6799 32-bit word that was used as a bit map. I have split them into two classes:
6800 those that are passed in the argument to log_write(), and those that are
6801 only ever tested independently outside of that function. These are now in
6802 separate 32-bit words, so there is plenty of room for expansion again.
6803 There is no change in the user interface or the logging behaviour.
6805 (b) When building, for example, log lines, the code previously used a
6806 macro that called string_cat() twice, in order to add two strings. This is
6807 not really sufficiently general. Furthermore, there was one instance where
6808 it was actually wrong because one of the argument was used twice, and in
6809 one call a function was used. (As it happened, calling the function twice
6810 did not affect the overall behaviour.) The macro has been replaced by a
6811 function that can join an arbitrary number of extra strings onto a growing
6814 (c) The code for expansion conditions now uses a table and a binary chop
6815 instead of a serial search (which was left over from when there were very
6816 few conditions). Also, it now recognizes conditions like "pam" even when
6817 the relevant support is not compiled in: a suitably worded error message is
6818 given if an attempt is made to use such a condition.
6820 11. Added ${time_interval:xxxxx}.
6822 12. A bug was causing one of the ddress fields not to be passed back correctly
6823 from remote delivery subprocesses. The field in question was not being
6824 subsequently used, so this caused to problems in practice.
6826 13. Added new log selectors queue_time and deliver_time.
6828 14. Might have fixed a bug in maildirsizefile handling that threw up
6829 "unexpected character" debug warnings, and recalculated the data
6830 unnecessarily. In any case, I expanded the warning message to give more
6833 15. Added the message "Restricted characters in address" to the statements in
6834 the default ACL that block characters like @ and % in local parts.
6836 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6837 Three changes have been made:
6839 (a) There was a serious bug; a negative response to MAIL caused the whole
6840 recipient domain to be cached as invalid, thereby blocking all messages
6841 to all local parts at the same domain, from all senders. This bug has
6842 been fixed. The domain is no longer cached after a negative response to
6843 MAIL if the sender used is not empty.
6845 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6848 (c) A new callout option, "use_sender" has been added for people who want
6849 the modified behaviour.
6855 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6858 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6859 indeed breaks things for older releases.
6861 3. Added additional logging to the case where there is a problem reading data
6862 from a filter that is running in a subprocess using a pipe, in order to
6863 try to track down a specific problem.
6865 4. Testing facility fudge: when running in the test harness and attempting
6866 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6867 getting "No route to host". Convert this to a timeout.
6869 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6872 6. Some OS don't have socklen_t but use size_t instead. This affects the
6873 fifth argument of getsockopt() amongst other things. This is now
6874 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6875 can be set for individual OS. I have set it for SunOS5, OSF1, and
6876 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6877 some earlier ones do not.
6879 7. Change 4.30/15 was not doing the test caselessly.
6881 8. The standard form for an IPv6 address literal was being rejected by address
6882 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6883 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6884 this, as well as the form without the "IPv6" on the front (but only when
6885 address literals are enabled, of course).
6887 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6889 10. Exim crashed if a message with an empty sender address specified by -f
6890 encountered a router with an errors_to setting. This could be provoked only
6891 by a command such as
6895 where an empty string was supplied; "<>" did not hit this bug.
6897 11. Installed PCRE release 4.5.
6899 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6900 remained set. It is now erased.
6902 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6903 times from message ids (which are base 36 rather than the normal 62).
6905 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6906 were being counted as actual protocol errors, and logged if the log
6907 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6908 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6909 and DATA following a set of rejected RCPTs do not count as protocol errors.
6910 In other words, Exim assumes they were pipelined, though this may not
6911 actually be the case. Of course, in all cases the client gets an
6912 appropriate error code.
6914 15. If a lookup fails in an ACL condition, a message about the failure may
6915 be available; it is used if testing the ACL cannot continue, because most
6916 such messages specify what the cause of the deferral is. However, some
6917 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6918 that caused an old message to be retained and used if a later statement
6919 caused a defer, replacing the real cause of the deferral.
6921 16. If an IP address had so many PTR records that the DNS lookup buffer
6922 was not large enough to hold them, Exim could crash while trying to process
6923 the truncated data. It now detects and logs this case.
6925 17. Further to 4.21/58, another change has been made: if (and only if) the
6926 first line of a message (the first header line) ends with CRLF, a bare LF
6927 in a subsequent header line has a space inserted after it, so as not to
6928 terminate the header.
6930 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6931 unnecessarily, used atoi() instead of strtol(), and didn't check the
6932 termination when getting file sizes from file names by regex.
6934 19. Completely re-implemented the support for maildirsize files, in the light
6935 of a number of problems with the previous contributed implementation
6936 (4.30/29). In particular:
6938 . If the quota is zero, the maildirsize file is maintained, but no quota is
6941 . If the maildir directory does not exist, it is created before any attempt
6942 to write a maildirsize file.
6944 . The quota value in the file is just a cache; if the quota is changed in
6945 the transport, the new value overrides.
6947 . A regular expression is available for excluding directories from the
6950 20. The autoreply transport checks the characters in options that define the
6951 message's headers; it allows continued headers, but it was checking with
6952 isspace() after an embedded newline instead of explicitly looking for a
6955 21. If all the "regular" hosts to which an address was routed had passed their
6956 expiry times, and had not reached their retry times, the address was
6957 bounced, even if fallback hosts were defined. Now Exim should go on to try
6960 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6961 equivalent code in the SMTP transport. Some hosts send humungous responses
6962 to HELO/EHLO, more than 1024 it seems.
6964 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6965 gives compiler warnings in some environments. I've now done it "properly",
6968 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6969 (because of problems with the built-in one) was declared to return uschar *
6970 instead of char *, causing compiler failure.
6972 25. Fixed a file descriptor leak when processing alias/forward files.
6974 26. Fixed a minor format string issue in dbfn.c.
6976 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6978 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6979 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6982 29. When a sender address is verified, it is cached, to save repeating the test
6983 when there is more than one recipient in a message. However, when the
6984 verification involves a callout, it is possible for different callout
6985 options to be set for different recipients. It is too complicated to keep
6986 track of this in the cache, so now Exim always runs a verification when a
6987 callout is required, relying on the callout cache for the optimization.
