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
92 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
93 "taint" log_selector. These were previously deprecated.
95 JH/22 Fix static address-list lookups to properly return the matched item.
96 Previously only the domain part was returned.
98 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
99 the call into OpenSSL to send a TLS Close was being repeated; this
100 resulted in the library waiting for the peer's Close. If that was never
101 sent we waited forever. Fix by tracking send calls.
103 JH/24 The ${run} expansion item now expands its command string elements after
104 splitting. Previously it was before; the new ordering makes handling
105 zero-length arguments simpler. The old ordering can be obtained by
106 appending a new option "preexpand", after a comma, to the "run".
108 JH/25 Taint-check exec arguments for transport-initiated external processes.
109 Previously, tainted values could be used. This affects "pipe", "lmtp" and
110 "queryprogram" transport, transport-filter, and ETRN commands.
111 The ${run} expansion is also affected: in "preexpand" mode no part of
112 the command line may be tainted, in default mode the executable name
115 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
116 the the facility was not passed across execs, and only the first message
117 passed over a connection could use BDAT; any further ones using DATA.
119 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
120 uses $sending_ip_address and an interface is specified.
121 Previously any use of the local address in the EHLO name disabled
122 PIPECONNECT, the common case being to use the rDNS of it.
124 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
125 resumption. Previously verify failed because no certificate status is
126 passed on the wire for the restarted session. Fix by using the recorded
127 ocsp status of the stored session for the new connection.
129 JH/29 TLS resumption: the key for session lookup in the client now includes
130 more info that a server could potentially use in configuring a TLS
131 session, avoiding oferring mismatching sessions to such a server.
132 Previously only the server IP was used.
138 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
139 related applications. Previously an "H" was used where available info
140 says that "M" should be, so change to match.
142 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
143 as arguments, so an implementation trying to copy these into a local
144 buffer was taking a taint-enforcement trap. Fix by using dynamically
145 created buffers. Similar fix for radius expansion condition.
147 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
148 reasonable, eg. to count headers. Fix by using dynamically created
149 buffers rather than a local. Do similar fixes for ACL actions "dcc",
150 "log_reject_target", "malware" and "spam"; the arguments are expanded
151 so could be handling tainted values.
153 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
154 broken the (no-op) support for this sendmail command. Restore it
155 to doing nothing, silently, and returning good status.
157 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
158 record path was given (or the default used) without a leading directory
159 path, an error occurred on trying to open it. Use the transport's working
162 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
163 transport. Previously it was the name on the DNS A-record; use instead
164 the head of the CNAME chain leading there (if there is one). This seems
165 to align better with RFC 6125.
167 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
168 smtp_accept_max_per_host allocated resources which were not released
169 when the limit was exceeded. This eventually crashed the daemon. Fix
170 by adding a release action in that path.
172 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
173 expanded; previously using tainted values was rejected. Fix by using
174 dynamically-created buffers.
176 JH/09 Relax restrictions on ACL verify condition needing access to message
177 headers. Previously they were only permitted in data and non-smtp ACLs;
178 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
179 not_blind, header_sender and header_names_ascii verification.
181 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
182 Previously a macro used one argument twice; when called with the
183 argument as an expression having side-effects, incorrect operation
184 resulted. Use an inlineable function.
186 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
187 held open for a verify callout. Previously this wan not accounted for
188 and a corrupt onward SMTP conversation resulted.
190 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
191 Previously they were embedded naively in the constructed address; when
192 needed, strip the quoting and quote the entire local_part.
193 Also make the inbound_srs expansion condition handle quoting.
195 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
196 excluded, not matching the documentation.
198 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
199 was given for the sqlite_dbfile a trap resulted.
201 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
202 "name" argument resulted in a trap. There is no reason to disallow such;
203 this was a coding error.
205 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
206 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
207 spent suspended, ignoring the POSIX definition. Previously we assumed
208 it did and a constant offset from real time could be used as a correction.
209 Change to using the same clock source for the start-of-message and the
210 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
211 exists, just to get a clock slightly more aligned to reality.
213 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
214 RFC says it is optional some validators care. The missing char was not
215 intended but triggered by a line-wrap alignment. Discovery and fix by
216 Guillaume Outters, hacked on by JH.
218 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
219 name being quoted was tainted a trap would be taken. Fix by using
220 dynamically created buffers. The routine could have been called by a
221 rewrite with the "h" flag, by using the "-F" command-line option, or
222 by using a "name=" option on a control=submission ACL modifier.
224 JH/19 SPF: change the Authentication-Results expansion component to give
225 smtp.helo when the sender domain is empty. Previously it gave
228 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
229 not in 127.0.0.0/8 to help in spotting list domains taken over by a
230 domain-parking registrar.
232 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
233 Previously when a whitespace character was specified it was not inserted
234 after removing the newline.
236 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
237 the domain part of the recipient address. This overrides any tls_sni
238 option set, which was previously used.
240 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
243 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
244 is_tainted() had an off-by-one error in the overenthusiastic direction.
245 Find and fix by Gavan. Although NetBSD is not a supported platform for
246 4.94 this bug could affect other platforms.
248 PP/01 Fix default prime selection to be consistent.
249 One path used ike23 still, instead of exim.dev.20160529.3; now both
250 execution flows will use the same DH primes (currently
251 exim.dev.20160529.3).
253 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
254 option in smtp transport, to match the documentation. Previously
255 verification was not being done in some cases where it should have been.
257 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
258 than one server was defined and depending on the platform memory layout
259 details, an internal consistency trap could be hit while walking the list
262 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
263 files. The value is used by the authresults expansion item. Previously
264 if this was used in a router or transport, a crash could result.
266 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
267 have a sufficient incidence of EINTR returns from select that an
268 interaction with other operations done by the main daemon loop exposed
269 a bug in the error-handling. This was benign apart from the log
272 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
273 for consistency with delivered "=>" and failed "**" lines. While we're
274 there, handle PRX and TFO.
276 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
277 applied. This resulted, if any header-line rewrite rules were configured,
278 in a panic-log triggerable by sending a message with a long address in
279 a header. Fix by increasing the arbitrary limit to larger than a single
280 (dewrapped) 5322 header line maximum size.
282 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
283 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
284 RFC 6648 which deprecates X- options in protocols as a general practice.
285 Changeover between the implementations is handled by the mechanisms
288 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
289 a remote which uses callout/hold. Previously the local was queued.
291 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
294 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
295 log line. The asterisk is now consistently appended to the remote IP
296 (and port, if given), and will also be provided on defer and fail log
297 lines. Previously it could be placed on the local IP if that was being
298 logged, and was only provided on delivery lines.
300 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
302 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
303 authenticator client_send option. Previously the next char, after a pair
304 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
305 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
306 documentation. There is still no way to get a leading ^ immediately
307 after a NUL (ie. for the password of a PLAIN method authenticator.
309 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
310 files. For bad sizes read, delete the record and whine to paniclog.
312 JH/38 When logging an AUTH failure, as server, do not include sensitive
313 information. Previously, the credentials would be included if given
314 as part of the AUTH command line and an ACL denied authentication.
316 JH/39 Bug 2691: fix $local_part_data. When the matching list element
317 referred to a file, bad data was returned. This likely also affected
320 JH/40 The gsasl authenticator now supports caching of the salted password
321 generated by the client-side implementation. This required the addition
322 of a new variable: $auth4.
324 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
325 left undeleted; the attempt to re-create it then failed - resulting in
326 the usual "SIGHUP tp have daemon reload configuration" to not work.
327 This affected any platform not supporting "abstract" Unix-domain
328 sockets (i.e. not Linux).
330 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
331 recipients" response to RCPT in a later response, with a 250. The
332 previous coding assumed this would not happen, and under PIPELINING
333 would result in both lost and duplicate recipients for a message.
335 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
336 Previously the weighting was incorrectly applied. Similar fix for socks
337 proxies. Found and fixed by Heiko Schlichting.
339 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
340 not handle sub-lists included using the +namedlist syntax. While
341 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
342 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
345 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
346 Previously the permanent pool was used, so the sore could not be freed.
347 This meant a connection with many messages would use continually-growing
350 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
351 per-pool so as not to waste too much space. Previously a constant size
352 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
353 DOS attacks harder. The cost is wasted memory use in the larger blocks.
355 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
356 same space repeatedly, and a smaller process footprint.
358 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
359 for GnuTLS. Previously the name started "%s" which, while not a bug,
360 looked as if if might be one.
362 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
363 "continued connection") make the $proxy_* variables available. Previously
364 the information was not passed across the exec() call for subsequent
365 transport executions. This also mean that the log lines for the
366 messages can show the proxy information.
368 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
369 receive time. With modern systems the difference is significant.
370 The historical behaviour can be restored by disabling (a new) log_selector
371 "queue_time_exclusive".
373 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
374 content) were specifically tested for. Now, also cover expansions
375 resulting in ACL names and inline ACL content.
377 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
378 making it unusable in complex expressions.
380 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
381 fail with a taint-check on the spoolfile name, and leave the message
384 HS/01 Enforce absolute PID file path name.
386 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
388 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
390 PP/02 Bug 2643: Correct TLS DH constants.
391 A missing NUL termination in our code-generation tool had led to some
392 incorrect Diffie-Hellman constants in the Exim source.
393 Reported by kylon94, code-gen tool fix by Simon Arlott.
395 PP/03 Impose security length checks on various command-line options.
396 Fixes CVE-2020-SPRSS reported by Qualys.
398 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
399 better. Reported by Qualys.
401 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
402 providing a particularly obnoxious sender full name.
405 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
407 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
410 PP/08 Change default for recipients_max from unlimited to 50,000.
412 PP/09 Fix security issue with too many recipients on a message (to remove a
413 known security problem if someone does set recipients_max to unlimited,
414 or if local additions add to the recipient list).
415 Fixes CVE-2020-RCPTL reported by Qualys.
417 PP/10 Fix security issue in SMTP verb option parsing
418 Fixes CVE-2020-EXOPT reported by Qualys.
420 PP/11 Fix security issue in BDAT state confusion.
421 Ensure we reset known-good where we know we need to not be reading BDAT
422 data, as a general case fix, and move the places where we switch to BDAT
423 mode until after various protocol state checks.
424 Fixes CVE-2020-BDATA reported by Qualys.
426 HS/03 Die on "/../" in msglog file names
428 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
429 the Exim runtime user are allowed to create files.
431 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
434 QS/03 When reading the output from interpreted forward files we do not
435 pass the pipe between the parent and the interpreting process to
436 executed child processes (if any).
438 QS/04 Always die if requested from internal logging, even is logging is
441 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
442 the API; compilation noo longer completes with DMARC support included.
443 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
444 been reported on other platforms.
446 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
448 JH/56 Make the majority of info read from config files readonly, for defence-in-
449 depth against exploits. Suggestion by Qualys.
450 Not supported on Solaris 10.
452 JH/57 Fix control=fakreject for a custom message containing tainted data.
453 Previously this resulted in a log complaint, due to a re-expansion present
454 since fakereject was originally introduced.
456 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
457 modifier was given, a loop resulted.
459 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
460 pipelined SMTP command followed the BDAT LAST then it would be
461 incorrectly treated as part of the message body, causing a verification
464 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
465 require looser validation rules than those for 821-level addresses,
466 which only permit IP addresses.
472 JH/01 Avoid costly startup code when not strictly needed. This reduces time
473 for some exim process initialisations. It does mean that the logging
474 of TLS configuration problems is only done for the daemon startup.
476 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
478 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
479 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
482 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
483 DKIM signing is being done. Previously a transport_filter would always
484 disable CHUNKING, falling back to traditional DATA.
486 JH/05 Regard command-line recipients as tainted.
488 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
490 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
491 PAM library frees one of the arguments given to it, despite the
492 documentation. Therefore a plain malloc must be used.
494 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
495 on-stack buffers were used, resulting in a taint trap when DSN information
496 copied from a received message was written into the buffer.
498 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
499 the ordering of its ARC headers. This caused a crash.
501 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
502 a new record was being constructed with information from the peer, a trap
505 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
506 installation would get error messages from DMARC verify, when it hit the
507 nonexistent file indicated by the default. Distros wanting DMARC enabled
508 should both provide the file and set the option.
509 Also enforce no DMARC verification for command-line sourced messages.
511 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
512 could, depending on the platform, hang at the STARTTLS response.
514 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
515 message on a connection. Previously if one message had ARC headers and
516 the following one did not, a crash could result when adding an
517 Authentication-Results: header.
519 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
520 Exim main code and Exim-related utities. The introduction of taint
521 tracking also did many adjustments to string handling. Since then, eximon
522 frequently terminated with an assert failure.
524 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
525 check for 452 responses. This slightly helps the inefficieny of doing
526 a large alias-expansion into a recipient-limited target. The max_rcpt
527 transport option still applies (and at the current default, will override
528 the new feature). The check is done for either cause of synch, and forces
529 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
530 connection. The new facility is not tunable at this time.
532 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
533 library live data was being used, so the results became garbage. Make
534 copies while it is still usable.
536 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
537 on delivery deferred (==) and failed (**) lines (if a delivery was
538 attemtped). Previously it was only on completion (=>) lines.
540 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
541 for the expansion of the server_scram_iter and server_scram_salt options.
543 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
544 are now specifically given a NO_DATA response without hitting the system
545 resolver. The library goes on to do the now-standard TXT lookup.
546 Use of dnsdb lookups is not affected.
548 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
549 only retrieve the errormessage once. Previously two calls to dlerror()
550 were used, and the second one (for mainlog/paniclog) retrieved null
553 JH/20 Taint checking: disallow use of tainted data for
554 - the appendfile transport file and directory options
555 - the pipe transport command
556 - the autoreply transport file, log and once options
557 - file names used by the redirect router (including filter files)
559 - paths used by single-key lookups
560 Previously this was permitted.
562 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
563 adjusted the size of a major service buffer; this failed because the
564 buffer was in use at the time. Change to a compile-time increase in the
565 buffer size, when this authenticator is compiled into exim.
567 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
568 previous fast-mode was untenable in the face of glibs using mmap to
569 support larger malloc requests.
571 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
572 New values supported, if defined on system where compiled:
573 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
574 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
576 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
577 running a limited number of proceses in parallel, a benefit is gained. The
578 amount varies with the platform hardware and load. The use of the option
579 queue_run_in_order means we cannot do this, as ordering becomes
582 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
583 had introduced a string-copy (for ensuring NUL-termination) which was not
584 appropriate for that case, which can include embedded NUL bytes in the
585 block of data. Investigation showed the copy to actually be needless, the
586 data being length-specified.
588 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
589 done during a receiving connection, and both used TLS, global info was
590 used rather than per-connection info for tracking the state of data
591 queued for transmission. This could result in a connection hang.
593 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
594 Previously, when delivering serveral messages down a single connection
595 only the first would provide a SIZE. This was due to the size information
596 not being properly tracked.
598 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
599 TAI (at 37 seconds currently), pretend to be in UTC for time-related
600 expansion and logging. Previously, spurious values such as a future
601 minute could be seen.
603 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
604 it could crash from a null-deref. This could also affect the
605 ${addresses: } operator and ${readsock } item.
607 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
608 message following a mime one, the variable was not reset.
610 JH/30 When an pipelined-connect fails at the first response, assume incorrect
611 cached capability (perhaps the peer reneged?) and immediately retry in
614 JH/31 Fix spurious detection of timeout while writing to transport filter.
616 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
617 an attempt to copy the string was made before checking it.
619 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
620 taint of the lookup key was maintained; we now regard the presence in the
621 filesystem as sufficient validation.
623 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
624 argument is supplied.
626 JH/35 The dsearch lookup now requires that the directory is an absolute path.
627 Previously this was not checked, and nonempty relative paths made an
628 access under Exim's current working directory.
630 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
631 Previously no event was raised.
633 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
634 parameter supplied by the sender MAIL FROM command. Previously it was
635 ignored, and only the check_spool_space option value for the required
638 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
639 the size of the signing public-key. Previously it was instead giving
640 the size of the signature hash.
642 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
643 the default. See the (new) dkim_verify_min_keysizes option.
645 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
646 and an ACL use a lookup for checking either the local_part or domain,
647 stale data could be accessed. Ensure that variable references are
648 dropped between messages.
650 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
651 by the client was not checked as pointing within response data before
652 being used. A malicious client could thus cause an out-of-bounds read and
653 possibly gain authentication. Fix by adding the check.
655 JH/42 Internationalisation: change the default for downconversion in the smtp
656 transport to be "if needed". Previously it was "as previously set" for
657 the message, which usually meant "if needed" for message-submission but
658 "no" for everything else. However, MTAs have been seen using SMTPUTF8
659 even when the envelope addresses did not need it, resulting in forwarding
660 failures to non-supporting MTAs. A downconvert in such cases will be
661 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
662 transport. The change does mean that addresses needing conversion will
663 be converted when previously a delivery failure would occur.
665 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
666 response was received it would be used unchecked in a fail-DSN, violating
667 standards on line-length limits. Truncate if needed.
669 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
670 doesn't work. (Additionally add a new main config option to configure the
677 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
678 side, to decode a TLS 1.3 packet capture.
680 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
681 Previously the default library behaviour applied, sending two, each in
684 JH/03 Debug output for ACL now gives the config file name and line number for
687 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
689 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
691 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
692 buffer overrun for (non-chunking) other transports.
694 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
695 TLS1.3, means that a server rejecting a client certificate is not visible
696 to the client until the first read of encrypted data (typically the
697 response to EHLO). Add detection for that case and treat it as a failed
698 TLS connection attempt, so that the normal retry-in-clear can work (if
699 suitably configured).
701 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
702 and/or domain. Found and fixed by Jason Betts.
704 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
705 configuration). If a CNAME target was not a wellformed name pattern, a
708 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
709 the OS reports them interleaved with other addresses.
711 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
712 used both for input and for a verify callout, both encrypted, SMTP
713 responses being sent by the server could be lost. This resulted in
714 dropped connections and sometimes bounces generated by a peer sending
717 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
718 string. Previously it was possible to cause undefined behaviour in a
719 library routine (usually a crash). Found by "zerons".
721 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
724 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
725 API was removed, so update to use the newer ones.
727 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
728 any timeout set, is taking a long time. Previously we would hang on to a
729 rotated logfile "forever" if the input was arriving with long gaps
730 (a previous attempt to fix addressed lack, for a long time, of initial
733 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
734 shared (NFS) environment. The length of the tempfile name is now
735 4 + 16 ("hdr.$message_exim_id") which might break on file
736 systems which restrict the file name length to lower values.
739 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
740 shared (NFS) environment.
742 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
743 did for all versions <4.90). Notably -M, -m, --invert, -I may be
746 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
747 on some platforms for bit 31.
749 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
750 to changes apparently associated with TLS1.3 handling some of the APIs
751 previously used were either nonfunctional or inappropriate. Strings
752 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
753 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
754 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
755 This affects log line X= elements, the $tls_{in,out}_cipher variables,
756 and the use of specific cipher names in the encrypted= ACL condition.
758 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
760 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
761 verification result was not updated unless hosts_require_ocsp applied.
763 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
764 queue_list_requires_admin set to false, non-admin users were denied the
767 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
768 directory-of-certs mode. Previously they were advertised despite the
771 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
772 A single TCP connection by a client will now hold a TLS connection open
773 for multiple message deliveries, by default. Previously the default was to
776 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
777 default. If built with the facility, DANE will be used. The facility
778 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
780 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
781 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
782 must be defined and you must still, unless you define DISABLE_TLS, manage
783 the the include-dir and library-file requirements that go with that
784 choice. Non-TLS builds are still supported.
786 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
789 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
790 default. If the platform supports and has the facility enabled, it will
791 be requested on all coneections.
793 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
794 controlled by the build-time option SUPPORT_PIPE_CONNECT.
796 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
798 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
799 success-DSN messages. Previously the From: header was always the default
800 one for these; the option was ignored.
802 JH/28 Fix the timeout on smtp response to apply to the whole response.
803 Previously it was reset for every read, so a teergrubing peer sending
804 single bytes within the time limit could extend the connection for a
805 long time. Credit to Qualsys Security Advisory Team for the discovery.
807 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
808 delivery address, which leaked information of the results of local
809 forwarding. Change to the original envelope recipient address, per
812 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
813 requested. Previously not bounce was generated and a log entry of
814 error ignored was made.
816 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
818 JH/32 Introduce a general tainting mechanism for values read from the input
819 channel, and values derived from them. Refuse to expand any tainted
820 values, to catch one form of exploit.
822 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
823 was unused and the unexpanded text used for the test. Found and
824 fixed by Ruben Jenster.
826 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
827 an attempt to use a TLS library read routine dereffed a nul pointer,
830 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
831 them in our smtp response.
833 JH/36 Have the general router option retry_use_local_part default to true when
834 any of the restrictive preconditions are set (to anything). Previously it
835 was only for check_local user. The change removes one item of manual
836 configuration which is required for proper retries when a remote router
837 handles a subset of addresses for a domain.
839 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
840 link count into consideration.
842 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
843 caused the extension of big_buffer, the following lines were ignored.
845 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
846 accordance with RFC 2308. Previously there was no expiry, so a longlived
847 receive process (eg. due to ACL delays) versus a short SOA value could
850 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
852 JH/39 Promote DMARC support to mainline.
854 JH/40 Bug 2452: Add a References: header to DSNs.
856 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
857 parameters. The relevant library call is documented as "Deprecated: This
858 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
859 3.6.0, DH parameters are negotiated following RFC7919."
861 HS/06 Change the default of dnssec_request_domains to "*"
863 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
864 carried on and emitted a BDAT command, even when PIPELINING was not
867 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
868 buffer was used for the filename, resulting in a trap when tainted
869 arguments (eg. $domain) were used.
871 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
872 recommended to avoid a possible server-load attack. The feature can be
873 re-enabled via the openssl_options main cofiguration option.
875 JH/45 local_scan API: documented the current smtp_printf() call. This changed
876 for version 4.90 - adding a "more data" boolean to the arguments.
877 Bumped the ABI version number also, this having been missed previously;
878 release versions 4.90 to 4.92.3 inclusive were effectively broken in
879 respect of usage of smtp_printf() by either local_scan code or libraries
880 accessed via the ${dlfunc } expansion item. Both will need coding
881 adjustment for any calls to smtp_printf() to match the new function
882 signature; a FALSE value for the new argument is always safe.
884 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
885 the file-offset (which the Linux syscall does, and exim expects); this
886 resulted in an indefinite loop.
888 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
889 to do ARC verification. The Authentication-Results: header line added
890 by the configuration then had no ARC item.
892 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
893 an incorrect mode was used for reading the result, resulting in it being
900 JH/01 Remove code calling the customisable local_scan function, unless a new
901 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
903 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
904 non-signal-safe functions being used.
906 JH/03 Bug 2269: When presented with a received message having a stupidly large
907 number of DKIM-Signature headers, disable DKIM verification to avoid
908 a resource-consumption attack. The limit is set at twenty.
910 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
911 report of oldest_pass in ${authres } in consequence, and separate out
912 some descriptions of reasons for verification fail.
914 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
915 files in the spool were present and unlocked. A queue-runner could spot
916 them, resulting in a duplicate delivery. Fix that by doing the unlock
917 after the unlink. Investigation by Tim Stewart. Take the opportunity to
918 add more error-checking on spoolfile handling while that code is being
921 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
922 No known attacks, no CVE, this is defensive hardening.
924 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
925 a queue-runner could start a delivery while other operations were ongoing.
926 Cutthrough delivery was a common victim, resulting in duplicate delivery.
927 Found and investigated by Tim Stewart. Fix by using the open message data
928 file handle rather than opening another, and not locally closing it (which
929 releases a lock) for that case, while creating the temporary .eml format
930 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
932 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
933 $sender_verify_failure/$recipient_verify_failure to "random".