6988 The overhead is duplication of the address routing, but this should not be
6991 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6992 to be verified with callout=postmaster, and the main callout worked but the
6993 postmaster check failed, the verification correctly failed. However, if a
6994 subsequent RCPT command asked for sender verification *without* the
6995 postmaster check, incorrect caching caused this verification also to fail,
6998 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6999 it was not caching the DNS options (qualify_single, search_parents) that
7000 were used when the lookup failed. A subsequent lookup with different
7001 options therefore always gave the same answer, though there were cases
7002 where it should not have. (Example: a "domains = !$mx_any" option on a
7003 dnslookup router: the "domains" option is always processed without any
7004 widening, but the router might have qualify_single set.) Now Exim uses the
7005 cached value only when the same options are set.
7007 32. Added John Jetmore's "exipick" utility to the distribution.
7009 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7010 than a timeout (e.g. a certificate is required, and is not provided), an
7011 Exim server now closes the connection immediately. Previously it waited for
7012 the client to close - but if the client is SSL, it seems that they each
7013 wait for each other, leading to a delay before one of them times out.
7015 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7016 maintained 0.8.x compatibility because I don't think many are using it, and
7017 it is clearly obsolete.
7019 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7022 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7023 client certificate was expired. A simple patch fixes this, though I don't
7024 understand the full logic of why the verify callback is called multiple
7027 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7028 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7029 which causes problems with some clients (such as the Certicom SSL Plus
7030 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7031 disables the coutermeasure allowing Eudora to connect."
7033 38. Exim was not checking that a write() to a log file succeeded. This could
7034 lead to Bad Things if a log got too big, in particular if it hit a file
7035 size limit. Exim now panics and dies if it cannot write to a log file, just
7036 as it does if it cannot open a log file.
7038 39. Modified OS/Makefile-Linux so that it now contains
7040 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7042 The two -D definitions ensure that Exim is compiled with large file
7043 support, which makes it possible to handle log files that are bigger than
7046 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7047 instance) a domain was checked against a named list that involved a lookup,
7048 causing $domain_data to be set, then another domain was checked against the
7049 same list, then the first domain was re-checked, the value of $domain_data
7050 after the final check could be wrong. In particular, if the second check
7051 failed, it could be set empty. This bug probably also applied to
7054 41. The strip_trailing_dot option was not being applied to the address given
7055 with the -f command-line option.
7057 42. The code for reading a message's header from the spool was incrementing
7058 $received_count, but never initializing it. This meant that the value was
7059 incorrect (doubled) while delivering a message in the same process in which
7060 it was received. In the most common configuration of Exim, this never
7061 happens - a fresh exec is done - but it can happen when
7062 deliver_drop_privilege is set.
7064 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7065 it now includes up to 150 characters of the unexpected data in the log
7068 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7069 and building data strings. The size of both of these buffers was 10 000
7070 bytes - far larger than anybody would *ever* want, thought I. Needless to
7071 say, somebody hit the limit. I have increased the maximum line length to
7072 20 000 and the maximum data length of concatenated lines to 100 000. I have
7073 also fixed two bugs, because there was no checking on these buffers. Tsk,
7074 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7075 buffer is too small.
7077 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7078 lsearch lookups. Now it does.
7080 46. When parsing a route_list item in a manualroute router, a fixed-length
7081 buffer was used for the list of hosts. I made this 1024 bytes long,
7082 thinking that nobody would ever have a list of hosts that long. Wrong.
7083 Somebody had a whole pile of complicated expansion conditions, and the
7084 string was silently truncated, leading to an expansion error. It turns out
7085 that it is easier to change to an unlimited length (owing to other changes
7086 that have happened since this code was originally written) than to build
7087 structure for giving a limitation error. The length of the item that
7088 expands into the list of hosts is now unlimited.
7090 47. The lsearch lookup could not handle data where the length of text line was
7091 more than 4095 characters. Such lines were truncated, leading to shortened
7092 data being returned. It should now handle lines of any length.
7094 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7095 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7098 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7099 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7100 no longer happens in comment lines. A list of replacements is now placed
7101 at the head of all of the source files, except those whose only change is
7102 to replace PERL_COMMAND in the very first #! line.
7104 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7105 messages on the queue, with a bottom-up merge sort, using code contributed
7106 by Michael Haardt. This should make operations like -bp somewhat faster on
7107 large queues. It won't affect queue runners, except when queue_run_in_order
7110 51. Installed eximstats 1.31 in the distribution.
7112 52. Added support for SRV lookups to the dnslookup router.
7114 53. If an ACL referred to $message_body or $message_body_end, the value was not
7115 reset for any messages that followed in the same SMTP session.
7117 54. The store-handling optimization for building very long strings was not
7118 differentiating between the different store pools. I don't think this
7119 actually made any difference in practice, but I've tidied it.
7121 55. While running the routers to verify a sender address, $sender_address
7122 was still set to the sender address. This is wrong, because when routing to
7123 send a bounce to the sender, it would be empty. Therefore, I have changed
7124 it so that, while verifying a sender address, $sender_address is set to <>.
7125 (There is no change to what happens when verifying a recipient address.)
7127 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7128 target A or AAAA records (if not already returned) without resetting the
7129 qualify_single or search_parents options of the DNS resolver. These are
7130 inappropriate in this case because the targets of MX and SRV records must
7131 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7132 target that, when qualified, matched something in the local domain. These
7133 two options are now turned off when doing these lookups.
7135 57. It seems that at least some releases of Reiserfs (which does not have the
7136 concept of a fixed number of inodes) returns zero and not -1 for the
7137 number of available inodes. This interacted badly with check_spool_inodes,
7138 which assumed that -1 was the "no such thing" setting. What I have done is
7139 to check that the total number of inodes is greater than zero before doing
7140 the test of how many are available.
7142 58. When a "warn" ACL statement has a log_message modifier, the message is
7143 remembered, and not repeated. This is to avoid a lot of repetition when a
7144 message has many recipients that cause the same warning to be written.
7145 However, Exim was preserving the list of already written lines for an
7146 entire SMTP session, which doesn't seem right. The memory is now reset if a
7147 new message is started.
7149 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7150 debugging output unless log_rewrite was also set.
7152 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7153 of (int)(handle) when we know that handle contains (void *)(-1).
7155 61. The Exim daemon panic-logs an error return when it closes the incoming
7156 connection. However "connection reset by peer" seems to be common, and
7157 isn't really an error worthy of noting specially, so that particular error
7160 62. When Exim is trying to find all the local interfaces, it used to panic and
7161 die if the ioctl to get the interface flags failed. However, it seems that
7162 on at least one OS (Solaris 9) it is possible to have an interface that is
7163 included in the list of interfaces, but for which you get a failure error
7164 for this call. This happens when the interface is not "plumbed" into a
7165 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7166 failure of the "get flags" call assumes that the interface is down.
7168 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7169 makes life easier for people who are doing arithmetic on fields extracted
7170 from dates, where you often get leading zeros that should not be
7171 interpreted as octal.
7173 64. Added qualify_domain to the redirect router, to override the global
7176 65. If a pathologically long header line contained very many addresses (the
7177 report of this problem mentioned 10 000) and each of them was rewritten,
7178 Exim could use up a very large amount of memory. (It kept on making new
7179 copies of the header line as it rewrote, and never released the old ones.)
7180 At the expense of a bit more processing, the header rewriting function has
7181 been changed so that it no longer eats memory in this way.
7183 66. The generation of the Received: header has been moved from the time that a
7184 message starts to be received, to the time that it finishes. The timestamp
7185 in the Received: header should now be very close to that of the <= log
7186 line. There are two side-effects of this change:
7188 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7189 logged header lines no longer include the local Received: line, because
7190 it has not yet been created. The same applies to a copy of the message
7191 that is returned to a non-SMTP sender when a message is rejected.
7193 (b) When a filter file is tested using -bf, no additional Received: header
7194 is added to the test message. After some thought, I decided that this
7197 This change does not affect the value of $received_for. It is still set
7198 after address rewriting, but before local_scan() is called.
7200 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7202 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7203 gave an unhelpful panic error message, and a defer error. I have managed to
7204 change this behaviour so that it now rejects any supplied certificate,
7205 which seems right, as the list of acceptable certificates is empty.
7207 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7208 gave an unhelpful defer error. I have not managed to make this reject any
7209 supplied certificates, but the error message it gives is "no certificate
7210 supplied", which is not helpful.
7212 70. exigrep's output now also includes lines that are not associated with any
7213 message, but which match the given pattern. Implemented by a patch from
7214 Martin Sluka, which also tidied up the Perl a bit.
7216 71. Recipient callout verification, like sender verification, was using <> in
7217 the MAIL FROM command. This isn't really the right thing, since the actual
7218 sender may affect whether the remote host accepts the recipient or not. I
7219 have changed it to use the actual sender in the callout; this means that
7220 the cache record is now keyed on a recipient/sender pair, not just the
7221 recipient address. There doesn't seem to be a real danger of callout loops,
7222 since a callout by the remote host to check the sender would use <>.
7223 [SEE ABOVE: changed after hitting problems.]
7225 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7226 temporary errors. However, in the case of such a code being given after
7227 the end of a data transmission (i.e. after ".") Exim was failing to write
7228 a retry record for the message. (Yes, there was some broken host that was
7229 actually sending 8xx at this point.)
7231 73. An unknown lookup type in a host list could cause Exim to panic-die when
7232 the list was checked. (An example that provoked this was putting <; in the
7233 middle of a list instead of at the start.) If this happened during a DATA
7234 ACL check, a -D file could be left lying around. This kind of configuration
7235 error no longer causes Exim to die; instead it causes a defer error. The
7236 incident is still logged to the main and panic logs.
7238 74. Buglet left over from Exim 3 conversion. The message "too many messages
7239 in one connection" was written to the rejectlog but not the mainlog, except
7240 when address rewriting (yes!) was being logged.
7242 75. Added write_rejectlog option.
7244 76. When a system filter was run not as root (that is, when system_filter_user
7245 was set), the values of the $n variables were not being returned to the
7246 main process; thus, they were not subsequently available in the $sn
7249 77. Added +return_path_on_delivery log selector.
7251 78. A connection timeout was being treated differently from recipients deferred
7252 when testing hosts_max_try with a message that was older than the host's
7253 retry timeout. (The host should not be counted, thus allowing all hosts to
7254 be tried at least once before bouncing.) This may have been the cause of an
7255 occasionally reported bug whereby a message would remain on the queue
7256 longer than the retry timeout, but would be bounced if a delivery was
7257 forced. I say "may" because I never totally pinned down the problem;
7258 setting up timeout/retry tests is difficult. See also the next item.
7260 79. The ultimate address timeout was not being applied to errors that involved
7261 a combination of host plus message (for example, a timeout on a MAIL
7262 command). When an address resolved to a number of possible hosts, and they
7263 were not all tried for each delivery (e.g. because of hosts_max_try), a
7264 message could remain on the queue longer than the retry timeout.
7266 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7269 81. Fixed an obscure SMTP outgoing bug which required at least the following
7270 conditions: (a) there was another message waiting for the same server;
7271 (b) the server returned 5xx to all RCPT commands in the first message so
7272 that the message was not completed; (c) the server dropped the connection
7273 or gave a negative response to the RSET that Exim sends to abort the
7274 transaction. The observed case was a dropped connection after DATA that had
7275 been sent in pipelining mode. That is, the server had advertised PIPELINING
7276 but was not implementing it correctly. The effect of the bug was incorrect
7277 behaviour, such as trying another host, and this could lead to a crash.
7283 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7284 and daemon.c were passed as pointers to ints; they should have been
7285 pointers to socklen_t variables (which are typically unsigned ints).
7287 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7290 3. Fixed a really odd bug that affected only the testing scheme; patching a
7291 certain fixed string in the binary changed the value of another string that
7292 happened to be identical to the end of the original first string.
7294 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7295 name", it returns that address as the IP address. On some operating
7296 systems (e.g. Solaris), it also passes back the IP address string as the
7297 "host name". However, on others (e.g. Linux), it passes back an empty
7298 string. Exim wasn't checking for this, and was changing the host name to an
7299 empty string, assuming it had been canonicalized.
7301 5. Although rare, it is permitted to have more than one PTR record for a given
7302 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7303 all the names associated with an address, because they do in Solaris.
7304 However, it seems that they do not in Linux for data that comes from the
7305 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7306 I found this out when I moved to a new Linux workstation and tried to run
7307 the Exim test suite.