935 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
938 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
939 Previously this would segfault.
941 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
944 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
945 like zero, since the resolver should be doing this for us, But we need one
946 as a CNAME but no MX presence gets the CNAME returned; we need to check
947 that doesn't point to an MX to declare it "no MX returned" rather than
948 "error, loop". A new main option is added so the older capability of
949 following some limited number of chain links is maintained.
951 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
953 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
954 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
955 now-deprecated earlier definitions used only specified the range up to TLS
956 1.2 (in the older-version library docs).
958 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
960 JH/15 Rework TLS client-side context management. Stop using a global, and
961 explicitly pass a context around. This enables future use of TLS for
962 connections to service-daemons (eg. malware scanning) while a client smtp
963 connection is using TLS; with cutthrough connections this is quite likely.
965 JH/16 Fix ARC verification to do AS checks in reverse order.
967 JH/17 Support a "tls" option on the ${readsocket } expansion item.
969 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
970 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
971 Previously the "utf8" would be re-prepended for every additional message.
973 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
974 Previously thery were accepted, resulting in issues when attempting to
975 forward messages to a non-supporting MTA.
977 PP/02 Let -n work with printing macros too, not just options.
979 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
980 one parent address was copied, and bogus data was used at delivery-logging
981 time. Either a crash (after delivery) or bogus log data could result.
982 Discovery and analysis by Tim Stewart.
984 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
985 Previously if the string ended mid-character, we did not insert the
986 promised '?' replacement.
988 PP/04 Documentation: current string operators work on bytes, not codepoints.
990 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
991 should pack well giving a smaller memory footprint so better caching and
992 therefore performance. Group the declarations where this can't be done so
993 that the byte-sized flag variables are not interspersed among pointer
994 variables, giving a better chance of good packing by the compiler.
996 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
997 non-null, to avoid issues with sites running BATV. Previously reports were
998 sent with an empty envelope sender so looked like bounces.
1000 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1001 The ignore_error flag wasn't being returned from the filter subprocess so
1002 was not set for later routers. Investigation and fix by Matthias Kurz.
1004 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1005 and a msg:complete for the whole, when a message is manually removed using
1006 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1008 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1009 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1010 hope that is portable enough.
1012 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1013 requiring a known-CA anchor certificate; make it now rely entirely on the
1014 TLSA as an anchor. Checking the name on the leaf cert against the name
1015 on the A-record for the host is still done for TA (but not for EE mode).
1017 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1018 included in delivery lines for non-proxied connections, when compiled with
1019 SUPPORT_SOCKS and running with proxy logging enabled.
1021 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1022 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1023 move the existing event to fire before the normal logging of message
1024 failure so that custom logging is bracketed by normal logging.
1026 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1027 msg:fail:internal event. Developement by Matthias Kurz.
1029 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1030 far too small for todays use of crypto signatures stored there. Go all
1031 the way to the max DNS message size of 64kB, even though this might be
1032 overmuch for IOT constrained device use.
1034 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1035 copy a string over itself. The library routine is documented as not
1036 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1038 JH/32 For main options check_spool_space and check_inode_space, where the
1039 platform supports 64b integers, support more than the previous 2^31 kB
1040 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1041 the previous G, M, k.
1043 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1044 $authenticated_fail_id variable on authentication failure. Previously
1047 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1048 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1049 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1050 GNUTLS_SEC_PARAM_MEDIUM.
1052 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1053 side. Previously we would continue as if no SNI had been received.
1055 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1056 "<" character, which should be a list-separator specification, we walked
1057 off past the nul-terimation.
1059 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1060 causes) even when the retry time is not yet met. Previously they were
1061 not, meaning that when (say) an account was over-quota and temp-rejecting,
1062 and multiple senders' messages were queued, only one sender would get
1063 notified on each configured delay_warning cycle.
1065 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1067 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1068 cipher-suites, an error can be left on the stack even for a succeeding
1069 accept; this results in impossible error messages when a later operation
1072 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1073 return error codes indicating retry. Under TLS1.3 this becomes required.
1075 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1076 it only wrote the new authenticators, resulting in a lack of tracking of
1077 peer changes of ESMTP extensions until the next cache flush.
1079 JH/41 Fix the loop reading a message header line to check for integer overflow,
1080 and more-often against header_maxsize. Previously a crafted message could
1081 induce a crash of the recive process; now the message is cleanly rejected.
1083 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1084 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1090 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1091 When redis_servers is set to a list of > 1 element, and the Redis servers
1092 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1093 case of MOVED into a DEFER case instead, thus moving the query onto the
1094 next server in the list. For a cluster of N elements, all N servers must
1095 be defined in redis_servers.
1097 GF/02 Catch and remove uninitialized value warning in exiqsumm
1098 Check for existence of @ARGV before looking at $ARGV[0]
1100 JH/01 Replace the store_release() internal interface with store_newblock(),
1101 which internalises the check required to safely use the old one, plus
1102 the allocate and data copy operations duplicated in both (!) of the
1103 extant use locations.
1105 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1106 modifier. This matches the restriction on the commandline.
1108 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1109 Previously only the last row was returned.
1111 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1112 we assumed that tags in the header were well-formed, and parsed the
1113 element content after inspecting only the first char of the tag.
1114 Assumptions at that stage could crash the receive process on malformed
1117 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1118 While running the DKIM ACL we operate on the Permanent memory pool so that
1119 variables created with "set" persist to the DATA ACL. Also (at any time)
1120 DNS lookups that fail create cache records using the Permanent pool. But
1121 expansions release any allocations made on the current pool - so a dnsdb
1122 lookup expansion done in the DKIM ACL releases the memory used for the
1123 DNS negative-cache, and bad things result. Solution is to switch to the
1124 Main pool for expansions.
1125 While we're in that code, add checks on the DNS cache during store_reset,
1126 active in the testsuite.
1127 Problem spotted, and debugging aided, by Wolfgang Breyha.
1129 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1130 When none of the hosts presented to a transport match an already-open
1131 connection, close it and proceed with the list. Previously we would
1132 queue the message. Spotted by Lena with Yahoo, probably involving
1135 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1136 Previously a spurious "250 OK id=" response was appended to the proper
1139 JH/08 The "support for" informational output now, which built with Content
1140 Scanning support, has a line for the malware scanner interfaces compiled
1141 in. Interface can be individually included or not at build time.
1143 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1144 by the template makefile "src/EDITME". The "STREAM" support for an older
1145 ClamAV interface method is removed.
1147 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1148 rows affected is given instead).
1150 JH/11 The runtime Berkeley DB library version is now additionally output by
1151 "exim -d -bV". Previously only the compile-time version was shown.
1153 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1154 SMTP connection. Previously, when one had more recipients than the
1155 first, an abortive onward connection was made. Move to full support for
1156 multiple onward connections in sequence, handling cutthrough connection
1157 for all multi-message initiating connections.
1159 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1160 routers. Previously, a multi-recipient message would fail to match the
1161 onward-connection opened for the first recipient, and cause its closure.
1163 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1164 a timeout on read on a GnuTLS initiating connection, resulting in the
1165 initiating connection being dropped. This mattered most when the callout
1166 was marked defer_ok. Fix to keep the two timeout-detection methods
1169 JH/15 Relax results from ACL control request to enable cutthrough, in
1170 unsupported situations, from error to silently (except under debug)
1171 ignoring. This covers use with PRDR, frozen messages, queue-only and
1174 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1176 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1177 metadata, resulting in a crash in free().
1179 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1180 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1181 Broken also in d185889f4, with init system revamp.
1183 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1184 Previously we abruptly closed the connection after reading a malware-
1185 found indication; now we go on to read the "scan ok" response line,
1188 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1189 ACL. Previously, a crash would result.
1191 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1192 macro text after a replacement (previously it was only once per line) and
1193 by skipping builtin macros when searching for an uppercase lead character.
1195 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1196 for the build is renamed.
1198 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1199 was allocated for every new TLS startup, meaning one per message. Fix
1200 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1202 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1203 reported the original. Fix to report (as far as possible) the ACL
1204 result replacing the original.
1206 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1207 OpenSSL. Certificate information is loaded for every new TLS startup,
1208 and the resources needed to be freed.
1210 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1212 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1215 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1216 DATA response info to the (existing) per-recipient response info for
1217 the "C=" log element. It can have useful tracking info from the
1218 destination system. Patch from Simon Arlott.
1220 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1221 length value. Previously this would segfault.
1223 HS/02 Support Avast multiline protoocol, this allows passing flags to
1224 newer versions of the scanner.
1226 JH/28 Ensure that variables possibly set during message acceptance are marked
1227 dead before release of memory in the daemon loop. This stops complaints
1228 about them when the debug_store option is enabled. Discovered specifically
1229 for sender_rate_period, but applies to a whole set of variables.
1230 Do the same for the queue-runner and queue-list loops, for variables set
1231 from spool message files. Do the same for the SMTP per-message loop, for
1232 certain variables indirectly set in ACL operations.
1234 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1235 as a multi-recipient message from a mailinglist manager). The coding had
1236 an arbitrary cutoff number of characters while checking for more input;
1237 enforced by writing a NUL into the buffer. This corrupted long / fast
1238 input. The problem was exposed more widely when more pipelineing of SMTP
1239 responses was introduced, and one Exim system was feeding another.
1240 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1241 receiving system, and refused recipients seen by the sending system
1242 (propating to people being dropped from mailing lists).
1243 Discovered and pinpointed by David Carter.
1245 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1246 replaced by the ${authresults } expansion.
1248 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1250 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1251 allows proper process termination in container environments.
1253 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1254 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1256 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1257 and "err_perm", deprecated since 4.83 when the RFC-defined words
1258 "temperror" and "permerror" were introduced.
1260 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1261 transport-filters or DKIM-signing. The restriction was lost in the
1262 consolidation of verify-callout and delivery SMTP handling.
1263 Extend the restriction to also cover ARC-signing.
1265 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1266 in defer=pass mode supply a 450 to the initiator. Previously the message
1269 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1270 tls_require_ciphers is used as before.
1272 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1273 "pass_unscanned". Only tmpfails from the scanner are written to
1274 the paniclog, as they may require admin intervention (permission
1275 denied, license issues). Other scanner errors (like decompression
1276 bombs) do not cause a paniclog entry.
1278 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1279 Previously it was possible to log spurious information in receive log
1282 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1283 triggered odd behaviour from Outlook Express clients.
1285 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1288 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1289 since the IETF WG has not yet settled on that versus the original
1290 "bare" representation.
1292 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1293 Previously the millisecond value corrupted the output.
1294 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1295 corrupted the output.
1301 JH/01 Rework error string handling in TLS interface so that the caller in
1302 more cases is responsible for logging. This permits library-sourced
1303 string to be attached to addresses during delivery, and collapses
1304 pairs of long lines into single ones.
1306 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1307 during configuration. Wildcards are allowed and expanded.
1309 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1310 This permits better logging.
1312 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1313 do not close it down and have the child transport start it up again on
1314 the passed-on TCP connection. Instead, proxy the child (and any
1315 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1316 affected: the continued delivery log lines do not have any DNSSEC, TLS
1317 Certificate or OCSP information. TLS cipher information is still logged.
1319 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1320 identical IP addresses on different listening ports. Will also affect
1323 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1324 add noisy ifdef guards to special-case this sillyness.
1325 Patch from Bernd Kuhls.
1327 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1328 than 255 are no longer allowed.
1330 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1331 Disable the session-cache too, which might reduce our load. Since we
1332 currrectly use a new context for every connection, both as server and
1333 client, there is no benefit for these.
1334 GnuTLS appears to not support tickets server-side by default (we don't
1335 call gnutls_session_ticket_enable_server()) but client side is enabled
1336 by default on recent versions (3.1.3 +) unless the PFS priority string
1339 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1340 <https://reproducible-builds.org/specs/source-date-epoch/>.
1342 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1343 the check for any unsuccessful recipients did not notice the limit, and
1344 erroneously found still-pending ones.
1346 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1347 MSG_MORE. Only in-clear (not on TLS connections).
1349 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1350 transport-filter is involved we can buffer the headers in memory for
1351 creating the signature, and read the spool data file once for the
1352 signature and again for transmission.
1354 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1355 4.77 as the kernel support then wasn't solid, having issues in 64bit
1356 mode. Now, it's been long enough. Add support for FreeBSD also.
1358 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1359 case where the routing stage had gathered several addresses to send to
1360 a host before calling the transport for the first, we previously failed
1361 to close down TLS in the old transport process before passing the TCP
1362 connection to the new process. The new one sent a STARTTLS command
1363 which naturally failed, giving a failed delivery and bloating the retry
1364 database. Investigation and fix prototype from Wolfgang Breyha.
1366 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1367 false-negatives in the check that the sender had not preempted a response
1368 or prompt from Exim (running as a server), due to that code's lack of
1369 awareness of the SMTP input buffering.
1371 PP/04 Add commandline_checks_require_admin option.
1372 Exim drops privileges sanely, various checks such as -be aren't a
1373 security problem, as long as you trust local users with access to their
1374 own account. When invoked by services which pass untrusted data to
1375 Exim, this might be an issue. Set this option in main configuration
1376 AND make fixes to the calling application, such as using `--` to stop
1379 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1380 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1381 responses to those, into a single TLS record each way (this usually means
1382 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1385 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1386 affects you only if you're dancing at the edge of the param size limits.
1387 If you are, and this message makes sense to you, then: raise the
1388 configured limit or use OpenSSL 1.1. Nothing we can do for older
1391 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1392 cmdline element to get the documented default one. Previously it was
1395 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1396 are unsuable for non-message-specific reasons - eg. connection timeout,
1397 banner-time rejection.
1399 JH/16 Fix logging of delivery remote port, when specified by router, under
1402 PP/06 Repair manualroute's ability to take options in any order, even if one
1403 is the name of a transport.
1406 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1408 JH/17 Change the list-building routines interface to use the expanding-string
1409 triplet model, for better allocation and copying behaviour.
1411 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1412 Previously it was constructed the first time a possibly-matching string
1413 was met in the configuration file input during startup; now it is done
1416 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1417 compatible one, to avoid the (poorly documented) possibility of a config
1418 file in the working directory redirecting the DB files, possibly correpting
1419 some existing file. CVE-2017-10140 assigned for BDB.
1421 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1422 cache-hot. Previously, although the result was properly cached, the
1423 initial verify call returned a defer.
1425 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1426 the main verify for receipient in uncached-mode.
1428 JH/22 Retire historical build files to an "unsupported" subdir. These are
1429 defined as "ones for which we have no current evidence of testing".
1431 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1432 if present. Previously it was ignored.
1434 JH/24 Start using specified-initialisers in C structure init coding. This is
1435 a C99 feature (it's 2017, so now considered safe).
1437 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1438 if was a fixed-sized field and bitmask ops via macros; it is now more
1441 PP/07 GitHub PR 56: Apply MariaDB build fix.
1442 Patch provided by Jaroslav Škarvada.
1444 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1445 during Coverity cleanups [4.87 JH/47]
1446 Diagnosis and fix provided by Michael Fischer v. Mollard.
1448 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1449 the right size to place the terminating semicolon on its own folded
1450 line, the header hash was calculated to an incorrect value thanks to
1451 the (relaxed) space the fold became.
1453 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1454 and confused the parent.
1456 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1457 which could crash as a result. This could lead to undeliverable messages.
1459 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1462 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1463 matters on fast-turnover and PID-randomising systems, which were getting
1464 out-of-order delivery.
1466 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1467 a possibly-overlapping copy. The symptom was that "Remote host closed
1468 connection in response to HELO" was logged instead of the actual 4xx
1471 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1472 Previously only that bufferd was discarded, resulting in SYMTP command
1475 JH/32 DKIM: when a message has multiple signatures matching an identity given
1476 in dkim_verify_signers, run the dkim acl once for each. Previously only
1477 one run was done. Bug 2189.
1479 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1480 "panic the current process" to "deliberately defer". The panic log is
1481 still written with the problem list name; the mail and reject logs now
1482 get a temp-reject line for the message that was being handled, saying
1483 something like "domains check lookup or other defer". The SMTP 451
1484 message is still "Temporary local problem".
1486 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1487 A crafted sequence of BDAT commands could result in in-use memory beeing
1488 freed. CVE-2017-16943.
1490 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1491 from SMTP input. Previously it was always done; now only done for DATA
1492 and not BDAT commands. CVE-2017-16944.
1494 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1495 to the message (such as an overlong header line). Previously this was
1496 not done and we did not exit BDAT mode. Followon from the previous item
1497 though a different problem.
1503 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1504 than -2003 did; needs libidn2 in addition to libidn.
1506 JH/02 The path option on a pipe transport is now expanded before use.
1508 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1509 Patch provided by "Björn", documentation fix added too.
1511 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1512 missing a wire-to-host endian conversion.
1514 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1515 close after a BDAT command line could be taken as a following command,
1516 giving a synch failure. Fix by only checking for synch immediately
1517 before acknowledging the chunk.
1519 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1520 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1521 macro. Patches provided by Josh Soref.
1523 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1524 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1525 by IANA as a service extension. However, John Klensin suggests that we
1528 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1529 the dkim code may be unix-mode line endings rather than smtp wire-format
1530 CRLF, so prepend a CR to any bare LF.
1532 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1533 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1535 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1536 the first were themselves being wrongly included in the feed into dkim
1537 processing; with most chunk sizes in use this resulted in an incorrect
1538 body hash calculated value.
1540 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1541 DKIM signature block, for verification. Although advised against by
1542 standards it is specifically not ruled illegal.
1544 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1546 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1547 missing a body hash (the bh= tag).
1549 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1550 It seems that HAProxy sends the Proxy Protocol information in clear and
1551 only then does a TLS startup, so do the same.
1553 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1554 TCP connections (such as for Spamd) unless the daemon successfully set
1555 Fast Open mode on its listening sockets. This fixes breakage seen on
1556 too-old kernels or those not configured for Fast Open, at the cost of
1557 requiring both directions being enabled for TFO, and TFO never being used
1558 by non-daemon-related Exim processes.
1560 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1561 endings, at least on the first header line. Try to canonify any that get
1562 past that check, despite the cost.
1564 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1565 now limited to an arbitrary five deep, while parsing addresses with the
1566 strip_excess_angle_brackets option enabled.
1568 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1569 instead leave the unprompted TLS handshake in socket buffer for the
1570 TLS library to consume.
1572 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1574 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1576 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1577 region in one operation, for speed, and this leaves assigned pointers
1578 dangling. Add checks run only under the testsuite which checks all
1579 variables at a store-reset and panics on a dangling pointer; add code
1580 explicitly nulling out all the variables discovered. Fixes one known
1581 bug: a transport crash, where a dangling pointer for $sending_ip_address
1582 originally assigned in a verify callout, is re-used.
1584 PP/06 Drop '.' from @INC in various Perl scripts.
1586 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1588 PP/08 Reduce a number of compilation warnings under clang; building with
1589 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1590 should be warning-free.
1592 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1594 HS/01 Fix portability problems introduced by PP/08 for platforms where
1595 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1597 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1598 chunk. This allows us to accept broken chunked messages. We need a more
1599 general solution here.
1601 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1602 already-broken messages in the queue.
1604 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1606 JH/19 Fix reference counting bug in routing-generated-address tracking.
1612 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1613 supports it and a size is available (ie. the sending peer gave us one).
1615 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1616 years of being deprecated). The replacements are the ACLs
1617 acl_smtp_mime and acl_not_smtp_mime.
1619 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1620 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1621 or even an in-clear connection were permitted. Now, if the host lookup
1622 was dnssec and dane was requested then the host is only used if the
1623 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1624 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1625 if one fails this test.
1626 This means that a poorly-configured remote DNS will make it incommunicado;
1627 but it protects against a DNS-interception attack on it.
1629 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1630 non-noisy when a race steals the message being considered.
1632 JH/05 If main configuration option tls_certificate is unset, generate a
1633 self-signed certificate for inbound TLS connections.
1635 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1636 in rewrites and routers.
1638 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1639 and logged a warning sing 4.83; now they are a configuration file error.
1641 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1642 (lacking @domain). Apply the same qualification processing as RCPT.
1644 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1646 JH/10 Support ${sha256:} applied to a string (as well as the previous
1649 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1650 a cutthrough deliver is pending, as we always want to make a connection.
1651 This also avoids re-routing the message when later placing the cutthrough
1652 connection after a verify cache hit.
1653 Do not update it with the verify result either.
1655 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1656 when routing results in more than one destination address.
1658 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1659 signing (which inhibits the cutthrough capability). Previously only
1660 the presence of an option was tested; now an expansion evaluating as
1661 empty is permissible (obviously it should depend only on data available
1662 when the cutthrough connection is made).
1664 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1665 the relevant preceding SMTP command did not note the pipelining mode.
1667 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1668 Previously they were not counted.
1670 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1671 as one having no matching records. Previously we deferred the message
1672 that needed the lookup.
1674 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1675 distinguished as "(=".
1677 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1678 for missing MX records. Previously it only worked for missing A records.
1680 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1682 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1683 after the data-go-ahead and data-ack. Patch from Jason Betts.
1685 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1686 even for a "none" policy. Patch from Tony Meyer.
1688 JH/22 Fix continued use of a connection for further deliveries. If a port was
1689 specified by a router, it must also match for the delivery to be
1692 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1693 When one of the recipients of a message was unsuitable for the connection
1694 (has no matching addresses), we lost track of needing to mark it
1695 deferred. As a result mail would be lost.
1697 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1699 JH/25 Decoding ACL controls is now done using a binary search; the source code
1700 takes up less space and should be simpler to maintain. Merge the ACL
1701 condition decode tables also, with similar effect.
1703 JH/26 Fix problem with one_time used on a redirect router which returned the
1704 parent address unchanged. A retry would see the parent address marked as
1705 delivered, so not attempt the (identical) child. As a result mail would
1708 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1709 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1710 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1713 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1714 Main config options check_{log,spool}_{inodes,space} are now
1715 100 inodes, 10MB unless set otherwise in the configuration.
1717 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1718 Previously it only applied to the main-section connection policy
1721 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1723 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1724 by me. Added RFC7919 DH primes as an alternative.
1726 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1727 are not in the system include path.
1729 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1730 GnuTLS, when a session startup failed (eg because the client disconnected)
1731 Exim did stdio operations after fclose. This was exposed by a recent
1732 change which nulled out the file handle after the fclose.
1734 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1735 signed directly by the cert-signing cert, rather than an intermediate
1736 OCSP-signing cert. This is the model used by LetsEncrypt.
1738 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1740 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1741 an incoming connection.
1743 HS/02 Bug 1802: Do not half-close the connection after sending a request
1746 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1747 fallback to "prime256v1".
1749 JH/34 SECURITY: Use proper copy of DATA command in error message.
1750 Could leak key material. Remotely exploitable. CVE-2016-9963.
1756 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1757 and 3.4.4 - once the server is enabled to respond to an OCSP request
1758 it does even when not requested, resulting in a stapling non-aware
1759 client dropping the TLS connection.
1761 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1762 support variable-length bit vectors. No functional change.
1764 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1765 The I= interface field on outgoing lines is now after the H= remote
1766 host field, same as incoming lines. There is a separate
1767 outgoing_interface log selector which allows you to disable the
1770 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1771 If not running log_selector +smtp_connection the mainlog would be held
1772 open indefinitely after a "too many connections" event, including to a
1773 deleted file after a log rotate. Leave the per net connection logging
1774 leaving it open for efficiency as that will be quickly detected by the
1775 check on the next write.
1777 HS/01 Bug 1671: Fix post transport crash.
1778 Processing the wait-<transport> messages could crash the delivery
1779 process if the message IDs didn't exist for some reason. When
1780 using 'split_spool_directory=yes' the construction of the spool
1781 file name failed already, exposing the same netto behaviour.
1783 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1784 mime_regex ACL conditions.
1786 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1787 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1788 to HELO, local diagnostic string.
1790 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1791 log line to debug. Even when configured with a tls authenticator many
1792 client connections are expected to not authenticate in this way, so
1793 an authenticate fail is not an error.