7309 To get round this problem I have changed the code so that it now does its
7310 own call to the DNS to look up PTR records when searching for a host name.
7311 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7312 addresses that are only in /etc/hosts are still found.
7314 This behaviour is, however, controlled by an option called host_lookup_
7315 order, which defaults to "bydns:byaddr". If people want to use the other
7316 order, or indeed, just use one or the other means of lookup, they can
7317 specify it in this variable.
7319 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7320 some operating systems, this comes back from gethostbyaddr() as an empty
7321 string, and this is what Exim used to test for. However, it seems that in
7322 other systems, "." is yielded. Exim now tests for this case too.
7324 7. The values of check_spool_space and check_log_space are now held internally
7325 as a number of kilobytes instead of an absolute number of bytes. If a
7326 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7327 kilobyte. This means that much larger values can be stored.
7329 8. Exim monitor: an attempt to get the action menu when not actually pointing
7330 at a message produces an empty menu entitled "No message selected". This
7331 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7332 no entries in it ("Shell widget menu has zero width and/or height"). So I
7333 have added a single, blank menu entry in this case.
7335 9. Added ${quote_local_part.
7337 10. MIME decoding is now applied to the contents of Subject: header lines when
7340 11. Now that a reference to $sender_host_address automatically causes a reverse
7341 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7342 host lookup before query-style lookups in lists that might use this
7343 variable. This has therefore been abolished, and the "net-" prefix is no
7344 longer necessary for query-style lookups.
7346 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7347 have been a typo for LFLAGS, so it has been changed.
7349 13. The install script calls Exim with "-C /dev/null" in order to find the
7350 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7351 to be output. However, since Exim outputs its version number before the
7352 error, it didn't break the script. It just looked ugly. I fixed this by
7353 always allowing "-C /dev/null" if the caller is root.
7355 14. Ignore overlarge ACL variable number when reading spool file - insurance
7356 against a later release with more variables having written the file.
7358 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7359 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7360 this, as well as the form without the "IPv6" on the front.
7362 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7363 OS/Makefile-Darwin file.
7365 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7366 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7368 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7369 variable. However, some OS use ints for inodes. I've added cast to long int
7370 to get rid of the compiler warning.
7372 19. I had forgotten to lock out "/../" in configuration file names when
7373 ALT_CONFIG_PREFIX was set.
7375 20. Routers used for verification do not need to specify transports. However,
7376 if such a router generated a host list, and callout was configured, Exim
7377 crashed, because it could not find a port number from the (non-existent)
7378 transport. It now assumes port 25 in this circumstance.
7380 21. Added the -t option to exigrep.
7382 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7383 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7384 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7385 them is set, LOOKUP_LSEARCH is forced.
7387 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7389 24. Added sender and host information to the "rejected by local_scan()" log
7390 line; previously there was no indication of these.
7392 25. Added .include_if_exists.
7394 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7395 Linux. It turns out that not all file systems support this. Apparently some
7396 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7397 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7398 not supported on the file descriptor, is now ignored when Exim is trying to
7399 sync a directory. This applies only to Linux.
7401 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7403 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7404 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7407 29. Added support for maildirsize files from supplied patch (modified a bit).
7409 30. The use of :fail: followed by an empty string could lead Exim to respond to
7410 sender verification failures with (e.g.):
7412 550 Verification failed for <xxx>
7413 550 Sender verify failed
7415 where the first response line was missing the '-' that indicates it is not
7416 the final line of the response.
7418 31. The loop for finding the name of the user that called Exim had a hardwired
7419 limit of 10; it now uses the value of finduser_retries, which is used for
7420 all other user lookups.
7422 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7425 33. Exim was neglecting to zero errno before one call of strtol() when
7426 expanding a string and expecting an integer value. On some systems this
7427 resulted in spurious "integer overflow" errors. Also, it was casting the
7428 result into an int without checking.
7430 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7431 did not work. The code looks as if it has *never* worked, though it appears
7432 to have been documented since at least release 1.62. I have made it work.
7434 35. The "timeout_DNS" error in retry rules, also documented since at least
7435 1.62, also never worked. As it isn't clear exactly what this means, and
7436 clearly it isn't a major issue, I have abolished the feature by treating it
7437 as "timeout", and writing a warning to the main and panic logs.
7439 36. The display of retry rules for -brt wasn't always showing the error code
7442 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7443 timeout_connect_A, timeout_connect_MX.
7445 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7446 to the empty sender.
7448 39. The daemon was not analysing the content of -oX till after it had closed
7449 stderr and disconnected from the controlling terminal. This meant that any
7450 syntax errors were only noted on the panic log, and the return code from
7451 the command was 0. By re-arranging the code a little, I've made the
7452 decoding happen first, so such errors now appear on stderr, and the return
7453 code is 1. However, the actual setting up of the sockets still happens in
7454 the disconnected process, so errors there are still only recorded on the
7457 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7458 connections (as happens on some IP stacks) was logged at start up time as
7459 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7460 it from "IPv6 and IPv4", which means that two separate sockets are being
7463 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7464 says whether AF_INET or AF_INET6 was passed as an argument.
7466 42. Exiwhat output was messed up when time zones were included in log
7469 43. Exiwhat now gives more information about the daemon's listening ports,
7470 and whether -tls-on-connect was used.
7472 44. The "port" option of the smtp transport is now expanded.
7474 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7475 silently ignored. Now an error message is written to the main and panic
7478 46. There's a new ACL modifier called "logwrite" which writes to a log file
7479 as soon as it is encountered.
7481 47. Added $local_user_uid and $local_user_gid at routing time.
7483 48. Exim crashed when trying to verify a sender address that was being
7486 49. Exim was recognizing only a space character after ".include". It now also
7487 recognizes a tab character.
7489 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7490 extracting the relevant information from the specification. The man page no
7491 longer contains scrambled data for the -d option, and I've added a section
7492 at the front about calling Exim under different names.
7494 51. Added "extra_headers" argument to the "mail" command in filter files.
7496 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7499 53. Installed eximstats 1.29.
7501 54. Added transport_filter_timeout as a generic transport option.
7503 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7504 Cc: header lines. This was required by RFC 822, but it not required by RFC
7507 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7508 incoming messages that did not have them. Now it does so only if the
7509 message originates locally, that is, if there is no associated remote host
7510 address. When Resent- header lines are present, this applies to the Resent-
7511 lines rather than the non-Resent- lines.
7513 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7514 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7516 58. Messages for configuration errors now include the name of the main
7517 configuration file - useful now that there may be more than one file in a
7518 list (.included file names were always shown).
7520 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7521 for those rare installations that do not start the daemon as root or run it
7522 setuid root. I've cut out the call to initgroups() if the daemon is not
7525 60. The Exim user and group can now be bound into the binary as text strings
7526 that are looked up at the start of Exim's processing.
7528 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7530 62. Added $mailstore_basename variable.
7532 63. Installed patch to sieve.c from Michael Haardt.
7534 64. When Exim failed to open the panic log after failing to open the main log,
7535 the original message it was trying to log was written to stderr and debug
7536 output, but if they were not available (the usual case in production), it
7537 was lost. Now it is written to syslog before the two lines that record the
7538 failures to open the logs.
7540 65. Users' Exim filters run in subprocesses under the user's uid. It is
7541 possible for a "deliver" command or an alias in a "personal" command to
7542 provoke an address rewrite. If logging of address rewriting is configured,
7543 this fails because the process is not running as root or exim. There may be
7544 a better way of dealing with this, but for the moment (because 4.30 needs
7545 to be released), I have disabled address rewrite logging when running a
7546 filter in a non-root, non-exim process.
7552 1. The buildconfig auxiliary program wasn't quoting the value set for
7553 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7554 not defined. This bug was masked in 4.22 by the effect that was fixed in
7557 2. Some messages that were rejected after a message id was allocated were
7558 shown as "incomplete" by exigrep. It no longer does this for messages that
7559 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7561 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7562 have allow_domain_literals set, the ID did not get logged in the <= line.
7563 Domain literals are now always recognized in Message-ID: header lines.
7565 4. The first argument for a ${extract expansion item is the key name or field
7566 number. Leading and trailing spaces in this item were not being ignored,
7567 causing some misleading effects.
7569 5. When deliver_drop_privilege was set, single queue runner processes started
7570 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7571 same command in the process it spins off) were not dropping privilege.
7573 6. When the daemon running as "exim" started a queue runner, it always
7574 re-executed Exim in the spun-off process. This is a waste of effort when
7575 deliver_drop_privilege is set. The new process now just calls the
7576 queue-runner function directly.
7582 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7585 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7586 ignored. Though the use of -oP was forcing the writing of a pid file, it
7587 was always written to the default place.
7589 3. If the message "no IP address found for host xxxx" is generated during
7590 incoming verification, it is now followed by identification of the incoming
7591 connection (so you can more easily find what provoked it).
7593 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7595 5. Added some features to "harden" Exim a bit more against certain attacks:
7597 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7598 be put in Local/Makefile. This is like the never_users runtime option,
7599 but it cannot be overridden. The default setting is "root".
7601 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7602 prefix string with which any file named in a -C command line option
7605 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7606 is retained for -C and -D only if the caller of Exim is root. Without
7607 it, the exim user may also use -C and -D and retain privilege.
7609 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7610 command line option is disabled.
7612 6. Macro names set by the -D option must start with an upper case letter, just
7613 like macro names defined in the configuration file.
7615 7. Added "dereference=" facility to LDAP.
7617 8. Two instances of the typo "uknown" in the source files are fixed.
7619 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7620 the Configure-Makefile script screwed up while processing it.
7622 10. Incorporated PCRE 4.4.
7624 11. The SMTP synchronization check was not operating right at the start of an
7625 SMTP session. For example, it could not catch a HELO sent before the client
7626 waited for the greeting. There is now a check for outstanding input at the
7627 point when the greeting is written. Because of the duplex, asynchronous
7628 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7629 way, but not yet received, when the check is performed.
7631 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7632 on TCP/IP sockets, because this apparently causes some broken clients to
7635 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7636 unchanged) from the Cygwin maintainer.
7638 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7639 was supported instead of testing for mbx. Effectively a typo.
7641 15. The spa authenticator server code was not checking that the input it
7642 received was valid base64.
7644 16. The debug output line for the "set" modifier in ACLs was not showing the
7645 name of the variable that was being set.
7647 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7649 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7650 lookup on its own. Something else was needed to trigger the lookup. For
7651 example, a match in host_lookup or the need for a host name in a host list.
7652 Now, if $sender_host_name is referenced and the host name has not yet been
7653 looked up, a lookup is performed. If the lookup fails, the variable remains
7654 empty, and $host_lookup_failed is set to "1".
7656 19. Added "eqi" as a case-independent comparison operator.
7658 20. The saslauthd authentication condition could segfault if neither service
7659 nor realm was specified.
7661 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7662 error message that was logged was misleading, and incoming SMTP
7663 connections were dropped. The message is now more accurate, and temporary
7664 errors are given to SMTP connections.
7666 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7667 (except RSET) with a 421 error, until QUIT is received. However, it was
7668 failing to send a response to QUIT.
7670 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7671 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7672 ACL is now run after the helo_try_verify_hosts code.
7674 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7675 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7676 case-independent, so other case variants are also recognized). Apparently
7677 some systems use these upper case variants.
7679 25. If more than two messages were waiting for the same host, and a transport
7680 filter was specified for the transport, Exim sent two messages over the
7681 same TCP/IP connection, and then failed with "socket operation on non-
7682 socket" when it tried to send the third.
7684 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7686 27. The extern definition of crypt16() in expand.c was not being excluded when
7687 the OS had its own crypt16() function.
7689 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7690 as a preferred synonym for return_size_limit, both as an option and as an
7693 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7695 30. Changed the default configuration ACL to relax the local part checking rule
7696 for addresses that are not in any local domains. For these addresses,
7697 slashes and pipe symbols are allowed within local parts, but the sequence
7698 /../ is explicitly forbidden.
7700 31. SPA server authentication was not clearing the challenge buffer before
7703 32. log_message in a "warn" ACL statement was writing to the reject log as
7704 well as to the main log, which contradicts the documentation and doesn't
7705 seem right (because no rejection is happening). So I have stopped it.
7707 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7708 However, I am unable to do any testing of this.