1795 HS/02 Add the Exim version string to the process info. This way exiwhat
1796 gives some more detail about the running daemon.
1798 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1799 matter for fast-change records such as DNSBLs.
1801 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1802 retry record, even if constant. There may be multiple transports with
1803 different interface settings and the retry behaviour needs to be kept
1806 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1808 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1810 JH/10 Bug 840: fix log_defer_output option of pipe transport
1812 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1813 hosts_randomize. This matters a lot when combined with mua_wrapper.
1815 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1816 ${quote_pgsql:<string>} operator.
1818 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1820 JH/14 Change header file name for freeradius-client. Relevant if compiling
1821 with Radius support; from the Gentoo tree and checked under Fedora.
1823 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1825 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1826 option transport_filter, meaning no filtering.
1828 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1830 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1831 defaults to "*" (all hosts). The variable is now available when not built
1832 with TLS, default unset, mainly to enable keeping the testsuite sane.
1833 If a server certificate is not supplied (via tls_certificate) an error is
1834 logged, and clients will find TLS connections fail on startup. Presumably
1835 they will retry in-clear.
1836 Packagers of Exim are strongly encouraged to create a server certificate
1837 at installation time.
1839 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1840 with the $config_file variable.
1842 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1843 in transport context, after the attempt, and per-recipient. The latter type
1844 is per host attempted. The event data is the error message, and the errno
1845 information encodes the lookup type (A vs. MX) used for the (first) host,
1846 and the trailing two digits of the smtp 4xx response.
1848 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1849 to write to mainlog (or rejectlog, paniclog) in the window between file
1850 creation and permissions/ownership being changed. Particularly affects
1851 installations where exicyclog is run as root, rather than exim user;
1852 result is that the running daemon panics and dies.
1854 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1856 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1857 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1858 "pri" and "weight". Note that the previous implicit priority given by the
1859 list order is no longer honoured.
1861 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1862 for DKIM processing.
1864 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1865 by defining SUPPORT_SOCKS.
1867 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1868 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1869 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1870 variables are renamed to proxy_{local,external}_{address,port}.
1872 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1873 for a build by defining SUPPORT_I18N
1875 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1876 of the query string, and make ${quote_redis:} do that quoting.
1878 JH/29 Move Events support from Experimental to mainline, enabled by default
1879 and removable for a build by defining DISABLE_EVENT.
1881 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1883 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1884 cached by the daemon.
1886 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1887 by defining LOOKUP_REDIS. The libhiredis library is required.
1889 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1890 keys are given for lookup.
1892 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1893 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1894 only supported when built with TLS support. The PolarSSL SHA routines
1895 are still used when the TLS library is too old for convenient support.
1897 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1898 openssl_options), for security. OpenSSL forces this from version 1.1.0
1899 server-side so match that on older versions.
1901 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1902 allocation for $value could be released as the expansion processing
1903 concluded, but leaving the global pointer active for it.
1905 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1906 and to use the domains and local_parts ACL conditions.
1908 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1909 incorrectly not doubled on cutthrough transmission, hence seen as a
1910 body-termination at the receiving system - resulting in truncated mails.
1911 Commonly the sender saw a TCP-level error, and retransmitted the message
1912 via the normal store-and-forward channel. This could result in duplicates
1913 received - but deduplicating mailstores were liable to retain only the
1914 initial truncated version.
1916 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1918 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1920 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1921 we're in there, support oversigning also; bug 1309.
1923 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1925 HS/04 Add support for keep_environment and add_environment options.
1927 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1928 either intentional arithmetic overflow during PRNG, or testing config-
1931 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1932 delivery resulted in actual delivery. Cancel cutthrough before DATA
1935 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1936 rejects a recipient: pass the reject to the originator.
1938 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1939 Many were false-positives and ignorable, but it's worth fixing the
1942 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1943 for the new environment-manipulation done at startup. Move the routines
1944 from being local to tls.c to being global via the os.c file.
1946 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1947 an extract embedded as result-arg for a map, the first arg for extract
1948 is unavailable so we cannot tell if this is a numbered or keyed
1949 extraction. Accept either.
1955 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1958 JH/02 The smtp transport option "multi_domain" is now expanded.
1960 JH/03 The smtp transport now requests PRDR by default, if the server offers
1963 JH/04 Certificate name checking on server certificates, when exim is a client,
1964 is now done by default. The transport option tls_verify_cert_hostnames
1965 can be used to disable this per-host. The build option
1966 EXPERIMENTAL_CERTNAMES is withdrawn.
1968 JH/05 The value of the tls_verify_certificates smtp transport and main options
1969 default to the word "system" to access the system default CA bundle.
1970 For GnuTLS, only version 3.0.20 or later.
1972 JH/06 Verification of the server certificate for a TLS connection is now tried
1973 (but not required) by default. The verification status is now logged by
1974 default, for both outbound TLS and client-certificate supplying inbound
1977 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1980 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1981 Status Notification (bounce) messages are now MIME format per RFC 3464.
1982 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1983 under the control of the dsn_advertise_hosts option, and routers may
1984 have a dsn_lasthop option.
1986 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1987 default, modifiable by a malware= option. The list separator for
1988 the options can now be changed in the usual way. Bug 68.
1990 JH/10 The smtp_receive_timeout main option is now expanded before use.
1992 JH/11 The incoming_interface log option now also enables logging of the
1993 local interface on delivery outgoing connections.
1995 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1996 if the interface and destination host and port all match.
1998 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2001 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2002 Patch from Andrew Lewis.
2004 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2005 now supports optional time-restrictions, weighting, and priority
2006 modifiers per server. Patch originally by <rommer@active.by>.
2008 JH/16 The spamd_address main option now supports a mixed list of local
2009 and remote servers. Remote servers can be IPv6 addresses, and
2010 specify a port-range.
2012 JH/17 Bug 68: The spamd_address main option now supports an optional
2013 timeout value per server.
2015 JH/18 Bug 1581: Router and transport options headers_add/remove can
2016 now have the list separator specified.
2018 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2021 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2024 JH/21 Support for the A6 type of dns record is withdrawn.
2026 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2027 rather than the verbs used.
2029 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2030 from 255 to 1024 chars.
2032 JH/24 Verification callouts now attempt to use TLS by default.
2034 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2035 are generic router options now. The defaults didn't change.
2037 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2038 Original patch from Alexander Shikoff, worked over by JH.
2040 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2041 files if ZCAT_COMMAND is not executable.
2043 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2045 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2047 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2048 Normally benign, it bites when the pair was led to by a CNAME;
2049 modern usage is to not canonicalize the domain to a CNAME target
2050 (and we were inconsistent anyway for A-only vs AAAA+A).
2052 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2054 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2055 when evaluating $sender_host_dnssec.
2057 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2058 $sender_helo_dnssec variable.
2060 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2062 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2064 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2066 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2067 documented as working, but never had. Support all but $spam_report.
2069 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2070 added for tls authenticator.
2072 HS/03 Add perl_taintmode main config option
2078 TL/01 When running the test suite, the README says that variables such as
2079 no_msglog_check are global and can be placed anywhere in a specific
2080 test's script, however it was observed that placement needed to be near
2081 the beginning for it to behave that way. Changed the runtest perl
2082 script to read through the entire script once to detect and set these
2083 variables, reset to the beginning of the script, and then run through
2084 the script parsing/test process like normal.
2086 TL/02 The BSD's have an arc4random API. One of the functions to induce
2087 adding randomness was arc4random_stir(), but it has been removed in
2088 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2089 function when detected.
2091 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2092 cause callback expansion.
2094 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2095 syntax errors in an expansion can be treated as a string instead of
2096 logging or causing an error, due to the internal use of bool_lax
2097 instead of bool when processing it.
2099 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2100 server certificates when making smtp deliveries.
2102 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2104 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2106 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2108 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2109 Merged patch from Sebastian Wiedenroth.
2111 JH/05 Fix results-pipe from transport process. Several recipients, combined
2112 with certificate use, exposed issues where response data items split
2113 over buffer boundaries were not parsed properly. This eventually
2114 resulted in duplicates being sent. This issue only became common enough
2115 to notice due to the introduction of connection certificate information,
2116 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2118 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2119 size buffer was used, resulting in syntax errors when an expansion
2122 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2123 version 3.3.6 or later.
2125 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2126 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2127 both become "event_action", the variables become $event_name, $event_data
2128 and $event_defer_errno. There is a new variable $verify_mode, usable in
2129 routers, transports and related events. The tls:cert event is now also
2130 raised for inbound connections, if the main configuration event_action
2133 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2134 early OCSP support, but no stapling (appears to be less than 1.0.0).
2136 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2137 server certificate names available under the smtp transport option
2138 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2141 JH/10 Time-related extraction expansions from certificates now use the main
2142 option "timezone" setting for output formatting, and are consistent
2143 between OpenSSL and GnuTLS compilations. Bug 1541.
2145 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2146 encoded parameter in the incoming message. Bug 1558.
2148 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2149 include certificate info, eximon was claiming there were spoolfile
2152 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2154 JH/14 Log delivery-related information more consistently, using the sequence
2155 "H=<name> [<ip>]" wherever possible.
2157 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2158 are problematic for Debian distribution, omit them from the release
2161 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2163 JH/16 Fix string representation of time values on 64bit time_t architectures.
2166 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2167 output list separator was used.
2172 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2173 checkers that were complaining about end of non-void function with no
2176 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2177 This was a regression introduced in 4.83 by another bugfix.
2179 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2181 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2182 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2188 TF/01 Correctly close the server side of TLS when forking for delivery.
2190 When a message was received over SMTP with TLS, Exim failed to clear up
2191 the incoming connection properly after forking off the child process to
2192 deliver the message. In some situations the subsequent outgoing
2193 delivery connection happened to have the same fd number as the incoming
2194 connection previously had. Exim would try to use TLS and fail, logging
2195 a "Bad file descriptor" error.
2197 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2198 utilities have not been installed.
2200 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2201 temporary space as the ACL may create new global variables.
2203 TL/01 LDAP support uses per connection or global context settings, depending
2204 upon the detected version of the libraries at build time.
2206 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2207 to extract and use the src ip:port in logging and expansions as if it
2208 were a direct connection from the outside internet. PPv2 support was
2209 updated based on HAProxy spec change in May 2014.
2211 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2213 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2214 Properly escape header and check for NULL return.
2216 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2219 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2221 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2222 characters in header names, implemented as a verify condition.
2223 Contributed by Michael Fischer v. Mollard.
2225 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2226 results permerror and temperror. Previous values are deprecated but
2227 still accepted. In a future release, err_perm and err_temp will be
2228 completely removed, which will be a backward incompatibility if the
2229 ACL tests for either of these two old results. Patch contributed by
2230 user bes-internal on the mailing list.
2232 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2234 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2235 selectors, in both main and reject logs.
2237 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2240 JH/07 Add malware type "sock" for talking to simple daemon.
2242 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2244 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2245 routers/transports under cutthrough routing.
2247 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2248 numbers. Touch up "bool" conditional to keep the same definition.
2250 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2252 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2254 JH/12 Expand items in router/transport headers_add or headers_remove lists
2255 individually rather than the list as a whole. Bug 1452.
2257 Required for reasonable handling of multiple headers_ options when
2258 they may be empty; requires that headers_remove items with embedded
2259 colons must have them doubled (or the list-separator changed).
2261 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2262 view the policy declared in the DMARC record. Currently, $dmarc_status
2263 is a combined value of both the record presence and the result of the
2266 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2268 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2269 dnslookup router and the smtp transport (applying to the forward
2272 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2273 of ldap servers used for a specific lookup. Patch provided by Heiko
2276 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2277 New variable $lookup_dnssec_authenticated for observability.
2279 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2280 Patch submitted by Lars Timman.
2282 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2284 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2285 Requires trusted mode and valid format message id, aborts otherwise.
2286 Patch contributed by Heiko Schlichting.
2288 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2289 certextract with support for various fields. Bug 1358.
2291 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2292 is requested by default, modifiable by smtp transport option
2295 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2296 operate on certificate variables to give certificate fingerprints
2297 Also new ${sha256:cert_variable}.
2299 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2301 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2302 Christian Aistleitner.
2304 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2306 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2307 file. Patch from Wolfgang Breyha.
2309 JH/25 Expand the coverage of the delivery $host and $host_address to
2310 client authenticators run in verify callout. Bug 1476.
2312 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2313 align with daemon_smtp_ports. Bug 72.
2315 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2316 support and error reporting did not work properly.
2318 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2319 and is readable. Patch from Andrew Colin Kissa.
2321 TL/14 Enhance documentation of ${run expansion and how it parses the
2322 commandline after expansion, particularly in the case when an
2323 unquoted variable expansion results in an empty value.
2325 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2327 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2328 by renaming away from it. Addresses GH issue 15, reported by
2331 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2333 TL/15 SECURITY: prevent double expansion in math comparison functions
2334 (can expand unsanitized data). Not remotely exploitable.
2341 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2343 PP/02 Make -n do something, by making it not do something.
2344 When combined with -bP, the name of an option is not output.
2346 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2349 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2350 $sender_host_name and config options to manage this, and basic check
2353 PP/05 DSCP support for outbound connections and control modifier for inbound.
2355 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2356 (Only plugin which currently uses this is kerberos4, which nobody should
2357 be using, but we should make it available and other future plugins might
2358 conceivably use it, even though it would break NAT; stuff *should* be
2359 using channel bindings instead).
2361 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2362 name; added for Sendmail compatibility; requires admin caller.
2363 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2364 just ignore it); requires trusted caller.
2365 Also parse but ignore: -Ac -Am -X<logfile>
2368 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2370 TL/02 Add +smtp_confirmation as a default logging option.
2372 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2373 Patch by Magnus Holmgren from 2007-02-20.
2375 TL/04 Bugzilla 1281 - Spec typo.
2376 Bugzilla 1283 - Spec typo.
2377 Bugzilla 1290 - Spec grammar fixes.
2379 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2381 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2383 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2384 mailing list by by Dmitry Isaikin.
2386 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2388 JH/02 Support "G" suffix to numbers in ${if comparisons.
2390 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2392 NM/01 Bugzilla 1197 - Spec typo
2393 Bugzilla 1196 - Spec examples corrections
2395 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2397 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2398 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2401 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2402 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2404 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2405 "acl {{name}{arg}...}", and optional args on acl condition
2408 JH/05 Permit multiple router/transport headers_add/remove lines.
2410 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2412 JH/07 Avoid using a waiting database for a single-message-only transport.
2413 Performance patch from Paul Fisher. Bugzilla 1262.
2415 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2418 JH/09 Add $headers_added variable, with content from use of ACL modifier
2419 add_header (but not yet added to the message). Bugzilla 199.
2421 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2422 Pulled from Bugzilla 817 by Wolfgang Breyha.
2424 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2426 (nb: this is the same fix as in Exim 4.80.1)
2428 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2431 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2434 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2436 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2437 advertise SMTP AUTH mechanism to us, instead of a generic
2438 protocol violation error. Also, make Exim more robust to bad
2439 data from the Dovecot auth socket.
2441 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2443 When a queue runner is handling a message, Exim first routes the
2444 recipient addresses, during which it prunes them based on the retry
2445 hints database. After that it attempts to deliver the message to
2446 any remaining recipients. It then updates the hints database using
2449 So if a recipient address works intermittently, it can get repeatedly
2450 deferred at routing time. The retry hints record remains fresh so the
2451 address never reaches the final cutoff time.
2453 This is a fairly common occurrence when a user is bumping up against
2454 their storage quota. Exim had some logic in its local delivery code
2455 to deal with this. However it did not apply to per-recipient defers
2456 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2458 This change adds a proper retry rule check during routing so that the
2459 final cutoff time is checked against the message's age. We only do
2460 this check if there is an address retry record and there is not a
2461 domain retry record; this implies that previous attempts to handle
2462 the address had the retry_use_local_parts option turned on. We use
2463 this as an approximation for the destination being like a local
2464 delivery, as in LMTP.
2466 I suspect this new check makes the old local delivery cutoff check
2467 redundant, but I have not verified this so I left the code in place.
2469 TF/02 Correct gecos expansion when From: is a prefix of the username.
2471 Test 0254 submits a message to Exim with the header
2475 When I ran the test suite under the user fanf2, Exim expanded
2476 the header to contain my full name, whereas it should have added
2477 a Resent-Sender: header. It erroneously treats any prefix of the
2478 username as equal to the username.
2480 This change corrects that bug.
2482 GF/01 DCC debug and logging tidyup
2483 Error conditions log to paniclog rather than rejectlog.
2484 Debug lines prefixed by "DCC: " to remove any ambiguity.
2486 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2488 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2489 Bug spotted by Jeremy Harris; was flawed since initial commit.
2490 Would have resulted in OCSP responses post-SNI triggering an Exim
2491 NULL dereference and crash.
2493 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2495 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2496 Bug detection, analysis and fix by Samuel Thibault.
2497 Bugzilla 1331, Debian bug #698092.
2499 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2501 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2502 Server implementation by Todd Lyons, client by JH.
2503 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2504 config variable "prdr_enable" controls whether the server
2505 advertises the facility. If the client requests PRDR a new
2506 acl_data_smtp_prdr ACL is called once for each recipient, after
2507 the body content is received and before the acl_smtp_data ACL.
2508 The client is controlled by both of: a hosts_try_prdr option
2509 on the smtp transport, and the server advertisement.
2510 Default client logging of deliveries and rejections involving
2511 PRDR are flagged with the string "PRDR".
2513 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2514 fclose(). Diagnosis by Todd Lyons.
2516 PP/17 Update configure.default to handle IPv6 localhost better.
2517 Patch by Alain Williams (plus minor tweaks).
2520 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2521 This is now consistent with GnuTLS, and is now documented: the
2522 previous undocumented portable approach to treating the option as
2523 unset was to force an expansion failure. That still works, and
2524 an empty string is now equivalent.
2526 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2527 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2528 not performing validation itself.
2530 PP/20 Added force_command boolean option to pipe transport.
2531 Patch from Nick Koston, of cPanel Inc.
2533 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2536 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2538 PP/21 Fix eximon continuous updating with timestamped log-files.
2539 Broken in a format-string cleanup in 4.80, missed when I repaired the
2540 other false fix of the same issue.
2541 Report and fix from Heiko Schlichting.
2544 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2545 Report from Prashanth Katuri.
2547 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2548 It's SecureTransport, so affects any MacOS clients which use the
2549 system-integrated TLS libraries, including email clients.
2551 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2552 using a MIME ACL for non-SMTP local injection.
2553 Report and assistance in diagnosis by Warren Baker.
2555 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2557 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2559 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2560 last id that failed so it may be referenced in subsequent ACL's.
2562 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2565 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2566 ldap library initialization, allowing self-signed CA's to be
2567 used. Also properly sets require_cert option later in code by
2568 using NULL (global ldap config) instead of ldap handle (per
2569 session). Bug diagnosis and testing by alxgomz.
2571 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2572 the src/util/ subdirectory.
2574 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2575 renamed to Transport Post Delivery Action by Jeremy Harris, as
2578 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2579 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2580 redis_servers = needs to be configured which will be used by the redis
2581 lookup. Patch from Warren Baker, of The Packet Hub.
2583 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2585 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2586 hostname or reverse DNS when processing a host list. Used suggestions
2587 from multiple comments on this bug.
2589 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2591 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2592 Missed a few lines, added it to make the runtest require no keyboard
2595 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2596 contains upper case chars. Make router use caseful_local_part.
2598 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2599 support when GnuTLS has been built with p11-kit.
2605 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2607 This, or similar/improved, will also be change PP/11 of 4.82.
2613 PP/01 Handle short writes when writing local log-files.
2614 In practice, only affects FreeBSD (8 onwards).
2615 Bugzilla 1053, with thanks to Dmitry Isaikin.
2617 NM/01 Bugzilla 949 - Documentation tweak
2619 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2622 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2624 PP/02 Implemented gsasl authenticator.
2626 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2628 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2629 `pkg-config foo` for cflags/libs.
2631 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2632 with rest of GSASL and with heimdal_gssapi.
2634 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2635 `pkg-config foo` for cflags/libs for the TLS implementation.
2637 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2638 properties get this fed in as external SSF. A number of robustness
2639 and debugging improvements to the cyrus_sasl authenticator.
2641 PP/08 cyrus_sasl server now expands the server_realm option.
2643 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2644 Patch by Jeremy Harris.
2646 PP/10 Added dbmjz lookup type.
2648 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2650 PP/12 MAIL args handles TAB as well as SP, for better interop with
2651 non-compliant senders.
2652 Analysis and variant patch by Todd Lyons.
2654 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2655 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2656 Patch from Dirk Mueller <dmueller@suse.com>
2658 PP/13 tls_peerdn now print-escaped for spool files.
2659 Observed some $tls_peerdn in wild which contained \n, which resulted
2660 in spool file corruption.
2662 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2663 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2664 or write after TLS renegotiation, which otherwise led to messages
2667 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2668 as a tracking header (ie: a signed header comes before the signature).
2669 Patch from Wolfgang Breyha.
2671 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2672 comma-sep list; embedded commas doubled.
2674 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2676 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2678 Report and patch from Dmitry Banschikov.
2680 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2681 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2682 failures. We appear to now support TLS1.1+ with Exim.
2684 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2685 lets Exim select keys and certificates based upon TLS SNI from client.
2686 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2687 before an outbound SMTP session. New log_selector, +tls_sni.
2689 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2690 NULL dereference. Report and patch from Alun Jones.
2692 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2693 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2694 Not seeing resolver debug output on NetBSD, but suspect this is a
2695 resolver implementation change.
2697 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2698 Left warnings. Added "eximon gdb" invocation mode.
2700 PP/21 Defaulting "accept_8bitmime" to true, not false.
2702 PP/22 Added -bw for inetd wait mode support.
2704 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2705 locate the relevant includes and libraries. Made this the default.
2707 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2708 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2710 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2711 This may cause build issues on older platforms.
2713 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2714 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2715 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2716 Added SNI support via GnuTLS too.
2717 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2719 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2721 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2722 Applied second patch from Janne, implementing suggestion to default
2723 multiple-strings-in-record handling to match SPF spec.
2725 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2727 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2728 read-only, out of scope).
2729 Patch from Wolfgang Breyha, report from Stuart Northfield.
2731 PP/29 Fix three issues highlighted by clang analyser static analysis.
2732 Only crash-plausible issue would require the Cambridge-specific
2733 iplookup router and a misconfiguration.
2734 Report from Marcin Mirosław.
2736 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2738 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2739 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2740 As part of this, removing so much warning spew let me fix some minor
2741 real issues in debug logging.
2743 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2744 assignment on my part. Fixed.
2746 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2747 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2748 Janne Snabb (who went above and beyond: thank you).
2750 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2751 string otherwise requires a connection and a bunch more work and it's
2752 relatively easy to get wrong. Should also expose TLS library linkage
2755 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2756 64-bit ${eval} (JH/03).
2758 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2759 GNU libc to support some of the 64-bit stuff, should not lead to
2760 conflicts. Defined before os.h is pulled in, so if a given platform
2761 needs to override this, it can.
2763 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2764 protection layer was required, which is not implemented.
2765 Bugzilla 1254, patch from Wolfgang Breyha.
2767 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2768 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2769 tls_dhparam take prime identifiers. Also unbreak combination of
2770 OpenSSL+DH_params+TLSSNI.
2772 PP/39 Disable SSLv2 by default in OpenSSL support.
2778 PP/01 Solaris build fix for Oracle's LDAP libraries.
2779 Bugzilla 1109, patch from Stephen Usher.
2781 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2783 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2786 TF/02 Fix a couple more cases where we did not log the error message
2787 when unlink() failed. See also change 4.74-TF/03.
2789 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2790 lock up or crash if it happened to be inside a call to libc when it
2791 got a SIGUSR1 from exiwhat.
2793 The SIGUSR1 handler appends the current process status to the process
2794 log which is later printed by exiwhat. It used to use the general
2795 purpose logging code to do this, but several functions it calls are
2796 not safe for signals.
2798 The new output code in the SIGUSR1 handler is specific to the process
2799 log, and simple enough that it's easy to inspect for signal safety.
2800 Removing some special cases also simplifies the general logging code.
2801 Removing the spurious timestamps from the process log simplifies
2804 TF/04 Improved ratelimit ACL condition.
2806 The /noupdate option has been deprecated in favour of /readonly which
2807 has clearer semantics. The /leaky, /strict, and /readonly update modes
2808 are mutually exclusive. The update mode is no longer included in the
2809 database key; it just determines when the database is updated. (This
2810 means that when you upgrade Exim will forget old rate measurements.)
2812 Exim now checks that the per_* options are used with an update mode that
2813 makes sense for the current ACL. For example, when Exim is processing a
2814 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2815 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2816 must specify per_mail/readonly. If you omit the update mode it defaults to
2817 /leaky where that makes sense (as before) or /readonly where required.
2819 The /noupdate option is now undocumented but still supported for
2820 backwards compatibility. It is equivalent to /readonly except that in
2821 ACLs where /readonly is required you may specify /leaky/noupdate or
2822 /strict/noupdate which are treated the same as /readonly.
2824 A useful new feature is the /count= option. This is a generalization
2825 of the per_byte option, so that you can measure the throughput of other
2826 aggregate values. For example, the per_byte option is now equivalent
2827 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2829 The per_rcpt option has been generalized using the /count= mechanism
2830 (though it's more complicated than the per_byte equivalence). When it is
2831 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2832 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2833 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2834 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2835 Note that using per_rcpt with a non-readonly update mode in more than
2836 one ACL will cause the recipients to be double-counted. (The per_mail
2837 and per_byte options don't have this problem.)
2839 The handling of very low rates has changed slightly. If the computed rate
2840 is less than the event's count (usually one) then this event is the first
2841 after a long gap. In this case the rate is set to the same as this event's
2842 count, so that the first message of a spam run is counted properly.
2844 The major new feature is a mechanism for counting the rate of unique
2845 events. The new per_addr option counts the number of different
2846 recipients that someone has sent messages to in the last time period. It
2847 behaves like per_rcpt if all the recipient addresses are different, but
2848 duplicate recipient addresses do not increase the measured rate. Like
2849 the /count= option this is a general mechanism, so the per_addr option
2850 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2851 example, measure the rate that a client uses different sender addresses
2852 with the options per_mail/unique=$sender_address. There are further
2853 details in the main documentation.
2855 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2857 TF/06 Removed a few PCRE remnants.
2859 TF/07 Automatically extract Exim's version number from tags in the git
2860 repository when doing development or release builds.
2862 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2863 Bugzilla 879. Patch from Paul Fisher.
2865 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2866 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2869 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2871 PP/05 Variable $av_failed, true if the AV scanner deferred.
2872 Bugzilla 1078. Patch from John Horne.
2874 PP/06 Stop make process more reliably on build failure.
2875 Bugzilla 1087. Patch from Heiko Schlittermann.
2877 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2878 Bugzilla 1089. Patch from Heiko Schlittermann.
2880 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2881 Bugzilla 1131. Patch from Holger Weiß.
2883 PP/09 Handle IPv6 addresses with SPF.
2884 Bugzilla 860. Patch from Wolfgang Breyha.
2886 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2888 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2891 PP/11 match_* no longer expand right-hand-side by default.
2892 New compile-time build option, EXPAND_LISTMATCH_RHS.
2893 New expansion conditions, "inlist", "inlisti".
2895 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2897 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2899 PP/14 fix log_write() format string regression from TF/03.
2900 Bugzilla 1152. Patch from Dmitry Isaikin.
2906 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2908 PP/02 Harmonised TLS library version reporting; only show if debugging.
2909 Layout now matches that introduced for other libraries in 4.74 PP/03.
2911 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2913 PP/04 New "dns_use_edns0" global option.
2915 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2918 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2919 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2921 TK/01 Updated PolarSSL code to 0.14.2.
2922 Bugzilla 1097. Patch from Andreas Metzler.
2924 PP/07 Catch divide-by-zero in ${eval:...}.
2925 Fixes bugzilla 1102.
2927 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2930 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2931 format-string attack -- SECURITY: remote arbitrary code execution.
2933 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2934 time unintentionally subject to list matching rules, letting the header
2935 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2936 arbitrary string expansion). This allowed for information disclosure.
2938 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2939 INT_MIN/-1 -- value coerced to INT_MAX.
2945 NM/01 Workaround for PCRE version dependency in version reporting
2948 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2949 This fixes portability to compilers other than gcc, notably
2950 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2952 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2953 makefiles for portability to HP-UX and POSIX correctness.
2955 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2956 Also via indented variable definition in the Makefile.
2957 (Debugging by Oliver Heesakkers).
2959 PP/02 Restore caching of spamd results with expanded spamd_address.
2960 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2962 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2963 Improves build reliability. Fix from: Frank Elsner
2965 NM/02 Fix wide character breakage in the rfc2047 coding
2966 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2968 NM/03 Allow underscore in dnslist lookups
2969 Fixes bug 1026. Patch from Graeme Fowler
2971 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2972 Code patches from Adam Ciarcinski of NetBSD.
2974 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2977 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2978 is logged, to avoid truncation. Patch from John Horne.
2980 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2981 Patch from Jakob Hirsch.
2983 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2984 SQL string expansion failure details.
2985 Patch from Andrey Oktyabrski.
2987 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2988 Patch from Simon Arlott.
2990 PP/09 New lookups functionality failed to compile on old gcc which rejects
2991 extern declarations in function scope.
2992 Patch from Oliver Fleischmann
2994 PP/10 Use sig_atomic_t for flags set from signal handlers.
2995 Check getgroups() return and improve debugging.
2996 Fixed developed for diagnosis in bug 927 (which turned out to be
2999 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3000 Patch from Mark Zealey.
3002 PP/12 Bugzilla 1056: Improved spamd server selection.
3003 Patch from Mark Zealey.
3005 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3006 Based on patch from Heiko Schlittermann.
3008 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3009 Patch from Uwe Doering, sign-off by Michael Haardt.
3011 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3012 variable declaration deep within a block. Bug and patch from
3015 PP/15 lookups-Makefile IRIX compatibility coercion.
3017 PP/16 Make DISABLE_DKIM build knob functional.
3019 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3020 Patch by Simon Arlott
3022 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3023 variable argument macros. Our copy now differs from upstream.
3029 TF/01 Failure to get a lock on a hints database can have serious
3030 consequences so log it to the panic log.
3032 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3033 controlled using the smtp_confirmation log selector.
3035 TF/03 Include the error message when we fail to unlink a spool file.
3037 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3038 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3039 for maintaining out-of-tree patches for some time.
3041 PP/01 Bugzilla 139: Documentation and portability issues.
3042 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3043 Handle per-OS dynamic-module compilation flags.
3045 PP/02 Let /dev/null have normal permissions.
3046 The 4.73 fixes were a little too stringent and complained about the
3047 permissions on /dev/null. Exempt it from some checks.
3048 Reported by Andreas M. Kirchwitz.
3050 PP/03 Report version information for many libraries, including
3051 Exim version information for dynamically loaded libraries. Created
3052 version.h, now support a version extension string for distributors
3053 who patch heavily. Dynamic module ABI change.
3055 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3056 privilege escalation vulnerability whereby the Exim run-time user
3057 can cause root to append content of the attacker's choosing to
3060 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3063 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3064 If dropping privileges for untrusted macros, we disabled normal logging
3065 on the basis that it would fail; for the Exim run-time user, this is not
3066 the case, and it resulted in successful deliveries going unlogged.
3067 Fixed. Reported by Andreas Metzler.
3073 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3074 only prepend for the Resent-* case. Fixes regression introduced in
3075 Exim 4.70 by NM/22 for Bugzilla 607.
3077 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3078 increasing numbers of administrators be bitten by this.
3080 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3082 PP/03 Bugzilla 994: added openssl_options main configuration option.
3084 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3086 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3088 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3090 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3091 without a peer certificate, leading to a segfault because of an
3092 assumption that peers always have certificates. Be a little more
3093 paranoid. Problem reported by Martin Tscholak.
3095 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3096 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3097 NB: ClamAV planning to remove STREAM in "middle of 2010".
3098 CL also introduces -bmalware, various -d+acl logging additions and
3099 more caution in buffer sizes.
3101 PP/09 Implemented reverse_ip expansion operator.
3103 PP/10 Bugzilla 937: provide a "debug" ACL control.
3105 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3107 PP/12 Bugzilla 973: Implement --version.
3109 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3111 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3113 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3115 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3116 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3117 ignore trailing whitespace.
3119 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3121 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3124 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3125 Notification from Dr Andrew Aitchison.
3127 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3128 ExtendedDetectionInfo response format.
3129 Notification from John Horne.
3131 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3134 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3135 XSL and documented dependency on system catalogs, with examples of how
3138 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3141 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3142 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3143 configuration file which is writeable by the Exim user or group.
3145 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3146 of configuration files to cover files specified with the -C option if
3147 they are going to be used with root privileges, not just the default
3150 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3151 option (effectively making it always true).
3153 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3154 files to be used while preserving root privileges.
3156 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3157 that rogue child processes cannot use them.
3159 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3160 run-time user, instead of root.
3162 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3163 Exim run-time user without dropping privileges.
3165 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3166 result string, instead of calling string_vformat() twice with the same
3169 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3170 for other users. Others should always drop root privileges if they use
3171 -C on the command line, even for a whitelisted configure file.
3173 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3175 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3181 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3182 $data_path, and $header_path variables; fixed documentation bugs and
3185 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3186 exipick to access non-standard spools, including the "frozen" queue
3189 NM/01 Bugzilla 965: Support mysql stored procedures.
3190 Patch from Alain Williams
3192 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3194 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3195 Patch from Andreas Metzler
3197 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3198 Patch from Kirill Miazine
3200 NM/05 Bugzilla 671: Added umask to procmail example.
3202 JJ/03 installed exipick 20100323.0, fixing doc bug
3204 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3205 directory. Notification and patch from Dan Rosenberg.
3207 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3209 TK/02 Improve log output when DKIM signing operation fails.
3211 MH/01 Treat the transport option dkim_domain as a colon separated
3212 list, not as a single string, and sign the message with each element,
3213 omitting multiple occurences of the same signer.
3215 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3216 Bugzilla 985, 986. Patch by Simon Arlott
3218 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3219 Patch by Simon Arlott
3221 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3222 MBX locking. Notification from Dan Rosenberg.
3228 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3230 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3232 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3234 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3236 NM/04 Bugzilla 915: Flag broken perl installation during build.
3242 TK/01 Added patch by Johannes Berg that expands the main option
3243 "spamd_address" if it starts with a dollar sign.
3245 TK/02 Write list of recipients to X-Envelope-Sender header when building
3246 the mbox-format spool file for content scanning (suggested by Jakob
3249 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3250 (http://www.dcc-servers.net/) support via dccifd. Activated by
3251 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3253 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3254 by Mark Daniel Reidel <mr@df.eu>.
3256 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3257 When building exim an external PCRE library is now needed -
3258 PCRE is a system library on the majority of modern systems.
3259 See entry on PCRE_LIBS in EDITME file.
3261 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3262 conversation. Added nologin parameter to request.
3263 Patch contributed by Kirill Miazine.
3265 TF/01 Do not log submission mode rewrites if they do not change the address.
3267 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3269 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3270 log files in place. Contributed by Roberto Lima.
3272 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3274 TF/03 Bugzilla 615: When checking the local_parts router precondition
3275 after a local_part_suffix or local_part_prefix option, Exim now
3276 does not use the address's named list lookup cache, since this
3277 contains cached lookups for the whole local part.
3279 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3280 Robert Millan. Documentation is in experimental-spec.txt.
3282 TF/04 Bugzilla 668: Fix parallel build (make -j).
3284 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3286 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3287 Patch provided by Jan Srzednicki.
3289 TF/05 Leading white space used to be stripped from $spam_report which
3290 wrecked the formatting. Now it is preserved.
3292 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3293 that they are available at delivery time.
3295 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3297 TF/08 TLS error reporting now respects the incoming_interface and
3298 incoming_port log selectors.
3300 TF/09 Produce a more useful error message if an SMTP transport's hosts
3301 setting expands to an empty string.
3303 NM/06 Bugzilla 744: EXPN did not work under TLS.
3304 Patch provided by Phil Pennock.
3306 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3307 Patch provided by Richard Godbee.
3309 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3310 acl_smtp_notquit, added index entry.
3312 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3313 Patch provided by Eugene Bujak.
3315 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3316 accept(). Patch provided by Maxim Dounin.
3318 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3319 Patch provided by Phil Pennock.
3321 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3323 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3324 Patch provided by Brad "anomie" Jorsch.
3326 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3327 Patch provided by Dean Brooks.
3329 TK/05 Add native DKIM support (does not depend on external libraries).
3331 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3332 Patch provided by Graeme Fowler.
3334 NM/16 Bugzilla 851: Documentation example syntax fix.
3336 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3338 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3341 NM/19 Bugzilla 745: TLS version reporting.
3342 Patch provided by Phil Pennock.
3344 NM/20 Bugzilla 167: bool: condition support.
3345 Patch provided by Phil Pennock.
3347 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3348 clients. Patch provided by Phil Pennock.
3350 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3351 Patch provided by Brad "anomie" Jorsch.
3353 NM/23 Bugzilla 687: Fix misparses in eximstats.
3354 Patch provided by Heiko Schlittermann.
3356 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3357 Patch provided by Heiko Schlittermann.
3359 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3360 plus update to original patch.
3362 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3364 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3365 Patch provided by David Brownlee.
3367 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3369 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3371 NM/30 Bugzilla 888: TLS documentation bugfixes.
3373 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3375 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3376 Unlike the original bugzilla I have changed all shell scripts in src tree.
3378 NM/33 Bugzilla 898: Transport filter timeout fix.
3379 Patch by Todd Rinaldo.
3381 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3382 Patch by Serge Demonchaux.
3384 NM/35 Bugzilla 39: Base64 decode bug fixes.
3385 Patch by Jakob Hirsch.
3387 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3389 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3391 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3393 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3399 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3400 ALT-N's libdkim that I have put here:
3401 http://duncanthrax.net/exim-experimental/
3403 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3404 were called 'true' and it seems that C99 defines that as a reserved
3405 keyword to be used with 'bool' variable types. That means you could
3406 not include C99-style headers which use bools without triggering
3407 build errors in sieve.c.
3409 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3410 as mailq or other aliases. Changed the --help handling significantly
3411 to do whats expected. exim_usage() emits usage/help information.
3413 SC/01 Added the -bylocaldomain option to eximstats.
3415 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3417 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3419 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3425 PH/01 Another patch from the Sieve maintainer.
3427 PH/02 When an IPv6 address is converted to a string for single-key lookup
3428 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3429 dots are used instead of colons so that keys in lsearch files need not
3430 contain colons. This was done some time before quoting was made available
3431 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3432 (notated using the quote facility) so as to distinguish them from IPv4
3433 keys. This meant that lookups for IP addresses in host lists did not work
3434 for iplsearch lookups.
3436 This has been fixed by arranging for IPv6 addresses to be expressed with
3437 colons if the lookup type is iplsearch. This is not incompatible, because
3438 previously such lookups could never work.
3440 The situation is now rather anomalous, since one *can* have colons in
3441 ordinary lsearch keys. However, making the change in all cases is
3442 incompatible and would probably break a number of configurations.
3444 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3447 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3448 conversion specification without a maximum field width, thereby enabling
3449 a rogue spamd server to cause a buffer overflow. While nobody in their
3450 right mind would setup Exim to query an untrusted spamd server, an
3451 attacker that gains access to a server running spamd could potentially
3452 exploit this vulnerability to run arbitrary code as the Exim user.
3454 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3455 $primary_hostname instead of what libspf2 thinks the hosts name is.
3457 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3458 a directory entry by the name of the lookup key. Previously, if a
3459 symlink pointed to a non-existing file or a file in a directory that
3460 Exim lacked permissions to read, a lookup for a key matching that
3461 symlink would fail. Now it is enough that a matching directory entry
3462 exists, symlink or not. (Bugzilla 503.)
3464 PH/03 The body_linecount and body_zerocount variables are now exported in the
3467 PH/04 Added the $dnslist_matched variable.
3469 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3470 This means they are set thereafter only if the connection becomes
3473 PH/06 Added the client_condition to authenticators so that some can be skipped
3474 by clients under certain conditions.
3476 PH/07 The error message for a badly-placed control=no_multiline_responses left
3477 "_responses" off the end of the name.
3479 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3481 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3482 (without spaces) instead of just copying the configuration text.
3484 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3486 PH/11 Added $max_received_linelength.
3488 PH/12 Added +ignore_defer and +include_defer to host lists.
3490 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3491 way in which PCRE > 7.0 is built.
3493 PH/14 Implemented queue_only_load_latch.
3495 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3496 MAIL command. The effect was to mangle the value on 64-bit systems.
3498 PH/16 Another patch from the Sieve maintainer.
3500 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3502 PH/18 If a system quota error occurred while trying to create the file for
3503 a maildir delivery, the message "Mailbox is full" was not appended to the
3504 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3505 only to a quota excession during the actual writing of the file.
3507 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3508 characters?) which causes problems in log lines. The DN values are now
3509 passed through string_printing() before being added to log lines.
3511 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3512 and InterBase are left for another time.)
3514 PH/21 Added message_body_newlines option.
3516 PH/22 Guard against possible overflow in moan_check_errorcopy().
3518 PH/23 POSIX allows open() to be a macro; guard against that.
3520 PH/24 If the recipient of an error message contained an @ in the local part
3521 (suitably quoted, of course), incorrect values were put in $domain and
3522 $local_part during the evaluation of errors_copy.
3528 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3529 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3532 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3533 issue a MAIL command.
3535 PH/02 In an ACL statement such as
3537 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3539 if a client was not listed at all, or was listed with a value other than
3540 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3541 the condition was not true (as it should be), so access was not denied.
3542 The bug was that the ! inversion was incorrectly passed on to the second
3543 item. This has been fixed.
3545 PH/03 Added additional dnslists conditions == and =& which are different from
3546 = and & when the dns lookup returns more than one IP address.
3548 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3549 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3551 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3552 FSYNC, which compiles an option called disable_fsync that allows for
3553 bypassing fsync(). The documentation is heavily laced with warnings.
3555 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3557 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3558 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3559 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3560 including adding "make clean"; (3) Added -fPIC when compiling the test
3561 dynamically loaded module, to get rid of a warning.
3563 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3564 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3565 = 0s. The bug is otherwise harmless.
3567 PH/07 There was a bug in the dovecot authenticator such that the value of
3568 $auth1 could be overwritten, and so not correctly preserved, after a
3569 successful authentication. This usually meant that the value preserved by
3570 the server_setid option was incorrect.
3572 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3574 PH/09 Installed PCRE release 7.0.
3576 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3577 run for batched SMTP input. It is now run at the start of every message
3578 in the batch. While fixing this I discovered that the process information
3579 (output by running exiwhat) was not always getting set for -bs and -bS
3580 input. This is fixed, and it now also says "batched" for BSMTP.
3582 PH/11 Added control=no_pipelining.
3584 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3585 patch, slightly modified), and move the expansion of helo_data till after
3586 the connection is made in the smtp transport (so it can use these
3589 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3591 PH/14 Added log_selector = +pid.
3593 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3595 PH/16 Add ${if forany and ${if forall.
3597 PH/17 Added dsn_from option to vary the From: line in DSNs.
3599 PH/18 Flush SMTP output before performing a callout, unless control =
3600 no_callout_flush is set.
3602 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3603 was true (the default) a successful delivery failed to delete the retry
3604 item, thus causing premature timeout of the address. The bug is now
3607 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3609 PH/21 Long custom messages for fakedefer and fakereject are now split up
3610 into multiline responses in the same way that messages for "deny" and
3611 other ACL rejections are.
3613 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3614 with slight modification.
3616 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3617 draft, changing the syntax and factoring some duplicate code".
3619 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3620 for deliveries of the second and subsequent messages over the same SMTP
3623 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3624 ${reduce, with only minor "tidies".
3626 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3628 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3629 expansion side effects.
3631 PH/27 When a message times out after an over-quota error from an Exim-imposed
3632 quota, the bounce message says "mailbox is full". This message was not
3633 being given when it was a system quota that was exceeded. It now should
3636 MH/03 Made $recipients available in local_scan(). local_scan() already has
3637 better access to the recipient list through recipients_list[], but
3638 $recipients can be useful in postmaster-provided expansion strings.
3640 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3641 in the case of a MAIL command with additional options following the
3642 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3643 were accidentally chopped off.
3645 PH/29 SMTP synchronization checks are implemented when a command is read -
3646 there is a check that no more input is waiting when there shouldn't be
3647 any. However, for some commands, a delay in an ACL can mean that it is
3648 some time before the response is written. In this time, more input might
3649 arrive, invalidly. So now there are extra checks after an ACL has run for
3650 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3651 pipelining has not been advertised.
3653 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3655 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3656 returned in the SMTP response when smtp_return_error_details was false.
3657 This has been fixed.
3659 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3660 instead of the C I/O that was originally supplied, because problems were
3661 reported on Solaris.
3663 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3664 Exim which did not show up earlier: it was assuming that a call to
3665 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3666 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3667 was a macro that became an assignment, so it seemed to work. This has
3668 changed to a proper function call with a void return, hence the compile
3669 error. Exim's code has been fixed.
3671 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3674 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3676 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3678 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3679 because it turns out to be too big or there is a timeout) there is no
3680 "Completed" line in the log. When some messages of this type were
3681 selected by exigrep, they were listed as "not completed". Others were
3682 picked up by some special patterns. I have improved the selection
3683 criteria to be more general.
3685 PH/38 The host_find_failed option in the manualroute router can now be set
3686 to "ignore", to completely ignore a host whose IP address cannot be
3687 found. If all hosts are ignored, the behaviour is controlled by the new
3688 host_all_ignored option.
3690 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3691 homing or because of multiple MX records with /mx) generated more than
3692 one IP address, and the following item turned out to be the local host,
3693 all the secondary addresses of the first item were incorrectly removed
3694 from the list, along with the local host and any following hosts (which
3695 is what is supposed to happen).
3697 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3698 whoever called Exim into the -H file. In the case of the daemon it was
3699 behaving confusingly. When first started, it used values for whoever
3700 started the daemon, but after a SIGHUP it used the Exim user (because it
3701 calls itself on a restart). I have changed the code so that it now always
3704 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3705 message are rejected with the same error (e.g. no authentication or bad
3706 sender address), and a DATA command is nevertheless sent (as can happen
3707 with PIPELINING or a stupid MUA), the error message that was given to the
3708 RCPT commands is included in the rejection of the DATA command. This is
3709 intended to be helpful for MUAs that show only the final error to their
3712 PH/42 Another patch from the Sieve maintainer.
3714 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3715 Eximstats - Fixed some broken HTML links and added missing column headers
3717 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3718 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3720 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3722 PH/43 Yet another patch from the Sieve maintainer.
3724 PH/44 I found a way to check for a TCP/IP connection going away before sending
3725 the response to the final '.' that terminates a message, but only in the
3726 case where the client has not sent further data following the '.'
3727 (unfortunately, this is allowed). However, in many cases there won't be
3728 any further data because there won't be any more messages to send. A call
3729 to select() can be used: if it shows that the input is "ready", there is
3730 either input waiting, or the socket has been closed. An attempt to read
3731 the next input character can distinguish the two cases. Previously, Exim
3732 would have sent an OK response which the client would never have see.
3733 This could lead to message repetition. This fix should cure that, at
3734 least in a lot of common cases.
3736 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3737 advertised in response to EHLO.