7710 34. Fixed an infelicity in the appendfile transport. When checking directories
7711 for a mailbox, to see if any needed to be created, it was accidentally
7712 using path names with one or more superfluous leading slashes; tracing
7713 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7715 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7716 contents are added to the log line that is written for every discarded
7717 recipient. (Previously a log_message setting was ignored.)
7719 36. The ${quote: operator now quotes the string if it is empty.
7721 37. The install script runs exim in order to find its version number. If for
7722 some reason other than non-existence or emptiness, which it checks, it
7723 could not run './exim', it was installing it with an empty version number,
7724 i.e. as "exim-". This error state is now caught, and the installation is
7727 38. An argument was missing from the function that creates an error message
7728 when Exim fails to connect to the socket for saslauthd authentication.
7729 This could cause Exim to crash, or give a corrupted message.
7731 39. Added isip, isip4, and isip6 to ${if conditions.
7733 40. The ACL variables $acl_xx are now saved with the message, and can be
7734 accessed later in routers, transports, and filters.
7736 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7737 strings in the file are not string-expanded.
7739 42. If a MAIL command specified a SIZE value that was too large to fit into an
7740 int variable, the check against message_size_limit failed. Such values are
7741 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7742 day this will have to be increased, but I don't think I want to be around
7743 when emails are that large.
7750 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7751 iconv() is not standard in FreeBSD.
7753 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7754 IPv6 enabled. The observed symptom was a segmentation fault on return from
7755 the function os_common_find_running_interfaces() in src/os.c.
7757 3. In the check_special_case() function in daemon.c I had used "errno" as an
7758 argument name, which causes warnings on some systems. This was basically a
7759 typo, since it was named "eno" in the comments!
7761 4. The code that waits for the clock to tick (at a resolution of some fraction
7762 of a second) so as to ensure message-id uniqueness was always waiting for
7763 at least one whole tick, when it could have waited for less. [This is
7764 almost certainly not relevant at current processor speeds, where it is
7765 unlikely to ever wait at all. But we try to future-proof.]
7767 5. The function that sleeps for a time interval that includes fractions of a
7768 second contained a race. It did not block SIGALRM between setting the
7769 timer, and suspending (a couple of lines later). If the interval was short
7770 and the sigsuspend() was delayed until after it had expired, the suspension
7771 never ended. On busy systems this could lead to processes getting stuck for
7774 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7775 process, before it forks any delivery processes. The open lookup caching
7776 mechanism meant that the open file or database connection was passed into
7777 the delivery process. The problem was that delivery processes always tidy
7778 up cached lookup data. This could cause a problem for the next delivery
7779 process started by the queue runner, because the external queue runner
7780 process does not know about the closure. So the next delivery process
7781 still has data in the lookup cache. In the case of a file lookup, there was
7782 no problem because closing a file descriptor in a subprocess doesn't affect
7783 the parent. However, if the lookup was caching a connection to a database,
7784 the connection was closed, and the second delivery process was likely to
7785 see errors such as "PGSQL: query failed: server closed the connection
7786 unexpectedly". The problem has been fixed by closing all cached lookups
7787 in a queue runner before running a delivery process.
7789 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7790 seem to have the "const" qualifier which it has on other OS. I've
7793 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7794 *of the same type* (client or server) with the same public name that an
7795 error should be diagnosed.
7797 9. When Exim looked up a host name for an IP address, but failed to find the
7798 original IP address when looking up the host name (a safety check), it
7799 output the message "<ip address> does not match any IP for NULL", which was
7800 confusing, to say the least. The bug was that the host name should have
7801 appeared instead of "NULL".
7803 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7804 user that is built into the binary, and the -C or -D options is used, root
7805 privilege is dropped before the configuration file is read. In addition,
7806 logging is switched to stderr instead of the normal log files. If the
7807 configuration then re-defines the Exim user, the unprivileged environment
7808 is probably not what is expected, so Exim logs a panic warning message (but
7811 However, if deliver_drop_privilege is set, the unprivileged state may well
7812 be exactly what is intended, so the warning has been cut out in that case,
7813 and Exim is allowed to try to write to its normal log files.
7819 1. smtp_return_error_details was not giving details for temporary sender
7820 or receiver verification errors.
7822 2. Diagnose a configuration error if two authenticators have the same public
7825 3. Exim used not to create the message log file for a message until the first
7826 delivery attempt. This could be confusing when incoming messages were held
7827 for policy or load reasons. The message log file is now created at the time
7828 the message is received, and an initial "Received" line is written to it.
7830 4. The automatically generated man page for command line options had a minor
7831 bug that caused no ill effects; however, a more serious problem was that
7832 the procedure for building the man page automatically didn't always
7833 operate. Consequently, release 4.20 contains an out-of-date version. This
7834 shouldn't happen again.
7836 5. When building Exim with embedded Perl support, the script that builds the
7837 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7838 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7840 6. The freeze_tell option was not being used for messages that were frozen on
7841 arrival, either by an ACL or by local_scan().
7843 7. Added the smtp_incomplete_transaction log selector.
7845 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7846 was accepting AUTH without a new EHLO.
7848 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7849 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7852 10. Logging of TCP/IP connections (when configured) now happens in the main
7853 daemon process instead of the child process, so that the TCP/IP connection
7854 count is more accurate (but it can never be perfect).
7856 11. The use of "drop" in a nested ACL was not being handled correctly in the
7857 outer ACL. Now, if condition failure induced by the nested "drop" causes
7858 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7859 or "require"), the connection is dropped.
7861 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7862 that yield "discard" can now be used with an "accept" or a "discard" verb,
7863 but an error is generated for any others (because I can't see a useful way
7864 to define what should happen).
7866 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7867 are now processed in the same way as lines in the Exim configuration file.
7868 In particular, continuation lines are supported.
7870 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7872 15. Added -ti meaning -t -i.
7874 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7875 domains, and warn by logging if others are found.
7877 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7878 returned from GIFCONF when Exim is trying to find the list of interfaces on
7879 a host. The code in os.c has been modified to copy each ifreq to an aligned
7880 structure in all cases.
7882 Also, in some cases, the returned ifreq's were being copied to a 'struct
7883 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7884 means the last couple of bytes of an IPv6 address could be chopped if the
7885 ifreq contained only a normal sockaddr (14 bytes storage).
7887 18. Named domain lists were not supported in the hosts_treat_as_local option.