3743 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3744 fixed by 4.65/MH/01 (is this a record?) are fixed:
3746 (i) An empty string was always treated as zero by the numeric comparison
3747 operators. This behaviour has been restored.
3749 (ii) It is documented that the numeric comparison operators always treat
3750 their arguments as decimal numbers. This was broken in that numbers
3751 starting with 0 were being interpreted as octal.
3753 While fixing these problems I realized that there was another issue that
3754 hadn't been noticed. Values of message_size_limit (both the global option
3755 and the transport option) were treated as octal if they started with 0.
3756 The documentation was vague. These values are now always treated as
3757 decimal, and I will make that clear in the documentation.
3763 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3764 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3767 MH/01 Don't check that the operands of numeric comparison operators are
3768 integers when their expansion is in "skipping" mode (fixes bug
3769 introduced by 4.64-PH/07).
3771 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3772 child addresses, Exim now panics and dies. Previously, because the count
3773 is held in a short int, deliveries were likely to be lost. As such a
3774 large number of recipients for a single message is ridiculous
3775 (performance will be very, very poor), I have chosen to impose a limit
3776 rather than extend the field.
3782 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3783 leftover -K file (the existence of which was triggered by #402).
3784 While we were at it, introduced process PID as part of the -K
3785 filename. This should rule out race conditions when creating
3788 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3789 processing considerably. Previous code took too long for large mails,
3790 triggering a timeout which in turn triggers #401.
3792 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3793 in the DK code in transports.c. sendfile() is not really portable,
3794 hence the _LINUX specificness.
3796 TF/01 In the add_headers option to the mail command in an Exim filter,
3797 there was a bug that Exim would claim a syntax error in any
3798 header after the first one which had an odd number of characters
3801 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3802 callout verification, Exim cached a "reject" for the entire domain. This
3803 is correct for most verifications, but it is not correct for a recipient
3804 verification with use_sender or use_postmaster set, because in that case
3805 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3806 case of MAIL FROM:<> rejection from other early rejections (e.g.
3807 rejection of HELO). When verifying a recipient using a non-null MAIL
3808 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3809 Whatever the result of the callout, the value of the domain cache is
3810 left unchanged (for any other kind of callout, getting as far as trying
3811 RCPT means that the domain itself is ok).
3813 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3816 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3817 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3818 session, the daemon ignores ECONNECT errors and logs others; it now
3819 ignores EPIPE as well.
3821 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3822 (quoted-printable decoding).
3824 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3825 later the small subsequent patch to fix an introduced bug.
3827 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3829 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3831 PH/08 An error is now given if message_size_limit is specified negative.
3833 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3834 to be given (somewhat) arbitrary names.
3836 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3839 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3840 miscellaneous code fixes
3842 PH/10 Added the log_reject_target ACL modifier to specify where to log
3845 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3846 hostname. This is wrong, because it relates to the incoming message (and
3847 probably the interface on which it is arriving) and not to the outgoing
3848 callout (which could be using a different interface). This has been
3849 changed to use the value of the helo_data option from the smtp transport
3850 instead - this is what is used when a message is actually being sent. If
3851 there is no remote transport (possible with a router that sets up host
3852 addresses), $smtp_active_hostname is used.
3854 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3855 tweaks were necessary in order to get it to work (see also 21 below):
3856 (a) The code assumed that strncpy() returns a negative number on buffer
3857 overflow, which isn't the case. Replaced with Exim's string_format()
3859 (b) There were several signed/unsigned issues. I just did the minimum
3860 hacking in of casts. There is scope for a larger refactoring.
3861 (c) The code used strcasecmp() which is not a standard C function.
3862 Replaced with Exim's strcmpic() function.
3863 (d) The code set only $1; it now sets $auth1 as well.
3864 (e) A simple test gave the error "authentication client didn't specify
3865 service in request". It would seem that Dovecot has changed its
3866 interface. Fortunately there's a specification; I followed it and
3867 changed what the client sends and it appears to be working now.
3869 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3872 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3873 address A is aliased to B and C, where B exists and C does not. Without
3874 -v the output is "A verified" because verification stops after a
3875 successful redirection if more than one address is generated. However,
3876 with -v the child addresses are also verified. Exim was outputting "A
3877 failed to verify" and then showing the successful verification for C,
3878 with its parentage. It now outputs "B failed to verify", showing B's
3879 parentage before showing the successful verification of C.
3881 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3882 look up a TXT record in a specific list after matching in a combined
3885 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3886 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3887 they consult the DNS. I had assumed they would set it the way they
3888 wanted; and indeed my experiments on Linux seem to show that in some
3889 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3890 To be on the safe side, however, I have now made the interface to
3891 host_find_byname() similar to host_find_bydns(), with an argument
3892 containing the DNS resolver options. The host_find_byname() function now
3893 sets these options at its start, just as host_find_bydns() does. The smtp
3894 transport options dns_qualify_single and dns_search_parents are passed to
3895 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3896 of host_find_byname() use the default settings of RES_DEFNAMES
3897 (qualify_single) but not RES_DNSRCH (search_parents).
3899 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3900 spool_read_header() do less string testing, by means of a preliminary
3901 switch on the second character of optional "-foo" lines. (This is
3902 overdue, caused by the large number of possibilities that now exist.
3903 Originally there were few.) While I was there, I also converted the
3904 str(n)cmp tests so they don't re-test the leading "-" and the first
3905 character, in the hope this might squeeze out yet more improvement.
3907 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3908 flag allowing group syntax was set by the header_syntax check but not
3909 turned off, possible causing trouble later; (2) The flag was not being
3910 set at all for the header_verify test, causing "group"-style headers to
3911 be rejected. I have now set it in this case, and also caused header_
3912 verify to ignore an empty address taken from a group. While doing this, I
3913 came across some other cases where the code for allowing group syntax
3914 while scanning a header line wasn't quite right (mostly, not resetting
3915 the flag correctly in the right place). These bugs could have caused
3916 trouble for malformed header lines. I hope it is now all correct.
3918 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3919 with the "reply" argument non-NULL. The code, however (which originally
3920 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3921 but it didn't always do it. This confused somebody who was copying the
3922 code for some other use. I have removed all the tests.
3924 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3925 feature that was used to support insecure browsers during the U.S. crypto
3926 embargo. It requires special client support, and Exim is probably the
3927 only MTA that supported it -- and would never use it because real RSA is
3928 always available. This code has been removed, because it had the bad
3929 effect of slowing Exim down by computing (never used) parameters for the
3930 RSA_EXPORT functionality.
3932 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3933 authenticator to fail if there's a tab character in the incoming data
3934 (there should never be unless someone is messing about, as it's supposed
3935 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3936 if the connection is using TLS or if the remote IP is the same as the
3937 local IP, and the "valid-client-cert option" if a client certificate has
3940 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3941 authenticators. This can be used for authorization after authentication
3942 succeeds. (In the case of plaintext, it servers for both authentication
3945 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3946 if any retry times were supplied.
3948 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3949 connection, where there is no client IP address to check. In this
3950 situation, the verify now always succeeds.
3952 PH/25 Applied John Jetmore's -Mset patch.
3954 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3956 PH/27 In a string expansion for a processed (not raw) header when multiple
3957 headers of the same name were present, leading whitespace was being
3958 removed from all of them, but trailing whitespace was being removed only
3959 from the last one. Now trailing whitespace is removed from each header
3960 before concatenation. Completely empty headers in a concatenation (as
3961 before) are ignored.
3963 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3964 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3966 PH/29 [Removed. This was a change that I later backed out, and forgot to
3967 correct the ChangeLog entry (that I had efficiently created) before
3968 committing the later change.]
3970 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3971 address errors (4xx response to RCPT) over the same connection as other
3972 messages routed to the same hosts. Such deliveries are always "forced",
3973 so retry times are not inspected. This resulted in far too many retries
3974 for the affected addresses. The effect occurred only when there were more
3975 hosts than the hosts_max_try setting in the smtp transport when it had
3976 the 4xx errors. Those hosts that it had tried were not added to the list
3977 of hosts for which the message was waiting, so if all were tried, there
3978 was no problem. Two fixes have been applied:
3980 (i) If there are any address or message errors in an SMTP delivery, none
3981 of the hosts (tried or untried) are now added to the list of hosts
3982 for which the message is waiting, so the message should not be a
3983 candidate for sending over the same connection that was used for a
3984 successful delivery of some other message. This seems entirely
3985 reasonable: after all the message is NOT "waiting for some host".
3986 This is so "obvious" that I'm not sure why it wasn't done
3987 previously. Hope I haven't missed anything, but it can't do any
3988 harm, as the worst effect is to miss an optimization.
3990 (ii) If, despite (i), such a delivery is accidentally attempted, the
3991 routing retry time is respected, so at least it doesn't keep
3992 hammering the server.
3994 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3995 in ${readsocket because some servers need this prod.
3997 PH/32 Added some extra debug output when updating a wait-xxx database.
3999 PH/33 The hint "could be header name not terminated by colon", which has been
4000 given for certain expansion errors for a long time, was not being given
4001 for the ${if def:h_colon_omitted{... case.
4003 PH/34 The spec says: "With one important exception, whenever a domain list is
4004 being scanned, $domain contains the subject domain." There was at least
4005 one case where this was not true.
4007 PH/35 The error "getsockname() failed: connection reset by peer" was being
4008 written to the panic log as well as the main log, but it isn't really
4009 panic-worthy as it just means the connection died rather early on. I have
4010 removed the panic log writing for the ECONNRESET error when getsockname()
4013 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4014 runs only) independently of the message's sender address. This meant
4015 that, if the 4xx error was in fact related to the sender, a different
4016 message to the same recipient with a different sender could confuse
4017 things. In particular, this can happen when sending to a greylisting
4018 server, but other circumstances could also provoke similar problems.
4019 I have changed the default so that the retry time for these errors is now
4020 based a combination of the sender and recipient addresses. This change
4021 can be overridden by setting address_retry_include_sender=false in the
4024 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4025 remote server are returned as part of bounce messages. This was not
4026 happening for LMTP over a pipe (the lmtp transport), but now it is the
4027 same for both kinds of LMTP.
4029 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4030 in header lines that were in fact CNAMEs. This is no longer the case.
4032 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4033 and queue runs started by the daemon processed all messages. This has
4034 been fixed so that -R and -S can now usefully be given with -q<time>.
4036 PH/40 Import PCRE release 6.7 (fixes some bugs).
4038 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4040 PH/42 Give an error if -q is specified more than once.
4042 PH/43 Renamed the variables $interface_address and $interface_port as
4043 $received_ip_address and $received_port, to make it clear that these
4044 values apply to message reception, and not to the outgoing interface when
4045 a message is delivered. (The old names remain recognized, of course.)
4047 PH/44 There was no timeout on the connect() call when using a Unix domain
4048 socket in the ${readsocket expansion. There now is.
4050 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4051 be meaningful with "accept".
4053 SC/01 Eximstats V1.43
4054 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4056 SC/02 Eximstats V1.44
4057 Use a glob alias rather than an array ref in the generated
4058 parser. This improves both readability and performance.
4060 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4061 Collect SpamAssassin and rejection statistics.
4062 Don't display local sender or destination tables unless
4063 there is data to show.
4064 Added average volumes into the top table text output.
4066 SC/04 Eximstats V1.46
4067 Collect data on the number of addresses (recipients)
4068 as well as the number of messages.
4070 SC/05 Eximstats V1.47
4071 Added 'Message too big' to the list of mail rejection
4072 reasons (thanks to Marco Gaiarin).
4074 SC/06 Eximstats V1.48
4075 Mainlog lines which have GMT offsets and are too short to
4076 have a flag are now skipped.
4078 SC/07 Eximstats V1.49 (Alain Williams)
4079 Added the -emptyok flag.
4081 SC/08 Eximstats V1.50
4082 Fixes for obtaining the IP address from reject messages.
4084 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4085 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4086 whitespace changes from 4.64-PH/27
4088 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4091 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4092 are found, allow negative numbers in numeric criteria)
4094 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4096 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4097 to match changes made in 4.64-PH/43
4099 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4101 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4102 "domain" is actually an IP address, and force a failure. This locks out
4103 those revolvers/nameservers that support "A-for-A" lookups, in
4104 contravention of the specifications.
4106 PH/48 When a host name was looked up from an IP address, and the subsequent
4107 forward lookup of the name timed out, the host name was left in
4108 $sender_host_name, contrary to the specification.
4110 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4111 restricted to single-key lookups, Exim was not diagnosing an error if
4112 * or *@ was used with a query-style lookup.
4114 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4116 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4117 long ago, but noone interested enough thought of it. Let's just say that
4118 the "1.1" means that there are some new functions that weren't there at
4119 some point in the past.
4121 PH/51 Error processing for expansion failure of helo_data from an smtp
4122 transport during callout processing was broken.
4124 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4125 tested/used via the -bh/-bhc/-bs options.
4127 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4128 bug, fixed in subsequent PCRE releases).
4130 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4131 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4133 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4139 SC/01 Use a glob alias rather than an array ref in eximstats generated
4140 parser. This improves both readability and performance.
4142 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4143 Don't display local sender or destination tables in eximstats unless
4144 there is data to show.
4145 Added average volumes into the eximstats top table text output.
4147 SC/03 Collect data on the number of addresses (recipients) as well
4148 as the number of messages in eximstats.
4150 TF/01 Correct an error in the documentation for the redirect router. Exim
4151 does (usually) call initgroups() when daemonizing.
4153 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4154 with consistent privilege compared to when running as a daemon.
4156 TF/03 Note in the spec that $authenticated_id is not set for local
4157 submissions from trusted users.
4159 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4160 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4162 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4163 by adding some example configuration directives to the default
4164 configuration file. A little bit of work is required to uncomment the
4165 directives and define how usernames and passwords are checked, but
4166 there is now a framework to start from.
4168 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4169 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4170 without this. I don't know how relevant this is to other LDAP libraries.
4172 PH/02 Add the verb name to the "unknown ACL verb" error.
4174 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4176 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4178 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4179 directory not expanded when it should be if an expanded home directory
4180 was set for the address (which is overridden by the transport).
4182 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4185 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4186 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4187 because it is too late at that time, and has no effect.
4189 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4190 security issue with \' (bugzilla #107). I could not use the
4191 PQescapeStringConn() function, because it needs a PGconn value as one of
4194 PH/08 When testing addresses using -bt, indicate those final addresses that
4195 are duplicates that would not cause an additional delivery. At least one
4196 person was confused, thinking that -bt output corresponded to deliveries.
4197 (Suppressing duplicates isn't a good idea as you lose the information
4198 about possibly different redirections that led to the duplicates.)
4200 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4201 systems where poll() doesn't work, in particular OS X.
4203 PH/10 Added more information to debugging output for retry time not reached.
4205 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4206 operations in malware.c.
4208 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4211 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4212 syslog_duplication set false, log lines that would normally be written
4213 both the the main log and to the reject log were not written to syslog at
4216 PH/14 In the default configuration, change the use of "message" in ACL warn
4217 statements to "add_header".
4219 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4220 not followed by a command (e.g. "seen endif").
4222 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4223 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4226 PH/17 Added extra conditions to the default value of delay_warning_condition
4230 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4231 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4232 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4235 The Auto-Submitted: and various List- headers are standardised, whereas I
4236 don't think Precedence: ever was.
4238 PH/18 Refactored debugging code in route_finduser() to show more information,
4239 in particular, the error code if getpwnam() issues one.
4241 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4242 This is apparently needed in addition to the PH/07 change above to avoid
4243 any possible encoding problems.
4245 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4246 but not after initializing Perl.
4248 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4249 output them only if debugging. By default they are written stderr,
4250 apparently, which is not desirable.
4252 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4255 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4258 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4260 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4261 authenticated or an ident call has been made. Suppress the default
4262 values for $authenticated_id and $authenticated_sender (but permit -oMai
4263 and -oMas) when testing with -bh.
4265 PH/24 Re-jigged the order of the tests in the default configuration so that the
4266 tests for valid domains and recipients precede the DNS black list and CSA
4267 tests, on the grounds that those ones are more expensive.
4269 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4270 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4271 This bug exists in every version of Exim that I still have, right back to
4274 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4275 However, an attempt to turn on case-sensitivity in a regex key by
4276 including (?-i) didn't work because the subject string was already
4277 lowercased, and the effects were non-intuitive. It turns out that a
4278 one-line patch can be used to allow (?-i) to work as expected.
4284 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4285 other effects) broke the use of negated acl sub-conditions.
4287 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4290 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4291 "Deny" causes Exim to reject the incoming connection with a 554 error.
4292 Unfortunately, if there is a major crisis, such as a disk failure,
4293 tcp-wrappers gives "deny", whereas what one would like would be some
4294 kind of temporary error. A kludge has been added to help with this.
4295 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4296 554 error is used if errno is still zero or contains ENOENT (which occurs
4297 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4300 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4302 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4303 errors. Otherwise a message that provokes a temporary error (when other
4304 messages do not) can cause a whole host to time out.
4306 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4307 addresses were routed directly to files or pipes from a redirect router.
4308 File deliveries just didn't batch; pipe deliveries might have suffered
4311 PH/06 A failure to get a lock for a hints database would erroneously always say
4312 "Failed to get write lock", even when it was really a read lock.
4314 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4315 of 0600. This has been changed to use the value of the lockfile_mode
4316 option (which defaults to 0600).
4318 PH/08 Applied small patch from the Sieve maintainer.
4320 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4321 folder from quota calculations, a direct delivery into this folder messed
4322 up the contents of the maildirsize file. This was because the regex was
4323 used only to exclude .Trash (or whatever) when the size of the mailbox
4324 was calculated. There was no check that a delivery was happening into an
4325 excluded directory. This bug has been fixed by ignoring all quota
4326 processing for deliveries into excluded directories.
4328 PH/10 Added the maildirfolder_create_regex option to appendfile.
4334 PH/01 The code for finding all the local interface addresses on a FreeBSD
4335 system running IPv6 was broken. This may well have applied to all BSD
4336 systems, as well as to others that have similar system calls. The broken
4337 code found IPv4 interfaces correctly, but gave incorrect values for the
4338 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4339 that it would not match correctly against @[] and not recognize the IPv6
4342 PH/02 The ipliteral router was not recognizing addresses of the form user@
4343 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4345 PH/03 Added disable_ipv6.
4347 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4348 decoded form, because it is most often used to construct To: headers
4349 lines in autoreplies, and the decoded form may well be syntactically
4350 invalid. However, $reply_address has leading white space removed, and all
4351 newlines turned into spaces so that the autoreply transport does not
4354 PH/05 If group was specified without a user on a router, and no group or user
4355 was specified on a transport, the group from the router was ignored.
4357 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4358 for visible compile-time settings that can be used to change these
4359 numbers, for those that want even more. Backwards compatibility with old
4360 spool files has been maintained. However, going back to a previous Exim
4361 release will lost any variables that are in spool files.
4363 PH/07 Two small changes when running in the test harness: increase delay when
4364 passing a TCP/IP connection to a new process, in case the original
4365 process has to generate a bounce, and remove special handling of
4366 127.0.0.2 (sic), which is no longer necessary.
4368 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4369 be the same on different OS.
4371 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4374 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4375 whether --show-vars was specified or not
4377 JJ/02 exipick: Added support for new ACL variable spool format introduced
4380 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4381 syntactically invalid From: or Reply-to: line, and a filter used this to
4382 generate an autoreply, and therefore failed to obtain an address for the
4383 autoreply, Exim could try to deliver to a non-existent relative file
4384 name, causing unrelated and misleading errors. What now happens is that
4385 it logs this as a hard delivery error, but does not attempt to create a
4388 PH/11 The exinext utility has a -C option for testing purposes, but although
4389 the given file was scanned by exinext itself; it wasn't being passed on
4390 when Exim was called.
4392 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4393 an end-of-file indication when reading a command response.
4395 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4396 compiled. In many other places in Exim, IPv6 addresses are always
4397 recognized, so I have changed this. It also means that IPv4 domain
4398 literals of the form [IPV4:n.n.n.n] are now always recognized.
4400 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4401 used if the router is not running as root, for example, when verifying at
4402 ACL time, or when using -bh. The debugging output from this situation was
4403 non-existent - all you got was a failure to exec. I have made two
4406 (a) Failures to set uid/gid, the current directory, or a process leader
4407 in a subprocess such as that created by queryprogram now generate
4408 suitable debugging output when -d is set.
4410 (b) The queryprogram router detects when it is not running as root,
4411 outputs suitable debugging information if -d is set, and then runs
4412 the subprocess without attempting to change uid/gid.
4414 PH/15 Minor change to Makefile for building test_host (undocumented testing
4417 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4418 additional section of a DNS packet that returns MX or SRV records.
4419 Instead, it always explicitly searches for A/AAAA records. This avoids
4420 major problems that occur when a DNS server includes only records of one
4421 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4422 fixed another bug: if SRV records were looked up and the corresponding
4423 address records were *not* found in the additional section, the port
4424 values from the SRV records were lost.
4426 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4427 using the correct key (the original address) when searching the retry
4428 rules in order to find which one to use for generating the retry hint.
4430 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4431 adding the default one. Similarly, if it contains a Reply-To: header, the
4432 errors_reply_to option, if set, is not used.
4434 PH/19 When calculating a retry time, Exim used to measure the "time since
4435 failure" by looking at the "first failed" field in the retry record. Now
4436 it does not use this if it is later than than the arrival time of the
4437 message. Instead it uses the arrival time. This makes for better
4438 behaviour in cases where some deliveries succeed, thus re-setting the
4439 "first failed" field. An example is a quota failure for a huge message
4440 when small messages continue to be delivered. Without this change, the
4441 "time since failure" will always be short, possible causing more frequent
4442 delivery attempts for the huge message than are intended.
4443 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4445 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4446 $1, $2, $3) because the numerical variables can be reset during some
4447 expansion items (e.g. "match"), thereby losing the authentication data.
4449 PH/21 Make -bV show the size of off_t variables so that the test suite can
4450 decide whether to run tests for quotas > 2G.
4452 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4453 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4454 filecount value is greater than 2G or if a quota value is greater than 2G
4455 on a system where the size of off_t is not greater than 4, a panic error
4458 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4459 never match. The debug and -bh output now contains an explicit error
4460 message indicating a malformed IPv4 address or mask.
4462 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4463 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4464 PH/23 above applies.
4466 PH/25 Do not write to syslog when running in the test harness. The only
4467 occasion when this arises is a failure to open the main or panic logs
4468 (for which there is an explicit test).
4470 PH/26 Added the /no_tell option to "control=freeze".
4472 PH/27 If a host name lookup failed very early in a connection, for example, if
4473 the IP address matched host_lookup and the reverse lookup yielded a name
4474 that did not have a forward lookup, an error message of the form "no IP
4475 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4476 could be logged. Now it outputs the IP address instead of "NULL".
4478 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4479 allows the sender and the authenticated sender to be set when
4480 submitting a message from within Exim. Since child_open_exim() is
4481 documented for local_scan(), the new function should be too.
4483 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4484 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4485 results in an empty string is now treated as unset.
4487 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4489 PH/31 Added sender_verify_fail logging option.
4491 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4492 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4493 tidied the source and removed it altogether.
4495 PH/33 When a queue run was abandoned because the load average was too high, a
4496 log line was always written; now it is written only if the queue_run log
4497 selector is set. In addition, the log line for abandonment now contains
4498 information about the queue run such as the pid. This is always present
4499 in "start" and "stop" lines but was omitted from the "abandon" line.
4501 PH/34 Omit spaces between a header name and the colon in the error message that
4502 is given when verify = headers_syntax fails (if there are lots of them,
4503 the message gets confusing).
4505 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4506 names, as there are now some PTR records that contain slashes. This check
4507 is only to protect against broken name servers that fall over on strange
4508 characters, so the fact that it applies to all lookups doesn't matter.
4510 PH/36 Now that the new test suite is complete, we can remove some of the
4511 special code in Exim that was needed for the old test suite. For example,
4512 sorting DNS records because real resolvers return them in an arbitrary
4513 order. The new test suite's fake resolver always returns records in the
4516 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4517 bounces) except when queue_only is set, to avoid logging races between
4518 the different processes.
4520 PH/38 Panic-die if .include specifies a non-absolute path.
4522 PH/39 A tweak to the "H" retry rule from its user.
4524 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4525 a label. They prevented compilation on older perls.
4527 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4528 a warning to be raised on newish perls.
4530 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4531 on queue. Changes to match documented behaviour of showing count of
4532 messages matching specified criteria.
4534 PH/40 Changed the default ident timeout from 30s to 5s.
4536 PH/41 Added support for the use of login_cap features, on those BSD systems
4537 that have them, for controlling the resources used by pipe deliveries.
4539 PH/42 The content-scanning code uses fopen() to create files in which to put
4540 message data. Previously it was not paying any attention to the mode of
4541 the files. Exim runs with umask(0) because the rest of the code creates
4542 files with open(), and sets the required mode explicitly. Thus, these
4543 files were ending up world-writeable. This was not a big issue, because,
4544 being within the spool directory, they were not world-accessible. I have
4545 created a function called modefopen, which takes an additional mode
4546 argument. It sets umask(777), creates the file, chmods it to the required
4547 mode, then resets the umask. All the relevant calls to fopen() in the
4548 content scanning code have been changed to use this function.
4550 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4551 to 24 hours. This avoids potential overflow problems when processing G
4552 and H retry rules. I suspect nobody ever tinkers with this value.
4554 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4556 PH/45 When the plaintext authenticator is running as a client, the server's
4557 challenges are checked to ensure they are valid base64 strings. By
4558 default, the authentication attempt is cancelled if an invalid string is
4559 received. Setting client_ignore_invalid_base64 true ignores these errors.
4560 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4561 they are received. Thus, the responses can be made to depend on the
4562 challenges. If an invalid string is ignored, an empty string is placed in
4565 PH/46 Messages that are created by the autoreply transport now contains a
4566 References: header, in accordance with RFCs 2822 and 3834.
4568 PH/47 Added authenticated_sender_force to the smtp transport.
4570 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4572 PH/49 Installed latest patch from the Sieve maintainer.
4574 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4575 was also set, the appendfile transport was unnecessarily scanning a
4576 directory of message files (e.g. for maildir delivery) to find the count
4577 of files (along with the size), even though it did not need this
4578 information. It now does the scan only if it needs to find either the
4579 size of the count of files.
4581 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4583 PH/52 Two bugs concerned with error handling when the smtp transport is
4586 (i) Exim was not creating retry information for temporary errors given
4587 for individual recipients after the DATA command when the smtp transport
4588 was used in LMTP mode. This meant that they could be retried too
4589 frequently, and not timed out correctly.
4591 (ii) Exim was setting the flag that allows error details to be returned
4592 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4593 recipients that were returned after the DATA command.
4595 PH/53 This is related to PH/52, but is more general: for any failing address,
4596 when detailed error information was permitted to be returned to the
4597 sender, but the error was temporary, then after the final timeout, only
4598 "retry timeout exceeded" was returned. Now it returns the full error as
4599 well as "retry timeout exceeded".
4601 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4602 do this, and (what is worse) MTAs that accept it.
4604 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4605 will now be deprecated.
4607 PH/56 New os.c-cygwin from the Cygwin maintainer.
4609 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4610 formats (previously only available in exim formats via -bpr, -bpru,
4611 and -bpra. Now also available in native and exiqgrep formats)
4613 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4614 with very large, slow to parse queues
4616 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4618 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4620 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4621 responses to authentication challenges, though it was showing the
4622 challenges; (ii) I've removed the CR characters from the debug output for
4625 PH/58 Allow for the insertion of a newline as well as a space when a string
4626 is turned into more than one encoded-word during RFC 2047 encoding. The
4627 Sieve code now uses this.
4629 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4630 data_4xx, lost_connection, tls_required.
4632 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4633 message was being sent as an SMTP response.
4635 PH/61 Add -l and -k options to exicyclog.
4637 PH/62 When verifying, if an address was redirected to one new address, so that
4638 verification continued, and the new address failed or deferred after
4639 having set something in $address_data, the value of $address_data was not
4640 passed back to the ACL. This was different to the case when no
4641 redirection occurred. The value is now passed back in both cases.
4643 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4644 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4645 use login_cap.h, so on its own it isn't the distinguishing feature. The
4646 new name refers directly to the setclassresources() function.
4648 PH/65 Added configuration files for NetBSD3.
4650 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4652 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4653 is preferred over IPv4.
4655 PH/68 The bounce_return_message and bounce_return_body options were not being
4656 honoured for bounces generated during the reception of non-SMTP messages.
4657 In particular, this applied to messages rejected by the ACL. This bug has
4658 been fixed. However, if bounce_return_message is true and bounce_return_
4659 body is false, the headers that are returned for a non-SMTP message
4660 include only those that have been read before the error was detected.
4661 (In the case of an ACL rejection, they have all been read.)
4663 PH/69 The HTML version of the specification is now built in a directory called
4664 spec_html instead of spec.html, because the latter looks like a path with
4665 a MIME-type, and this confuses some software.
4667 PH/70 Catch two compiler warnings in sieve.c.
4669 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4670 function verify_get_ident() calls ip_connect() to connect a socket, but
4671 if the "connect()" function timed out, ip_connect() used to close the
4672 socket. However, verify_get_ident() also closes the socket later, and in
4673 between Exim writes to the log, which may get opened at this point. When
4674 the socket was closed in ip_connect(), the log could get the same file
4675 descriptor number as the socket. This naturally causes chaos. The fix is
4676 not to close the socket in ip_connect(); the socket should be closed by
4677 the function that creates it. There was only one place in the code where
4678 this was missing, in the iplookup router, which I don't think anybody now
4679 uses, but I've fixed it anyway.
4681 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4682 well as to direct DNS lookups. Otherwise the handling of names in host
4683 lists is inconsistent and therefore confusing.
4689 PH/01 Two changes to the default runtime configuration:
4691 (1) Move the checks for relay_from_hosts and authenticated clients from
4692 after to before the (commented out) DNS black list checks.
4694 (2) Add control=submission to the relay_from_hosts and authenticated
4695 clients checks, on the grounds that messages accepted by these
4696 statements are most likely to be submissions.
4698 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4700 (1) Generate an error if the third argument for the ${prvs expansion is
4703 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4706 (3) Reset the variables that are obtained from the first argument of
4707 ${prvscheck and used in the second argument before leaving the code,
4708 because their memory is reclaimed, so using them afterwards may do
4711 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4712 one (it's much easier than Tom thought :-).
4714 (5) Because of (4), we can now allow for the use of $prvscheck_result
4715 inside the third argument.
4717 PH/03 For some reason, the default setting of PATH when running a command from
4718 a pipe transport was just "/usr/bin". I have changed it to
4721 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4722 anything to be listed in the output from -bV.
4724 PH/05 When a filter generated an autoreply, the entire To: header line was
4725 quoted in the delivery log line, like this:
4727 => >A.N.Other <ano@some.domain> <original@ddress> ...
4729 This has been changed so that it extracts the operative address. There
4730 may be more than one such address. If so, they are comma-separated, like
4733 => >ano@some.domain,ona@other.domain <original@ddress> ...
4735 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4736 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4737 not being looked up in the rDNS to get a host name, Exim was showing the
4738 IP address twice in Received: lines, even though the IP addresses were
4739 identical. For example:
4741 Received: from [1.2.3.4] (helo=[1.2.3.4])
4743 However, if the real host name was known, it was omitting the HELO data
4744 if it matched the actual IP address. This has been tidied up so that it
4745 doesn't show the same IP address twice.
4747 PH/07 When both +timestamp and +memory debugging was on, the value given by
4748 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4749 called by the debug printing, thereby overwriting the timestamp buffer.
4750 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4752 PH/08 When the original message was included in an autoreply transport, it
4753 always said "this is a copy of the message, including all the headers",
4754 even if body_only or headers_only was set. It now gives an appropriate
4757 PH/09 Applied a patch from the Sieve maintainer which:
4759 o fixes some comments
4760 o adds the (disabled) notify extension core
4761 o adds some debug output for the result of if/elsif tests
4762 o points to the current vacation draft in the documentation
4763 and documents the missing references header update
4767 o fixes a bug in processing the envelope test (when testing
4768 multiple envelope elements, the last element determined the
4771 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4772 Electronic Mail") by including:
4774 Auto-submitted: auto-generated
4776 in the messages that it generates (bounce messages and others, such as
4777 warnings). In the case of bounce messages for non-SMTP messages, there was
4778 also a typo: it was using "Auto_submitted" (underscore instead of
4779 hyphen). Since every message generated by Exim is necessarily in response
4780 to another message, thes have all been changed to:
4782 Auto-Submitted: auto-replied
4784 in accordance with these statements in the RFC:
4786 The auto-replied keyword:
4788 - SHOULD be used on messages sent in direct response to another
4789 message by an automatic process,
4791 - MUST NOT be used on manually-generated messages,
4793 - MAY be used on Delivery Status Notifications (DSNs) and Message
4794 Disposition Notifications (MDNs),
4796 - MUST NOT be used on messages generated by automatic or periodic
4797 processes, except for messages which are automatic responses to
4800 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4801 to the default Received: header definition.
4803 PH/12 Added log selector acl_warn_skipped (default on).
4805 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4806 variables because (a) they are in the wrong storage pool and (b) even if
4807 they were copied, it wouldn't work properly because of the caching.
4809 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4810 checking when decoding. Apparently there are clients that generate
4811 overlong encoded strings. Why am I not surprised?
4813 PH/15 If the first argument of "${if match_address" was not empty, but did not
4814 contain an "@" character, Exim crashed. Now it writes a panic log message
4815 and treats the condition as false.
4817 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4819 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4820 extension "envelope-auth". The code is finished and in agreement with
4821 other implementations, but there is no documentation so far and in fact,
4822 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4823 not changing the active code.
4825 Print executed "if" and "elsif" statements when debugging is used. This
4826 helps a great deal to understand what a filter does.
4828 Document more things not specified clearly in RFC3028. I had all this
4829 sorted out, when out of a sudden new issues came to my mind. Oops."
4831 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4834 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4835 canonical form (as documented). However, after a host name lookup from
4836 the IP address, check_host() was doing a simple string comparison with
4837 addresses acquired from the DNS when checking that the found name did
4838 have the original IP as one of its addresses. Since any found IPv6
4839 addresses are likely to be in abbreviated form, the comparison could
4840 fail. Luckily, there already exists a function for doing the comparison
4841 by converting both addresses to binary, so now that is used instead of
4842 the text comparison.
4844 PH/20 There was another similar case to PH/19, when a complete host name was
4845 given in a host list; looking up its IP address could give an abbreviated
4846 form, whereas the current host's name might or might not be abbreviated.
4847 The same fix has been applied.
4853 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4854 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4857 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4858 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4860 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4862 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4863 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4864 Also, for "Auto-Submitted", treat anything other than "no" as
4865 non-personal, in accordance with RFC 3834. (Previously it treated
4866 anything starting "auto-" as non-personal.)
4868 TF/01 The control=submission/name=... option had a problem with syntax
4869 errors if the name included a slash character. The /name= option
4870 now slurps the rest of the string, so it can include any characters
4871 but it must come last in the list of options (after /sender_retain
4874 PH/05 Some modifications to the interface to the fake nameserver for the new
4882 TK/01 Added the "success_on_redirect" address verification option. See
4883 NewStuff for rationale and an example.
4885 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4887 PH/02 Patch to exigrep to allow it to work on syslog lines.
4889 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4890 fread() to skip over the body file's header line, because in Cygwin the
4891 header line is locked and is inaccessible.
4893 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4894 co-exist for some time) to make it clear that it is the Exim ID that is
4895 referenced, not the Message-ID: header line.
4897 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4898 string_format() function, because snprintf() does not exist on all
4901 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4902 ${stat: expansion item.
4904 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4905 protocol synchronization error", to keep the pedants happy.
4907 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4908 well as for IRIX systems, when gcc is being used. See the host.c source
4911 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4913 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4916 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4917 transport and to the smtp transport in LMTP mode.
4919 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4921 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4922 run a filter in a subprocess. This could lead to confusion in subsequent
4923 lookups in the parent process. There should also be a search_tidyup() at
4924 the end of the subprocess.
4926 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4927 only if the host matched helo_try_verify_hosts, which caused the
4928 verification to occur when the EHLO/HELO command was issued. The ACL just
4929 tested the remembered result. Now, if a previous verification attempt has
4930 not happened, "verify = helo" does it there and then.
4932 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4934 TK/03 Fix log output including CR from clamd.
4936 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4937 exist provoked a memory error which could cause a segfault.
4939 PH/15 Installed PCRE 6.2
4941 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4943 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4944 of the problem. Specifically, suggested +O2 rather than +O1 for the
4947 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4949 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4950 record, the port was not correctly logged when the outgoing_port log
4951 selector was set (it logged the transort's default port).
4953 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4954 fallback_hosts, and "hosts" in the smtp transport.
4956 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4957 host errors such as "Connection refused".
4959 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4960 authentication with radiusclient 0.4.9:
4962 - Error returned from rc_read_config was caught wrongly
4963 - Username/password not passed on to radius server due to wrong length.
4965 The presumption is that some radiusclient API changes for 4.51/PH/17
4966 were not taken care of correctly. The code is still untested by me (my
4967 Linux distribution still has 0.3.2 of radiusclient), but it was
4968 contributed by a Radius user.
4970 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4971 expanding the "port" option of the smtp transport.
4973 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4974 while reading a MIME header. Thanks to Tom Hughes for a patch.
4976 PH/24 Include config.h inside local_scan.h so that configuration settings are
4979 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4980 that in an ACL for RCPT (for example), you can examine exactly what was
4983 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4984 commands, but it was not correctly comparing the address with the actual
4985 client host address. Thus, it would show the EHLO address in Received:
4986 header lines when this was not necessary.
4988 PH/27 Added the % operator to ${eval:}.
4990 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4991 it should be ignoring failures (because with -C, for example, it has lost
4992 privilege). It wasn't ignoring creation failures other than "already
4995 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4998 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4999 because an input line was too long, either on its own, or by virtue of
5000 too many continuations, the temporary file was not being removed, and the
5001 return code was incorrect.
5003 PH/31 Missing "BOOL" in function definition in filtertest.c.
5005 PH/32 Applied Sieve patches from the maintainer.
5007 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5009 PH/33 Added "verify = not_blind".
5011 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5012 Local/Makefile (with some defaults set). These are used in built scripts
5013 such as exicyclog, but they have never been used in the exim_install
5014 script (though there are many overriding facilities there). I have
5015 arranged that the exim_install script now takes note of these two
5018 PH/35 Installed configuration files for Dragonfly.
5020 PH/36 When a locally submitted message by a trusted user did not contain a
5021 From: header, and the sender address was obtained from -f or from an SMTP
5022 MAIL command, and the trusted user did not use -F to supply a sender
5023 name, $originator_name was incorrectly used when constructing a From:
5024 header. Furthermore, $originator_name was used for submission mode
5025 messages from external hosts without From: headers in a similar way,
5026 which is clearly wrong.
5028 PH/37 Added control=suppress_local_fixups.
5030 PH/38 When log_selector = +received_sender was set, and the addition of the
5031 sender made the log line's construction buffer exactly full, or one byte
5032 less than full, an overflow happened when the terminating "\n" was
5035 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5036 when the result of a list match is failure because a DNS lookup failed.
5038 PH/40 RM_COMMAND is now used in the building process.
5040 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5041 the "build-* directories that it finds.
5043 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5044 address in a domain literal was a prefix of an interface address.
5046 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5047 when verifying a sender address, unless rewrite_headers is false.
5049 PH/44 Wrote a long comment about why errors_to addresses are verified as
5050 recipients, not senders.
5052 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5053 the ratelimit ACL was added.
5055 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5057 PH/46 Added extra information about PostgreSQL errors to the error string.
5059 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5060 avoiding the need to install special zones in a real server. This is
5061 backwards compatible; if it can't find the fake resolver, it drops back.
5062 Thus, both old and new test suites can be run.
5064 TF/02 Added util/ratelimit.pl
5066 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5067 clock is set back in time.
5069 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5070 Candler <B.Candler@pobox.com>.
5072 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5073 OK for addresses that are the result of redirections.
5075 PH/48 A number of further additions for the benefit of the new test suite,
5076 including a fake gethostbyname() that interfaces to the fake DNS resolver
5079 TF/06 The fix for widen_domains has also been applied to qualify_single and
5080 search_parents which are the other dnslookup options that can cause
5083 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5086 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5088 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5089 service" if it encountered an MX record with an empty target hostname.
5090 The message is now "an MX or SRV record indicated no SMTP service".
5092 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5093 if the DNS of the sending site is misconfigured. This is quite a
5094 common situation. This change restores the behaviour of treating a
5095 helo verification defer as a failure.
5097 PH/51 If self=fail was set on a router, the bounce message did not include the
5098 actual error message.
5104 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5106 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5107 command itself ended in error, the underlying message about the transport
5108 filter timeout was being overwritten with the pipe command error. Now the
5109 underlying error message should be appended to the second error message.
5111 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5113 PH/02 Reduce the amount of output that "make" produces by default. Full output
5114 can still be requested.
5116 PH/03 The warning log line about a condition test deferring for a "warn" verb
5117 was being output only once per connection, rather than after each
5118 occurrence (because it was using the same function as for successful
5119 "warn" verbs). This seems wrong, so I have changed it.
5121 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5122 it should not have, which might have caused a crash in the right
5123 circumstances, but probably never did.
5125 PH/04 Installed a modified version of Tony Finch's patch to make submission
5126 mode fix the return path as well as the Sender: header line, and to
5127 add a /name= option so that you can make the user's friendly name appear
5130 TF/03 Added the control = fakedefer ACL modifier.
5132 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5133 Mark Lowes for thorough testing.
5135 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5137 TK/03 Merged latest SRS patch from Miles Wilton.
5139 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5140 with the definition in sysexits.h (which is #included earlier).
5141 Fortunately, Exim does not actually use EX_OK. The code used to try to
5142 preserve the sysexits.h value, by assuming that macro definitions were
5143 scanned for macro replacements. I have been disabused of this notion,
5144 so now the code just undefines EX_OK before #including unistd.h.
5146 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5147 in the smtp transport. When a block could not be written in a single
5148 write() function, the timeout was being re-applied to each part-write.
5149 This seems wrong - if the receiver was accepting one byte at a time it
5150 would take for ever. The timeout is now adjusted when this happens. It
5151 doesn't have to be particularly precise.
5153 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5154 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5156 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5157 without a domain if the check to <postmaster@domain> fails.
5159 SC/01 Eximstats: added -xls and the ability to specify output files
5160 (patch written by Frank Heydlauf).
5162 SC/02 Eximstats: use FileHandles for outputting results.
5164 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5166 SC/04 Eximstats: fixed display of large numbers with -nvr option
5168 SC/05 Eximstats: fixed merging of reports with empty tables.
5170 SC/06 Eximstats: added the -include_original_destination flag
5172 SC/07 Eximstats: removed tabs and trailing whitespace.
5174 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5176 TK/06 MBOX spool code: Add real "From " MBOX separator line
5177 so the .eml file is really in mbox format (even though
5178 most programs do not really care). Patch from Alex Miller.
5180 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5181 The latter is generated from $received_to and is only set if the
5182 message has one envelope recipient. SA can use these headers,
5183 obviously out-of-the-box. Patch from Alex Miller.
5185 PH/08 The ${def test on a variable was returning false if the variable's
5186 value was "0", contrary to what the specification has always said!
5187 The result should be true unless the variable is empty.
5189 PH/09 The syntax error of a character other than { following "${if
5190 def:variable_name" (after optional whitespace) was not being diagnosed.
5191 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5192 accidental colon was present, for example, could give incorrect results.
5194 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5195 result is used (not including appendfile, where other changes are about
5198 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5199 This involved changing a lot of size variables from int to off_t. It
5200 should work with maildirs and everything.
5202 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5203 spamd dying while we are connected to it.
5205 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5208 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5209 function for BDB 4.3.
5211 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5213 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5214 the sources for PCRE have been re-arranged and more files are now
5217 PH/15 The code I had for printing potentially long long variables in PH/11
5218 above was not the best (it lost precision). The length of off_t variables
5219 is now inspected at build time, and an appropriate printing format (%ld
5220 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5221 to be "long long int" or "long int". This is needed for the internal
5222 formatting function string_vformat().
5224 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5225 the configuration file to be ":syslog", then the script "guesses" where
5226 the logs files are, rather than using the compiled in default. In our
5227 case the guess is not the same as the compiled default, so the script
5228 suddenly stopped working when I started to use syslog. The patch checks
5229 to see if log_file_path is "". If so, it attempts to read it from exim
5230 with no configuration file to get the compiled in version, before it
5231 falls back to the previous guessing code."
5233 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5234 implementing BATV in an Exim configuration. See NewStuff for the gory
5237 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5238 Makefile that are specific to HP-UX.
5240 PH/18 If the "use_postmaster" option was set for a recipient callout together
5241 with the "random" option, the postmaster address was used as the MAIL
5242 FROM address for the random test, but not for the subsequent recipient
5243 test. It is now used for both.
5245 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5246 patch removes a few documentation additions to RFC 3028, because the
5247 latest draft now contains them. It adds the new en;ascii-case comparator
5248 and a new error check for 8bit text in MIME parts. Comparator and
5249 require names are now matched exactly. I enabled the subaddress
5250 extension, but it is not well tested yet (read: it works for me)."
5252 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5253 rework some of the code of TK/09 above to avoid the hardwired use of
5254 "%lld" and "long long". Replaced the call to snprintf() with a call to
5257 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5258 records point to non-existent hosts", "retry timeout exceeded", and
5259 "retry time not reached for any host after a long failure period".
5261 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5262 experimental DomainKeys support:
5264 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5265 (2) On an error such as an illegally used "control", the wrong name for
5266 the control was given.
5268 These problems did NOT occur unless DomainKeys support was compiled.
5270 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5272 PH/24 Added ${if match_ip condition.
5274 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5275 databases so that it will be absolutely obvious if a crash occurs in the
5276 DB library. This is a regular occurrence (often caused by mis-matched
5279 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5280 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5281 picked up on a user's system that detects such things. There doesn't seem
5282 to be a gcc warning option for this - only an attribute that has to be
5283 put on the function's prototype. It seems that in Fedora Core 4 they have
5284 set this on a number of new functions. No doubt there will be more in due
5287 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5288 specify a transport. However, if an address that was verified by such a
5289 router was the subject of a callout, Exim crashed because it tried to
5290 read the rcpt_include_affixes from the non-existent transport. Now it
5291 just assumes that the setting of that option is false. This bug was
5292 introduced by 4.51/PH/31.
5294 PH/28 Changed -d+all to exclude +memory, because that information is very
5295 rarely of interest, but it makes the output a lot bigger. People tend to
5296 do -d+all out of habit.
5298 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5299 code in os-type was giving problems when libc.so lives in lib64, like on
5302 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5303 aren't the modern standard, and it seems that some systems' include files
5304 don't always have them. Exim was already checking for some of the newer
5305 ones like T_AAAA, and defining it itself. I've added checks for all the
5306 record types that Exim uses.
5308 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5309 not automatically generating a new one, as it is supposed to. This
5310 prevented TLS from working. If the file did exist, but contained invalid
5311 data, a new version was generated, as expected. It was only the case of a
5312 non-existent file that was broken.
5314 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5315 with a change in libdomainkeys > 0.64.
5317 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5318 from DNS. If the selector record carries the flag, it now has
5319 precedence over the domain-wide flag.