7888 An entry such as +xxxx was not recognized, and was treated as a literal
7891 19. Ensure that header lines added by a DATA ACL are included in the reject log
7892 if the ACL subsequently rejects the message.
7894 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7895 MD5 (which is deprecated).
7897 21. When testing a filter file using -bf, Exim was writing a message when it
7898 took the sender from a "From " line in the message, but it was not doing so
7899 when it took $return_path from a Return-Path: header line. It now does.
7901 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7902 with a valid header line field name (a series of printing characters
7903 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7905 23. Changed "disc" in the source to "disk" to conform to the documentation and
7906 the book and for uniformity.
7908 24. Ignore Sendmail's -Ooption=value command line item.
7910 25. When execve() failed while trying to run a command in a pipe transport,
7911 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7912 could be confused with a return value of 69 from the command itself. This
7913 has been changed to 127, the value the shell returns if it is asked to run
7914 a non-existent command. The wording for the related log line suggests a
7915 non-existent command as the problem.
7917 26. If received_header_text expands to an empty string, do not add a Received:
7918 header line to the message. (Well, it adds a token one on the spool, but
7919 marks it "old" so that it doesn't get used or transmitted.)
7921 27. Installed eximstats 1.28 (addition of -nt option).
7923 28. There was no check for failure on the call to getsockname() in the daemon
7924 code. This can fail if there is a shortage of resources on the system, with
7925 ENOMEM, for example. A temporary error is now given on failure.
7927 29. Contrary to the C standard, it seems that in some environments, the
7928 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7929 program. Exim now does this explicitly; it affects the formatting of
7930 timestamps using strftime().
7932 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7933 complaints. I've now initialized all the variables, to avoid this.
7935 32. Header lines added by a system filter were not being "seen" during
7936 transport-time rewrites.
7938 33. The info_callback() function passed to OpenSSL is set up with type void
7939 (*)(SSL *, int, int), as described somewhere. However, when calling the
7940 function (actually a macro) that sets it up, the type void(*)() is
7941 expected. I've put in a cast to prevent warnings from picky compilers.
7943 34. If a DNS black list lookup found a CNAME record, but there were no A
7944 records associated with the domain it pointed at, Exim crashed.
7946 35. If a DNS black list lookup returned more than one A record, Exim ignored
7947 all but the first. It now scans all returned addresses if a particular IP
7948 value is being sought. In this situation, the contents of the
7949 $dnslist_value variable are a list of all the addresses, separated by a
7952 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7953 to accept a host name and all its aliases if the forward lookup for any of
7954 them yielded the IP address of the incoming connection. Now it accepts only
7955 those names whose forward lookup yields the correct IP address. Any other
7956 names are discarded. This closes a loophole whereby a rogue DNS
7957 administrator could create reverse DNS records to break through a
7958 wildcarded host restriction in an ACL.
7960 37. If a user filter or a system filter that ran in a subprocess used any of
7961 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7962 the wrong values were passed to the pipe command ($thisaddress had the
7963 value of $0, $0 had the value of $1, etc). This bug was introduced by
7964 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7966 38. Improved the line breaking for long SMTP error messages from ACLs.
7967 Previously, if there was no break point between 40 and 75 characters, Exim
7968 left the rest of the message alone. Two changes have been made: (a) I've
7969 reduced the minimum length to 35 characters; (b) if it can't find a break
7970 point between 35 and 75 characters, it looks ahead and uses the first one
7971 that it finds. This may give the occasional overlong line, but at least the
7972 remaining text gets split now.
7974 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7975 file descriptors might be low, and that setting 1000 would always raise it.
7976 It turns out that in some environments, the limit is already over 1000 and
7977 that lowering it causes trouble. So now Exim takes care not to decrease it.
7979 40. When delivering a message, the value of $return_path is set to $sender_
7980 address at the start of routing (routers may change the value). By an
7981 oversight, this default was not being set up when an address was tested by
7982 -bt or -bv, which affected the outcome if any router or filter referred to
7985 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7986 log only when "message" or "log_message" are set. However, if one of the
7987 conditions was an address verification, or a call to a nested ACL, the
7988 messages generated by the underlying test were being passed through. This
7989 no longer happens. The underlying message is available in $acl_verify_
7990 message for both "message" and "log_message" expansions, so it can be
7991 passed through if needed.
7993 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7994 new expansion $bh_ to give the encoded byte string without charset
7995 translation. Translation happens only if iconv() is available; HAVE_ICONV
7996 indicates this at build time. HEADERS_CHARSET gives the charset to
7997 translate to; headers_charset can change it in the configuration, and
7998 "headers charset" can change it in an individual filter file.
8000 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8001 that creates RFC 2047 encoded "words" labels them as that charset instead
8002 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8003 expansion operator; (ii) when Exim creates a From: line for a local
8004 message; (iii) when a header line is rewritten to include a "phrase" part.
8006 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8007 buggy, causing it to skip the first lines of messages whose message ID
8008 ended in 'D'. This would not have bitten before Exim release 4.14, because
8009 message IDs were unlikely to end in 'D' before then. The effect was to have
8010 incorrect size information for certain domains.
8012 45. #include "config.h" was missing at the start of the crypt16.c module. This
8013 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8016 46. If there was a timeout during a "random" callout check, Exim treated it as
8017 a failure of the random address, and carried on sending RSET and the real
8018 address. If the delay was just some slowness somewhere, the response to the
8019 original RCPT would be taken as a response to RSET and so on, causing
8020 mayhem of various kinds.
8022 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8023 when I implemented it. It didn't allow for the fact that some option values
8024 may legitimately be negative (e.g. size_addition), and it didn't even do
8025 the right test for positive values.
8027 48. Domain names in DNS records are case-independent. Exim always looks them up
8028 in lower case. Some resolvers return domain names in exactly the case they
8029 appear in the zone file, that is, they may contain uppercase letters. Not
8030 all resolvers do this - some return always lower case. Exim was treating a
8031 change of case by a resolver as a change of domain, similar to a widening
8032 of a domain abbreviation. This triggered its re-routing code and so it was
8033 trying to route what was effectively the same domain again. This normally
8034 caused routing to fail (because the router wouldn't handle the domain
8035 twice). Now Exim checks for this case specially, and just changes the
8036 casing of the domain that it ultimately uses when it transmits the message
8039 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8042 50. If a filter generated a file delivery with a non-absolute name (possible if
8043 no home directory exists for the router), the forbid_file option was not
8046 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8047 the existing equality matching.