5321 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5323 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5324 the use of a port name that isn't defined in /etc/services) occurred, the
5325 message was deferred as in a normal delivery, and thus remained on the
5326 spool, instead of being failed because of the mua_wrapper setting. This
5327 is now fixed, and I tidied up some of the mua_wrapper messages at the
5330 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5331 the messages in a hash of arrays rather than using individual hashes.
5332 This is a bit cleaner and results in dramatic memory savings, albeit
5333 at a slight CPU cost.
5335 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5336 as requested by Marc Sherman.
5338 SC/10 Eximstats: added histograms for user specified patterns as requested
5341 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5343 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5344 fopen() in the content-scanning modules that did not already have it.
5350 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5351 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5353 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5355 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5357 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5358 to test Sieve filters that use "vacation".
5360 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5361 that changes the way the GnuTLS parameters are stored in the cache file.
5362 The new format can be generated externally. For backward compatibility,
5363 if the data in the cache doesn't make sense, Exim assumes it has read an
5364 old-format file, and it generates new data and writes a new file. This
5365 means that you can't go back to an older release without removing the
5368 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5369 work if there are any delivery delays because "one_time" forces the
5370 parent to be marked "delivered", so its unseen clone is never tried
5371 again. For this reason, Exim now forbids the simultaneous setting of
5374 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5375 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5376 entry if you want to know the details. The fix, however, neglected to
5377 consider the case when local delivery batching is involved. The test for
5378 "previously delivered" was not happening when checking to see if an
5379 address could be batched with a previous (undelivered) one; under
5380 certain circumstances this could lead to multiple deliveries to the same
5383 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5384 in its include files, and this causes problems building Exim.
5386 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5387 header_syntax) but Exim was just ignoring anything given after a slash.
5388 In particular, this caused confusion with an attempt to use "verify =
5389 reverse_host_lookup/defer_ok". An error is now given when options are
5390 supplied for verify items that do not have them. (Maybe reverse_host_
5391 lookup should have a defer_ok option, but that's a different point.)
5393 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5394 defined by RFC 821) to 2048, because there were problems with some AUTH
5395 commands, and RFC 1869 says the size should be increased for extended
5396 SMTP commands that take arguments.
5398 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5401 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5402 "unknown" error; now it says that the functionality isn't in the binary.
5404 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5405 an address' error message when a string expansion fails (syntax or
5406 whatever). Otherwise the password may appear in the log. Following change
5407 PH/42 below, there is no longer a chance of it appearing in a bounce
5410 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5412 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5413 of its addresses was ever tried. (Bugzilla bug #2.)
5415 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5416 the result incorrectly in the debug output. (It correctly added a newline
5417 to what was transported.)
5419 TF/01 Added $received_time.
5421 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5422 commented out examples of how to interface to a virus scanner and to
5423 SpamAssassin. Also added commented examples of av_scanner and
5424 spamd_address settings.
5426 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5427 and controls are allowed in which ACLs. There were a couple of minor
5428 errors. Some of the entries in the conditions table (which is a table of
5429 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5430 negation of where the condition IS allowed.
5432 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5434 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5435 header file does not have a version number, so I've had to invent a new
5436 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5437 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5438 radiusclient), but it was contributed by a Radius user.
5440 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5441 files or directories, for OpenSSL.
5443 PH/19 When an Exim process that is running as root has to create an Exim log
5444 file, it does so in a subprocess that runs as exim:exim so as to get the
5445 ownership right at creation (otherwise, other Exim processes might see
5446 the file with the wrong ownership). There was no test for failure of this
5447 fork() call, which would lead to the process getting stuck as it waited
5448 for a non-existent subprocess. Forks do occasionally fail when resources
5449 run out. I reviewed all the other calls to fork(); they all seem to check
5452 PH/20 When checking for unexpected SMTP input at connect time (before writing
5453 the banner), Exim was not dealing correctly with a non-positive return
5454 from the read() function. If the client had disconnected by this time,
5455 the result was a log entry for a synchronization error with an empty
5456 string after "input=" when read() returned zero. If read() returned -1
5457 (an event I could not check), uninitialized data bytes were printed.
5458 There were reports of junk text (parts of files, etc) appearing after
5461 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5463 PH/22 Added support for macro redefinition, and (re)definition in between
5464 driver and ACL definitions.
5466 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5467 forgetting to use the resulting value; it was using the unexpanded value.
5469 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5470 hadn't been configured. The fix is from Juergen Kreileder, who
5471 understands it better than I do:
5473 "Here's what I see happening with three configured cyrus_sasl
5474 authenticators configured (plain, login, cram-md5):
5476 On startup auth_cyrus_sasl_init() gets called for each of these.
5477 This means three calls to sasl_listmech() without a specified mech_list.
5478 => SASL tests which mechs of all available mechs actually work
5479 => three warnings about OTP not working
5480 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5482 With the patch, sasl_listmech() also gets called three times. But now
5483 SASL's mech_list option is set to the server_mech specified in the the
5484 authenticator. Or in other words, the answer from sasl_listmech()
5485 gets limited to just the mech you're testing for (which is different
5487 => the return list contains just 'plain' or 'login', 'cram-md5' or
5488 nothing depending on the value of ob->server_mech.
5490 I've just tested the patch: Authentication still works fine,
5491 unavailable mechs specified in the exim configuration are still
5492 caught, and the auth.log warnings about OTP are gone."
5494 PH/25 When debugging is enabled, the contents of the command line are added
5495 to the debugging output, even when log_selector=+arguments is not
5498 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5499 answer is "GNU", and only if the return is "GNU/something" is the answer
5502 PH/27 $acl_verify_message is now set immediately after the failure of a
5503 verification in an ACL, and so is available in subsequent modifiers. In
5504 particular, the message can be preserved by coding like this:
5506 warn !verify = sender
5507 set acl_m0 = $acl_verify_message
5509 Previously, $acl_verify_message was set only while expanding "message"
5510 and "log_message" when a very denied access.
5512 PH/28 Modified OS/os.c-Linux with
5514 -#ifndef OS_LOAD_AVERAGE
5515 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5517 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5518 nomenclature these days.)
5520 PH/29 Installed patch from the Sieve maintainer that adds the options
5521 sieve_useraddress and sieve_subaddress to the redirect router.
5523 PH/30 In these circumstances:
5524 . Two addresses routed to the same list of hosts;
5525 . First host does not offer TLS;
5526 . First host accepts first address;
5527 . First host gives temporary error to second address;
5528 . Second host offers TLS and a TLS session is established;
5529 . Second host accepts second address.
5530 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5531 and peerdn, if requested) that were in fact used only for the second
5534 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5535 attention to any local part prefix or suffix that was matched by the
5536 router that accepted the address. It now behaves in the same way as it
5537 does for delivery: the affixes are removed from the local part unless
5538 rcpt_include_affixes is set on the transport.
5540 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5541 timeout during the DATA phase of an incoming message.
5543 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5544 applied a patch sanctioned by the Sieve maintainer.
5546 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5547 the uid or gid is negative. A case of a negative gid caused this to be
5548 noticed. The fix allows for either to be negative.
5550 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5551 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5554 PH/36 The change PH/12 above was broken. Fixed it.
5556 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5557 the grounds that routing the same address twice would always produce the
5558 same answer. This might have been true once, but it is certainly no
5559 longer true now. Routing a child address may depend on the previous
5560 routing that produced that child. Some complicated redirection strategies
5561 went wrong when messages had multiple recipients, and made Exim's
5562 behaviour dependent on the order in which the addresses were given.
5564 I have moved the duplicate checking until after the routing is complete.
5565 Exim scans the addresses that are assigned to local and remote
5566 transports, and removes any duplicates. This means that more work will be
5567 done, as duplicates will always all be routed, but duplicates are
5568 presumably rare, so I don't expect this is of any significance.
5570 For deliveries to pipes, files, and autoreplies, the duplicate checking
5571 still happens during the routing process, since they are not going to be
5574 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5575 It corrects a timeout issue with spamd. This is Ian's comment: "The
5576 background is that sometimes spamd either never reads data from a
5577 connection it has accepted, or it never writes response data. The exiscan
5578 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5579 blindly assumes that writes won't block so it may never time out."
5581 PH/39 Allow G after quota size as well as K and M.
5583 PH/40 The value set for $authenticated_id in an authenticator may not contain
5584 binary zeroes or newlines because the value is written to log lines and
5585 to spool files. There was no check on this. Now the value is run through
5586 the string_printing() function so that such characters are converted to
5587 printable escape sequences.
5589 PH/41 $message_linecount is a new variable that contains the total number of
5590 lines in the message. Compare $body_linecount, which is the count for the
5593 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5594 bounce and delay warning messages, except in certain special cases, which
5597 (a) An SMTP error message from a remote host;
5598 (b) A message specified in a :fail: redirection;
5599 (c) A message specified in a "fail" command in a system filter;
5600 (d) A message specified in a FAIL return from the queryprogram router;
5601 (e) A message specified by the cannot_route_message router option.
5603 In these cases only, Exim does include the error details in bounce and
5604 warning messages. There are also a few cases where bland messages such
5605 as "unrouteable address" or "local delivery error" are given.
5607 PH/43 $value is now also set for the "else" part of a ${run expansion.
5609 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5610 being worked on, but at least Exim now implements the latest version to
5613 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5614 process to complete was treated as a delivery failure, a timeout while
5615 writing the message to the pipe was logged, but erroneously treated as a
5616 successful delivery. Such timeouts include transport filter timeouts. For
5617 consistency with the overall process timeout, these timeouts are now
5618 treated as errors, giving rise to delivery failures by default. However,
5619 there is now a new Boolean option for the pipe transport called
5620 timeout_defer, which, if set TRUE, converts the failures into defers for
5621 both kinds of timeout. A transport filter timeout is now identified in
5624 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5625 systems where "make" and "gmake" are different, calling "gmake" at top
5626 level broke things. I've arranged for the value of $(MAKE) to be passed
5627 from the Makefile to this script so that it can call the same version of
5631 A note about Exim versions 4.44 and 4.50
5632 ----------------------------------------
5634 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5635 changes of various kinds. As a consequence, a big documentation update was
5636 needed. This delayed the release for rather longer than seemed good, especially
5637 in the light of a couple of (minor) security issues. Therefore, the changes
5638 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5639 release. So 4.44 and 4.50 are in effect two different branches that both start
5642 I have left the 4.50 change log unchanged; it contains all the changes since
5643 4.43. The change log for 4.44 is below; many of its items are identical to
5644 those for 4.50. This seems to be the most sensible way to preserve the
5645 historical information.
5651 1. Minor wording change to the doc/README.SIEVE file.
5653 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5654 computation of the current number of files was incorrect.
5656 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5657 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5660 4. Give more explanation in the error message when the command for a transport
5661 filter fails to execute.
5663 5. There are several places where Exim runs a non-Exim command in a
5664 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5665 was being done only for the command run by the queryprogram router. It is
5666 now done for all such subprocesses. The other cases are: ${run, transport
5667 filters, and the commands run by the lmtp and pipe transports.
5669 6. Added CONFIGURE_GROUP build-time option.
5671 7. Some older OS have a limit of 256 on the maximum number of file
5672 descriptors. Exim was using setrlimit() to set 1000 as a large value
5673 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5674 systems. I've change it so that if it can't get 1000, it tries for 256.
5676 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5677 was an oversight, and furthermore, ever since the addition of extra
5678 controls (e.g. 4.43/32), the checks on when to allow different forms of
5679 "control" were broken. There should now be diagnostics for all cases when a
5680 control that does not make sense is encountered.
5682 9. Added the /retain_sender option to "control=submission".
5684 10. $recipients is now available in the predata ACL (oversight).
5686 11. Tidy the search cache before the fork to do a delivery from a message
5687 received from the command line. Otherwise the child will trigger a lookup
5688 failure and thereby defer the delivery if it tries to use (for example) a
5689 cached ldap connection that the parent has called unbind on.
5691 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5692 of $address_data from the recipient verification was clobbered by the
5693 sender verification.
5695 13. The value of address_data from a sender verification is now available in
5696 $sender_address_data in subsequent conditions in the ACL statement.
5698 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5700 15. Added a new option "connect=<time>" to callout options, to set a different
5703 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5704 was its contents. (It was OK if the option was not defined at all.)
5706 17. A "Completed" log line is now written for messages that are removed from
5707 the spool by the -Mrm option.
5709 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5710 information about exactly what failed.
5712 19. Added -dd to debug only the daemon process.
5714 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5715 handles timeouts, both on the server side and network timeouts. Renamed the
5716 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5718 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5719 It is now set to "smtps".
5721 22. $host_address is now set to the target address during the checking of
5722 ignore_target_hosts.
5724 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5725 being passed; this would have caused $sender_host_name to have been used if
5726 matching the list had actually called for a host name (not very likely,
5727 since this list is usually IP addresses). A host name is now passed as
5730 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5731 code that specifies a non-restarting handler (typically sigaction() in
5732 modern systems) in an attempt to fix a rare and obscure crash bug.
5734 25. Narrowed the window for a race in the daemon that could cause it to ignore
5735 SIGCHLD signals. This is not a major problem, because they are used only to
5736 wake it up if nothing else does.
5738 26. A malformed maildirsize file could cause Exim to calculate negative values
5739 for the mailbox size or file count. Odd effects could occur as a result.
5740 The maildirsize information is now recalculated if the size or filecount
5743 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5744 support for a long time. Removed HAVE_SYS_VFS_H.
5746 28. Installed the latest version of exipick from John Jetmore.
5748 29. In an address list, if the pattern was not a regular expression, an empty
5749 subject address (from a bounce message) matched only if the pattern was an
5750 empty string. Non-empty patterns were not even tested. This was the wrong
5751 because it is perfectly reasonable to use an empty address as part of a
5752 database query. An empty address is now tested by patterns that are
5753 lookups. However, all the other forms of pattern expect the subject to
5754 contain a local part and a domain, and therefore, for them, an empty
5755 address still always fails if the pattern is not itself empty.
5757 30. Exim went into a mad DNS loop when attempting to do a callout where the
5758 host was specified on an smtp transport, and looking it up yielded more
5759 than one IP address.
5761 31. Re-factored the code for checking spool and log partition space into a
5762 function that finds that data and another that does the check. The former
5763 is then used to implement four new variables: $spool_space, $log_space,
5764 $spool_inodes, and $log_inodes.
5766 32. The RFC2047 encoding function was originally intended for short strings
5767 such as real names; it was not keeping to the 75-character limit for
5768 encoded words that the RFC imposes. It now respects the limit, and
5769 generates multiple encoded words if necessary. To be on the safe side, I
5770 have increased the buffer size for the ${rfc2047: expansion operator from
5773 33. It is now permitted to omit both strings after an "if" condition; if the
5774 condition is true, the result is "true". As before, when the second string
5775 is omitted, a false condition yields an empty string. This makes it less
5776 cumbersome to write custom ACL and router conditions.
5778 34. Failure to deliver a bounce message always caused it to be frozen, even if
5779 there was an errors_to setting on the router. The errors_to setting is now
5782 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5783 canonical form (fully expanded) before being placed in
5784 $sender_host_address.
5786 36. The table in the code that translates DNS record types into text (T_A to
5787 "A" for instance) was missing entries for NS and CNAME. It is just possible
5788 that this could have caused confusion if both these types were looked up
5789 for the same domain, because the text type is used as part of Exim's
5790 per-process caching. But the chance of anyone hitting this buglet seems
5793 37. The dnsdb lookup has been extended in a number of ways.
5795 (1) There is a new type, "zns", which walks up the domain tree until it
5796 finds some nameserver records. It should be used with care.
5798 (2) There is a new type, "mxh", which is like "mx" except that it returns
5799 just the host names, not the priorities.
5801 (3) It is now possible to give a list of domains (or IP addresses) to be
5802 looked up. The behaviour when one of the lookups defers can be
5803 controlled by a keyword.
5805 (4) It is now possible to specify the separator character for use when
5806 multiple records are returned.
5808 38. The dnslists ACL condition has been extended: it is now possible to supply
5809 a list of IP addresses and/or domains to be looked up in a particular DNS
5812 39. Added log_selector=+queue_time_overall.
5814 40. When running the queue in the test harness, wait just a tad after forking a
5815 delivery process, to get repeatability of debugging output.
5817 41. Include certificate and key file names in error message when GnuTLS fails
5818 to set them up, because the GnuTLS error message doesn't include the name
5819 of the failing file when there is a problem reading it.
5821 42. Allow both -bf and -bF in the same test run.
5823 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5825 44. The "Exiscan patch" is now merged into the mainline Exim source.
5827 45. Sometimes the final signoff response after QUIT could fail to get
5828 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5829 < 0 before doing a fflush(). This bug looks as though it goes back to the
5830 introduction of TLS in release 3.20, but "sometimes" must have been rare
5831 because the tests only now provoked it.
5833 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5834 (this can affect the format of dates).
5836 47. exim_tidydb, when checking for the continued existence of a message for
5837 which it has found a message-specific retry record, was not finding
5838 messages that were in split spool directories. Consequently, it was
5839 deleting retry records that should have stayed in existence.
5841 48. Steve fixed some bugs in eximstats.
5843 49. The SPA authentication driver was not abandoning authentication and moving
5844 on to the next authenticator when an expansion was forced to fail,
5845 contradicting the general specification for all authenticators. Instead it
5846 was generating a temporary error. It now behaves as specified.
5848 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5849 (the order specifies the preference for clients). The order is now AES256,
5850 AES128, 3DES, ARCFOUR128.
5852 51. Small patch to Sieve code - explicitly set From: when generating an
5855 52. Exim crashed if a remote delivery caused a very long error message to be
5856 recorded - for instance if somebody sent an entire SpamAssassin report back
5857 as a large number of 550 error lines. This bug was coincidentally fixed by
5858 increasing the size of one of Exim's internal buffers (big_buffer) that
5859 happened as part of the Exiscan merge. However, to be on the safe side, I
5860 have made the code more robust (and fixed the comments that describe what
5863 53. Now that there can be additional text after "Completed" in log lines (if
5864 the queue_time_overall log selector is set), a one-byte patch to exigrep
5865 was needed to allow it to recognize "Completed" as not the last thing in
5868 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5869 patch that reportedly fixes this has been added. I am not expert enough to
5870 create a test for it. This is what the patch creator wrote:
5872 "I found a little strange behaviour of ldap code when working with
5873 Windows 2003 AD Domain, where users was placed in more than one
5874 Organization Units. When I tried to give exim partial DN, the exit code
5875 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5876 But simultaneously result of request was absolutely normal ldap result,
5877 so I produce this patch..."
5879 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5880 REFERENCE, so I have modified the code to exclude the patch when that macro
5883 55. Some experimental protocols are using DNS PTR records for new purposes. The
5884 keys for these records are domain names, not reversed IP addresses. The
5885 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5886 leaves it alone. Component reversal etc. now happens only for IP addresses.
5889 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5891 57. Double the size of the debug message buffer (to 2048) so that more of very
5892 long debug lines gets shown.
5894 58. The exicyclog utility now does better if the number of log files to keep
5895 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5897 59. Two changes related to the smtp_active_hostname option:
5899 (1) $smtp_active_hostname is now available as a variable.
5900 (2) The default for smtp_banner uses $smtp_active_hostname instead
5901 of $primary_hostname.
5903 60. The host_aton() function is supposed to be passed a string that is known
5904 to be a valid IP address. However, in the case of IPv6 addresses, it was
5905 not checking this. This is a hostage to fortune. Exim now panics and dies
5906 if the condition is not met. A case was found where this could be provoked
5907 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5908 components; fortuitously, this particular loophole had already been fixed
5909 by change 4.50/55 above.
5911 If there are any other similar loopholes, the new check in host_aton()
5912 itself should stop them being exploited. The report I received stated that
5913 data on the command line could provoke the exploit when Exim was running as
5914 exim, but did not say which command line option was involved. All I could
5915 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5916 running as the user.
5919 61. There was a buffer overflow vulnerability in the SPA authentication code
5920 (which came originally from the Samba project). I have added a test to the
5921 spa_base64_to_bits() function which I hope fixes it.
5924 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5925 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5927 63. The daemon start-up calls getloadavg() while still root for those OS that
5928 need the first call to be done as root, but it missed one case: when
5929 deliver_queue_load_max is set with deliver_drop_privilege. This is
5930 necessary for the benefit of the queue runner, because there is no re-exec
5931 when deliver_drop_privilege is set.
5933 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5934 This has been fixed.
5936 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5937 was in use, was not putting the data itself into the right store pool;
5938 consequently, it could be overwritten for a subsequent message in the same
5939 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5942 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5944 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5945 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5946 the function were treating the return as a boolean value, which happened to
5947 work because 0=false and not-0=true, but is not correct code.
5949 68. The host_aton() function was not handling scoped IPv6 addresses (those
5950 with, for example, "%eth0" on the end) correctly.
5952 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5953 negated items (that is, ~something) in unsigned ints. Some compilers
5954 apparently mutter when there is no cast.
5956 70. If an address verification called from an ACL failed, and did not produce a
5957 user-specific message (i.e. there was only a "system" message), nothing was
5958 put in $acl_verify_message. In this situation, it now puts the system
5961 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5962 session; change 4.31/43 added the unwanted input to the log line - except
5963 that it did not do this in the start of session case. It now does.
5965 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5966 This is wrong and can cause the other end to generate a synchronization
5967 error if it is another Exim or anything else that does the synchronization
5968 check. A QUIT command is no longer sent after a timeout.
5970 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5971 during host lookups.
5973 74. The defer_ok option of callout verification was not working if it was used
5974 when verifying addresses in header lines, that is, for this case:
5976 verify = header_sender/callout=defer_ok
5978 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5979 those file descriptors could be used for SMTP connections. If anything
5980 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5981 could be sent to the SMTP client, causing chaos. The daemon now opens
5982 stdin, stdout, and stderr to /dev/null when it puts itself into the
5985 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5986 log by default. The user can override this with suitable Perl magic.
5988 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5989 the log message when discard triggers, was not working for the DATA ACL or
5990 for the non-SMTP ACL.
5992 78. Error message wording change in sieve.c.
5994 79. If smtp_accept_max_per_host was set, the number of connections could be
5995 restricted to fewer than expected, because the daemon was trying to set up
5996 a new connection before checking whether the processes handling previous
5997 connections had finished. The check for completed processes is now done
5998 earlier. On busy systems, this bug wouldn't be noticed because something
5999 else would have woken the daemon, and it would have reaped the completed
6002 80. If a message was submitted locally by a user whose login name contained one
6003 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6004 It caused a spool format error. I have fixed the spool reading code. A
6005 related problem was that the "from" clause in the Received: line became
6006 illegal because of the space(s). It is now covered by ${quote_local_part.
6008 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6011 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6013 83. Patch from Sieve maintainer to fix unterminated string problem in
6014 "vacation" handling.
6016 84. Some minor changes to the Linux configuration files to help with other
6017 OS variants using glibc.
6019 85. One more patch for Sieve to update vacation handling to latest spec.
6022 ----------------------------------------------------
6023 See the note above about the 4.44 and 4.50 releases.
6024 ----------------------------------------------------
6030 1. Change 4.43/35 introduced a bug that caused file counts to be
6031 incorrectly computed when quota_filecount was set in an appendfile
6034 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6035 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6038 3. Give more explanation in the error message when the command for a transport
6039 filter fails to execute.
6041 4. There are several places where Exim runs a non-Exim command in a
6042 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6043 was being done only for the command run by the queryprogram router. It is
6044 now done for all such subprocesses. The other cases are: ${run, transport
6045 filters, and the commands run by the lmtp and pipe transports.
6047 5. Some older OS have a limit of 256 on the maximum number of file
6048 descriptors. Exim was using setrlimit() to set 1000 as a large value
6049 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6050 systems. I've change it so that if it can't get 1000, it tries for 256.
6052 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6053 was an oversight, and furthermore, ever since the addition of extra
6054 controls (e.g. 4.43/32), the checks on when to allow different forms of
6055 "control" were broken. There should now be diagnostics for all cases when a
6056 control that does not make sense is encountered.
6058 7. $recipients is now available in the predata ACL (oversight).
6060 8. Tidy the search cache before the fork to do a delivery from a message
6061 received from the command line. Otherwise the child will trigger a lookup
6062 failure and thereby defer the delivery if it tries to use (for example) a
6063 cached ldap connection that the parent has called unbind on.
6065 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6066 of $address_data from the recipient verification was clobbered by the
6067 sender verification.
6069 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6070 was its contents. (It was OK if the option was not defined at all.)
6072 11. A "Completed" log line is now written for messages that are removed from
6073 the spool by the -Mrm option.
6075 12. $host_address is now set to the target address during the checking of
6076 ignore_target_hosts.
6078 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6079 being passed; this would have caused $sender_host_name to have been used if
6080 matching the list had actually called for a host name (not very likely,
6081 since this list is usually IP addresses). A host name is now passed as
6084 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6085 code that specifies a non-restarting handler (typically sigaction() in
6086 modern systems) in an attempt to fix a rare and obscure crash bug.
6088 15. Narrowed the window for a race in the daemon that could cause it to ignore
6089 SIGCHLD signals. This is not a major problem, because they are used only to
6090 wake it up if nothing else does.
6092 16. A malformed maildirsize file could cause Exim to calculate negative values
6093 for the mailbox size or file count. Odd effects could occur as a result.
6094 The maildirsize information is now recalculated if the size or filecount
6097 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6098 support for a long time. Removed HAVE_SYS_VFS_H.
6100 18. Updated exipick to current release from John Jetmore.
6102 19. Allow an empty sender to be matched against a lookup in an address list.
6103 Previously the only cases considered were a regular expression, or an
6106 20. Exim went into a mad DNS lookup loop when doing a callout where the
6107 host was specified on the transport, if the DNS lookup yielded more than
6110 21. The RFC2047 encoding function was originally intended for short strings
6111 such as real names; it was not keeping to the 75-character limit for
6112 encoded words that the RFC imposes. It now respects the limit, and
6113 generates multiple encoded words if necessary. To be on the safe side, I
6114 have increased the buffer size for the ${rfc2047: expansion operator from
6117 22. Failure to deliver a bounce message always caused it to be frozen, even if
6118 there was an errors_to setting on the router. The errors_to setting is now
6121 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6122 canonical form (fully expanded) before being placed in
6123 $sender_host_address.
6125 24. Updated eximstats to version 1.33
6127 25. Include certificate and key file names in error message when GnuTLS fails
6128 to set them up, because the GnuTLS error message doesn't include the name
6129 of the failing file when there is a problem reading it.
6131 26. Expand error message when OpenSSL has problems setting up cert/key files.
6134 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6135 (this can affect the format of dates).
6137 28. exim_tidydb, when checking for the continued existence of a message for
6138 which it has found a message-specific retry record, was not finding
6139 messages that were in split spool directories. Consequently, it was
6140 deleting retry records that should have stayed in existence.
6142 29. eximstats updated to version 1.35
6143 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6144 1.35 - bugfix such that pie charts by volume are generated correctly
6146 30. The SPA authentication driver was not abandoning authentication and moving
6147 on to the next authenticator when an expansion was forced to fail,
6148 contradicting the general specification for all authenticators. Instead it
6149 was generating a temporary error. It now behaves as specified.
6151 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6152 (the order specifies the preference for clients). The order is now AES256,
6153 AES128, 3DES, ARCFOUR128.
6155 31. Small patch to Sieve code - explicitly set From: when generating an
6158 32. Exim crashed if a remote delivery caused a very long error message to be
6159 recorded - for instance if somebody sent an entire SpamAssassin report back
6160 as a large number of 550 error lines. This bug was coincidentally fixed by
6161 increasing the size of one of Exim's internal buffers (big_buffer) that
6162 happened as part of the Exiscan merge. However, to be on the safe side, I
6163 have made the code more robust (and fixed the comments that describe what
6166 33. Some experimental protocols are using DNS PTR records for new purposes. The
6167 keys for these records are domain names, not reversed IP addresses. The
6168 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6169 leaves it alone. Component reversal etc. now happens only for IP addresses.
6172 34. The host_aton() function is supposed to be passed a string that is known
6173 to be a valid IP address. However, in the case of IPv6 addresses, it was
6174 not checking this. This is a hostage to fortune. Exim now panics and dies
6175 if the condition is not met. A case was found where this could be provoked
6176 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6177 components; fortuitously, this particular loophole had already been fixed
6178 by change 4.50/55 or 4.44/33 above.
6180 If there are any other similar loopholes, the new check in host_aton()
6181 itself should stop them being exploited. The report I received stated that
6182 data on the command line could provoke the exploit when Exim was running as
6183 exim, but did not say which command line option was involved. All I could
6184 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6185 running as the user.
6188 35. There was a buffer overflow vulnerability in the SPA authentication code
6189 (which came originally from the Samba project). I have added a test to the
6190 spa_base64_to_bits() function which I hope fixes it.
6193 36. The daemon start-up calls getloadavg() while still root for those OS that
6194 need the first call to be done as root, but it missed one case: when
6195 deliver_queue_load_max is set with deliver_drop_privilege. This is
6196 necessary for the benefit of the queue runner, because there is no re-exec
6197 when deliver_drop_privilege is set.
6199 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6200 was in use, was not putting the data itself into the right store pool;
6201 consequently, it could be overwritten for a subsequent message in the same
6202 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6205 38. Sometimes the final signoff response after QUIT could fail to get
6206 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6207 < 0 before doing a fflush(). This bug looks as though it goes back to the
6208 introduction of TLS in release 3.20, but "sometimes" must have been rare
6209 because the tests only now provoked it.
6215 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6216 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6217 Now it returns an integer. A number of calls were still expecting a T/F
6218 return. Fortuitously, in all cases, the tests worked in OK situations,
6219 which is the norm. However, things would have gone wrong on any write
6220 failures on the smtp file descriptor. This function is used when sending
6221 messages over SMTP and also when doing verify callouts.
6223 2. When Exim is called to do synchronous delivery of a locally submitted
6224 message (the -odf or -odi options), it no longer closes stderr before doing
6227 3. Implemented the mua_wrapper option.
6229 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6231 5. Implemented the functions header_remove(), header_testname(),
6232 header_add_at_position(), and receive_remove_recipient(), and exported them
6235 6. If an ACL "warn" statement specified the addition of headers, Exim already
6236 inserted X-ACL-Warn: at the start if there was no header name. However, it
6237 was not making this test for the second and subsequent header lines if
6238 there were newlines in the string. This meant that an invalid header could
6239 be inserted if Exim was badly configured.
6241 7. Allow an ACL "warn" statement to add header lines at the start or after all
6242 the Received: headers, as well as at the end.
6244 8. Added the rcpt_4xx retry error code.
6246 9. Added postmaster_mailfrom=xxx to callout verification option.
6248 10. Added mailfrom=xxxx to the callout verification option, for verify=
6251 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6252 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6254 12. Inserted some casts to stop certain compilers warning when using pointer
6255 differences as field lengths or precisions in printf-type calls (mostly
6256 affecting debugging statements).
6258 13. Added optional readline() support for -be (dynamically loaded).
6260 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6261 same clock tick as a message's arrival, so that its received time was the
6262 same as the "first fail" time on the retry record, and that message
6263 remained on the queue past the ultimate address timeout, every queue runner
6264 would try a delivery (because it was past the ultimate address timeout) but
6265 after another failure, the ultimate address timeout, which should have then
6266 bounced the address, did not kick in. This was a "< instead of <=" error;
6267 in most cases the first failure would have been in the next clock tick
6268 after the received time, and all would be well.
6270 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6271 being recognized when the domain list was tested by the match_domain
6272 condition in an expansion string.
6274 16. Added the ${str2b64: operator.
6276 17. Exim was always calling setrlimit() to set a large limit for the number of
6277 processes, without checking whether the existing limit was already
6278 adequate. (It did check for the limit on file descriptors.) Furthermore,
6279 errors from getrlimit() and setrlimit() were being ignored. Now they are
6280 logged to the main and panic logs, but Exim does carry on, to try to do its
6281 job under whatever limits there are.
6283 18. Imported PCRE 5.0.
6285 19. Trivial typo in log message " temporarily refused connection" (the leading
6288 20. If the log selector return_path_on_delivery was set and an address was
6289 redirected to /dev/null, the delivery process crashed because it assumed
6290 that a return path would always be set for a "successful" delivery. In this
6291 case, the whole delivery is bypassed as an optimization, and therefore no
6294 21. Internal re-arrangement: the function for sending a challenge and reading
6295 a response while authentication was assuming a zero-terminated challenge
6296 string. It's now changed to take a pointer and a length, to allow for
6297 binary data in such strings.
6299 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6301 23. Exim was not respecting finduser_retries when seeking the login of the
6302 uid under which it was called; it was always trying 10 times. (The default
6303 setting of finduser_retries is zero.) Also, it was sleeping after the final
6304 failure, which is pointless.
6306 24. Implemented tls_on_connect_ports.
6308 25. Implemented acl_smtp_predata.
6310 26. If the domain in control=submission is set empty, Exim assumes that the
6311 authenticated id is a complete email address when it generates From: or
6312 Sender: header lines.
6314 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6315 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6316 chown and chgrp in /bin and hostname in /usr/bin.
6318 28. Exim was keeping the "process log" file open after each use, just as it
6319 does for the main log. This opens the possibility of it remaining open for
6320 long periods when the USR1 signal hits a daemon. Occasional processlog
6321 errors were reported, that could have been caused by this. Anyway, it seems
6322 much more sensible not to leave this file open at all, so that is what now
6325 29. The long-running daemon process does not normally write to the log once it
6326 has entered its main loop, and it closes the log before doing so. This is
6327 so that log files can straightforwardly be renamed and moved. However,
6328 there are a couple of unusual error situations where the daemon does write
6329 log entries, and I had neglected to close the log afterwards.
6331 30. The text of an SMTP error response that was received during a remote
6332 delivery was being truncated at 512 bytes. This is too short for some of
6333 the long messages that one sometimes sees. I've increased the limit to
6336 31. It is now possible to make retry rules that apply only when a message has a
6337 specific sender, in particular, an empty sender.
6339 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6340 it possible to be selective about when SMTP synchronization is enforced.
6342 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6344 32. Implemented hosts_connection_nolog.
6346 33. Added an ACL for QUIT.
6348 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6351 35. Added mailbox_size and mailbox_filecount to appendfile.
6353 36. Added control = no_multiline_responses to ACLs.
6355 37. There was a bug in the logic of the code that waits for the clock to tick
6356 in the case where the clock went backwards by a substantial amount such
6357 that the microsecond fraction of "now" was more than the microsecond
6358 fraction of "then" (but the whole seconds number was less).
6360 38. Added support for the libradius Radius client library this is found on
6361 FreeBSD (previously only the radiusclient library was supported).
6367 1. When certain lookups returned multiple values in the form name=value, the
6368 quoting of the values was not always being done properly. Specifically:
6369 (a) If the value started with a double quote, but contained no whitespace,
6371 (b) If the value contained whitespace other than a space character (i.e.
6372 tabs or newlines or carriage returns) it was not quoted.
6373 This fix has been applied to the mysql and pgsql lookups by writing a
6374 separate quoting function and calling it from the lookup code. The fix
6375 should probably also be applied to nisplus, ibase and oracle lookups, but
6376 since I cannot test any of those, I have not disturbed their existing code.
6378 2. A hit in the callout cache for a specific address caused a log line with no
6379 reason for rejecting RCPT. Now it says "Previous (cached) callout
6380 verification failure".
6382 3. There was an off-by-one bug in the queryprogram router. An over-long
6383 return line was truncated at 256 instead of 255 characters, thereby
6384 overflowing its buffer with the terminating zero. As well as fixing this, I
6385 have increased the buffer size to 1024 (and made a note to document this).
6387 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6388 when Exim is waiting for an SMTP response from a remote server, Exim
6389 restarts its select() call on the socket, thereby resetting its timeout.
6390 This is not a problem when such interrupts are rare. Somebody set up a cron
6391 job to run exiwhat every 2 minutes, which is less than the normal select()
6392 timeout (5 or 10 minutes). This meant that the select() timeout never
6393 kicked in because it was always reset. I have fixed this by comparing the
6394 time when an interrupt arrives with the time at the start of the first call
6395 to select(). If more time than the timeout has elapsed, the interrupt is
6396 treated as a timeout.
6398 5. Some internal re-factoring in preparation for the addition of Sieve
6399 extensions (by MH). In particular, the "personal" test is moved to a
6400 separate function, and given an option for scanning Cc: and Bcc: (which is
6401 not set for Exim filters).
6403 6. When Exim created an email address using the login of the caller as the
6404 local part (e.g. when creating a From: or Sender: header line), it was not
6405 quoting the local part when it contained special characters such as @.
6407 7. Installed new OpenBSD configuration files.
6409 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6410 try to make them clearer.
6412 9. Callout options, other than the timeout value, were being ignored when
6413 verifying sender addresses in header lines. For example, when using
6415 verify = header_sender/callout=no_cache
6417 the cache was (incorrectly) being used.
6419 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6420 only the Cygwin environment.
6422 11. When return_path_on_delivery was set as a log selector, if different remote
6423 addresses in the same message used different return paths and parallel
6424 remote delivery occurred, the wrong values would sometimes be logged.
6425 (Whenever a remote delivery process finished, the return path value from
6426 the most recently started remote delivery process was logged.)
6428 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6429 lines when AUTH and/or TLS are in use. This is the "received protocol"
6430 field. Exim used to use "asmtp" for authenticated SMTP, without any
6431 indication (in the protocol name) for TLS use. Now it follows the RFC and
6432 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6433 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6434 names appear in log lines as well as in Received: header lines.
6436 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6437 extensions, and comparison tests, and to fix some bugs.
6439 14. Changes to the "personal" filter test:
6441 (1) The test was buggy in that it was just doing the equivalent of
6442 "contains" tests on header lines. For example, if a user's address was
6443 anne@some.where, the "personal" test would incorrectly be true for
6445 To: susanne@some.where
6447 This test is now done by extracting each address from the header in turn,
6448 and checking the entire address. Other tests that are part of "personal"
6449 are now done using regular expressions (for example, to check local parts
6450 of addresses in From: header lines).
6452 (2) The list of non-personal local parts in From: addresses has been
6453 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6454 taken from the Sieve specification recommendations.
6456 (3) If the message contains any header line starting with "List-" it is
6457 treated as non-personal.
6459 (4) The test for "circular" in the Subject: header line has been removed
6460 because it now seems ill-conceived.
6462 15. Minor typos in src/EDITME comments corrected.
6464 16. Installed latest exipick from John Jetmore.
6466 17. If headers_add on a router specified a text string that was too long for
6467 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6468 of string_sprintf() is now avoided.
6470 18. $message_body_size was not set (it was always zero) when running the DATA
6471 ACL and the local_scan() function.
6473 19. For the "mail" command in an Exim filter, no default was being set for
6474 the once_repeat time, causing a random time value to be used if "once" was
6475 specified. (If the value happened to be <= 0, no repeat happened.) The
6476 default is now 0s, meaning "never repeat". The "vacation" command was OK
6477 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6478 (I found it when inspecting the code).
6480 20. There is now an overall timeout for performing a callout verification. It
6481 defaults to 4 times the callout timeout, which applies to individual SMTP
6482 commands during the callout. The overall timeout applies when there is more
6483 than one host that can be tried. The timeout is checked before trying the
6484 next host. This prevents very long delays if there are a large number of
6485 hosts and all are timing out (e.g. when the network connections are timing
6486 out). The value of the overall timeout can be changed by specifying an
6487 additional sub-option for "callout", called "maxwait". For example:
6489 verify = sender/callout=5s,maxwait=20s
6491 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6492 to the end before writing, but this should make it even safer).
6494 22. Exim was forgetting that it had advertised PIPELINING for the second and
6495 subsequent messages on an SMTP connection. It was also not resetting its
6496 memory on STARTTLS and an internal HELO.
6498 23. When Exim logs an SMTP synchronization error within a session, it now
6499 records whether PIPELINING has been advertised or not.
6501 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6502 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6503 rather than long int.
6505 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6507 26. Added the never_mail option to autoreply.
6513 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6514 crash if the getsockname() call failed; this can happen if a connection is
6515 closed very soon after it is established. The problem was simply in the
6516 order in which certain operations were done, causing Exim to try to write
6517 to the SMTP stream before it had set up the file descriptor. The bug has
6518 been fixed by making things happen in the correct order.
6524 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6525 before the connection was closed, thus losing the rejection response.
6527 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6528 some early Solaris releases, but causes trouble in current releases where
6529 socklen_t is defined.
6531 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6534 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6537 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6538 that releasing the top of it at the end releases what was used for sub-
6539 expansions (unless the block got too big). However, discard this block if
6540 the first thing is a variable or header, so that we can use its block when
6541 it is dynamic (useful for very large $message_headers, for example).
6543 6. Lookups now cache *every* query, not just the most recent. A new, separate
6544 store pool is used for this. It can be recovered when all lookup caches are
6545 flushed. Lookups now release memory at the end of their result strings.
6546 This has involved some general refactoring of the lookup sources.
6548 7. Some code has been added to the store_xxx() functions to reduce the amount
6549 of flapping under certain conditions.
6551 8. log_incoming_interface used to affect only the <= reception log lines. Now
6552 it causes the local interface and port to be added to several more SMTP log
6553 lines, for example "SMTP connection from", and rejection lines.
6555 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6557 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6559 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6560 could be overwritten at the end of the current message (or the start of a
6561 new message if it was set in a HELO ACL). The value is now preserved for
6562 the duration of the SMTP connection.
6564 12. If a transport had a headers_rewrite setting, and a matching header line
6565 contained an unqualified address, that address was qualified, even if it
6566 did not match any rewriting rules. The underlying bug was that the values
6567 of the flags that permit the existence of unqualified sender and recipient
6568 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6569 non-local messages, and by -bnq for local messages) were not being
6570 preserved with the message after it was received.
6572 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6573 "next input=" as part of the text comprising the host identity instead of
6574 the correct text. The code was using the same buffer for two different
6575 strings. However, depending on which order the printing function evaluated
6576 its arguments, the bug did not always show up. Under Linux, for example, my
6577 test suite worked just fine.
6579 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6580 doesn't work with some older versions of Perl. It has been changed to "my",
6581 which in any case is probably the better facility to use.
6583 15. A really picky compiler found some instances of statements for creating
6584 error messages that either had too many or two few arguments for the format
6587 16. The size of the buffer for calls to the DNS resolver has been increased
6588 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6589 for addresses that have a lot of PTR records. This alleviates a problem; it
6590 does not fully solve it.
6592 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6593 buffer now truncates the list and logs the incident, which is the same
6594 action as happens when Exim is looking up a host name and its aliases.
6595 Previously in this situation something unpredictable would happen;
6596 sometimes it was "internal error: store_reset failed".
6598 18. If a server dropped the connection unexpectedly when an Exim client was
6599 using GnuTLS and trying to read a response, the client delivery process
6600 crashed while trying to generate an error log message.
6602 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6603 string, for example:
6605 warn message = H1: something\nH2: something
6607 the text was added as a single header line from Exim's point of view
6608 though it ended up OK in the delivered message. However, searching for the
6609 second and subsequent header lines using $h_h2: did not work. This has been
6610 fixed. Similarly, if a system filter added multiple headers in this way,
6611 the routers could not see them.
6613 20. Expanded the error message when iplsearch is called with an invalid key to
6614 suggest using net-iplsearch in a host list.
6616 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6617 ACLs are no longer actually imposed (and a message to that effect is
6620 22. If a "gecos" field in a passwd entry contained escaped characters, in
6621 particular, if it contained a \" sequence, Exim got it wrong when building
6622 a From: or a Sender: header from that name. A second bug also caused
6623 incorrect handling when an unquoted " was present following a character
6624 that needed quoting.
6626 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6627 was not being matched caselessly.
6629 24. Arranged for all hyphens in the exim.8 source to be escaped with
6632 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6633 properly. Recipient callout cache records were still being keyed to include
6634 the sender, even when use_sender was set false. This led to far more
6635 callouts that were necessary. The sender is no longer included in the key
6636 when use_sender is false.
6638 26. Added "control = submission" modifier to ACLs.
6640 27. Added the ${base62d: operator to decode base 62 numbers.
6642 28. dnsdb lookups can now access SRV records.
6644 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6645 the configuration file.
6647 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6648 (-v) mode. This makes the output for a verbose queue run more intelligible.
6650 31. Added a use_postmaster feature to recipient callouts.
6652 32. Added the $body_zerocount variable, containing the number of binary zero
6653 bytes in the message body.
6655 33. The time of last modification of the "new" subdirectory is now used as the
6656 "mailbox time last read" when there is a quota error for a maildir
6659 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6661 35. Added +ignore_unknown as a special item in host lists.
6663 36. Code for decoding IPv6 addresses in host lists is now included, even if
6664 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6665 address was recognized as an IP address, but was then not correctly decoded
6666 into binary, causing unexpected and incorrect effects when compared with
6673 1. Very minor rewording of debugging text in manualroute to say "list of
6674 hosts" instead of "hostlist".
6676 2. If verify=header_syntax was set, and a header line with an unqualified
6677 address (no domain) and a large number of spaces between the end of the
6678 name and the colon was received, the reception process suffered a buffer
6679 overflow, and (when I tested it) crashed. This was caused by some obsolete
6680 code that should have been removed. The fix is to remove it!
6682 3. When running in the test harness, delay a bit after writing a bounce
6683 message to get a bit more predictability in the log output.
6685 4. Added a call to search_tidyup() just before forking a reception process. In
6686 theory, someone could use a lookup in the expansion of smtp_accept_max_
6687 per_host which, without the tidyup, could leave open a database connection.
6689 5. Added the variables $recipient_data and $sender_data which get set from a
6690 lookup success in an ACL "recipients" or "senders" condition, or a router
6691 "senders" option, similar to $domain_data and $local_part_data.
6693 6. Moved the writing of debug_print from before to after the "senders" test
6696 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6697 problems for message scanning, either using a data ACL, or using
6698 local_scan() because the Received: header was not generated till after they
6699 were called (in order to set the time as the time of reception completion).
6700 I have revised the way this works. The header is now generated after the
6701 body is received, but before the ACL or local_scan() are called. After they
6702 are run, the timestamp in the header is updated.
6708 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6709 before starting a queue runner without re-exec. This happened only when
6710 deliver_drop_privilege was set or when the Exim user was set to root. The
6711 effect of the bug was that timeouts during subsequent deliveries caused
6712 crashes instead of being properly handled. The handler is now left at its
6713 default (and expected) setting.
6715 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6716 message, again when deliver_drop_privilege is set or Exim is run as root.
6717 The bug described in (1) was not present in this case, but the tidying up
6718 of the other signals was missing. I have made the two cases consistent.
6720 3. The ignore_target_hosts setting on a manualroute router was being ignored
6721 for hosts that were looked up using the /MX notation.
6723 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6726 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6727 operated on the sender address. After changing the $sender_address to <>
6728 for the sender address verify, Exim was re-instated it as the original
6729 (before rewriting) address, but remembering that it had rewritten it, so it
6730 wasn't rewriting it again. This bug also had the effect of breaking the
6731 sender address verification caching when the sender address was rewritten.
6733 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6734 This has been changed so that if the ip literal address matches
6735 ignore_target_hosts, the router declines.
6737 7. Added expansion conditions match_domain, match_address, and match_local_
6738 part (NOT match_host).
6740 8. The placeholder for the Received: header didn't have a length field set.
6742 9. Added code to Exim itself and to exim_lock to test for a specific race
6743 condition that could lead to file corruption when using MBX delivery. The
6744 issue is with the lockfile