8049 52. Exim was using ints instead of ino_t variables in some places where it was
8050 dealing with inode numbers.
8052 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8053 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8054 called TMPDIR, and if it finds it is different, it changes its value.
8056 54. The smtp_printf() function is now made available to local_scan() so
8057 additional output lines can be written before returning. There is also an
8058 smtp_fflush() function to enable the detection of a dropped connection.
8059 The variables smtp_input and smtp_batched_input are exported to
8062 55. Changed the default runtime configuration: the message "Unknown user"
8063 has been removed from the ACL, and instead placed on the localuser router,
8064 using the cannot_route_message feature. This means that any verification
8065 failures that generate their own messages won't get overridden. Similarly,
8066 the "Unrouteable address" message that was in the ACL for unverifiable
8067 relay addresses has also been removed.
8069 56. Added hosts_avoid_esmtp to the smtp transport.
8071 57. The exicyclog script was not checking for the esoteric option
8072 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8073 will work only if exicyclog is run under the appropriate euid.
8075 58. Following a discussion on the list, the rules by which Exim recognises line
8076 endings on incoming messages have been changed. The -dropcr and drop_cr
8077 options are now no-ops, retained only for backwards compatibility. The
8078 following line terminators are recognized: LF CRLF CR. However, special
8079 processing applies to CR:
8081 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8082 nor a local message in the state where . is a terminator.
8084 (ii) If a bare CR is encountered in a header line, an extra space is added
8085 after the line terminator so as not to end the header. The reasoning
8086 behind this is that bare CRs in header lines are most likely either
8087 to be mistakes, or people trying to play silly games.
8089 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8090 was being incorrectly given as 18 bytes larger than it should have been.
8091 This is a VOB (very old bug).
8093 60. This may never have affected anything current, but just in case it has:
8094 When the local host is found other than at the start of a list of hosts,
8095 the local host, those with the same MX, and any that follow, are discarded.
8096 When the list in question was part of a longer list of hosts, the following
8097 hosts (not currently being processed) were also being discarded. This no
8098 longer happens. I'm not sure if this situation could ever has previously
8101 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8104 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8105 header. This is something that is recommended in a new Internet Draft, and
8106 is something that is documented as being done by Sendmail. There are two
8107 possible values. For messages generated by the autoreply transport, Exim
8110 Auto-Submitted: auto-replied
8112 whereas for all other generated messages (e.g. bounces) it adds
8114 Auto-Submitted: auto-generated
8116 63. The "personal" condition in filters now includes a test for the
8117 Auto-Submitted: header. If it contains the string "auto-" the message it
8118 not considered personal.
8120 64. Added rcpt_include_affixes as a generic transport option.
8122 65. Added queue_only_override (default true).
8124 66. Added the syslog_duplication option.
8126 67. If what should have been the first header line of a message consisted of
8127 a space followed by a colon, Exim was mis-interpreting it as a header line.
8128 It isn't of course - it is syntactically invalid and should therefore be
8129 treated as the start of the message body. The misbehaviour could have
8130 caused a number of strange effects, including loss of data in subsequent
8131 header lines, and spool format errors.
8133 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8134 client host had authenticated. This control can now be exercised by an ACL
8135 for more flexibility.
8137 69. By default, callouts do not happen when testing with -bh. There is now a
8138 variant, -bhc, which does actually run the callout code, including
8139 consulting and updating the callout cache.
8141 70. Added support for saslauthd authentication, courtesy of Alexander
8144 71. If statvfs() failed on the spool or log directories while checking their
8145 size for availability, Exim confusingly gave the error "space shortage".
8146 Furthermore, in debugging mode it crashed with a floating point exception.
8147 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8148 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8149 serious problem, Exim now writes to the main and panic logs when this
8150 happens, with details of the failure. It then refuses to accept the
8151 incoming message, giving the message "spool directory problem" or "log
8152 directory problem" with a 421 code for SMTP messages.
8154 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8155 0, 1, and 2 exist, because some OS complain for execs without them (see
8156 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8157 descriptors. However, the code omitted to check that the open succeeded,
8158 causing mysterious errors if for some reason the permissions on /dev/null
8159 got screwed. Now Exim writes a message to the main and panic logs, and
8160 bombs out if it can't open /dev/null.
8162 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8163 interact so that it is all more flexible. It is supposed to remain
8164 backwards compatible. Also added extra_local_interfaces.
8166 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8167 to bomb out with an assertion failure - to the client this appears as a
8168 connection drop. This problem occurs in the part of the code that was taken
8169 from the Samba project. Fortunately, the assertion is in a very simple
8170 function, so I have fixed this by reproducing the function inline in the
8171 one place where it is called, and arranging for authentication to fail
8172 instead of killing the process with assert().
8174 75. The SPA client code was not working when the server requested OEM rather
8175 than Unicode encoding.
8177 76. Added code to make require_files with a specific uid setting more usable in
8178 the case where statting the file as root fails - usually a non-root-mounted
8179 NFS file system. When this happens and the failure is EACCES, Exim now
8180 forks a subprocess and does the per-uid checking as the relevant uid.
8182 77. Added process_log_path.
8184 78. If log_file_path was not explicitly set, a setting of check_log_space or
8185 check_log_inodes was ignored.
8187 79. If a space check for the spool or log partitions fails, the incident is now
8188 logged. Of course, in the latter case the data may get lost...
8190 80. Added the %p formatting code to string_format() so that it can be used to
8191 print addresses in debug_print(). Adjusted all the address printing in the
8192 debugging in store.c to use %p rather than %d.
8194 81. There was a concern that a line of code in smtp_in.c could overflow a
8195 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8196 initially expressed, the concern was not well-founded, because trailing
8197 spaces are removed early. However, if the trailing spaces were followed by
8198 a NULL, they did not get removed, so the overflow was possible. Two fixes
8201 (a) I re-wrote the offending code in a cleaner fashion.
8202 (b) If an incoming SMTP command contains a NULL character, it is rejected
8205 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8206 runs initgroups(), so that if the Exim user is in any additional groups,
8207 they will be used during message reception.
8213 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob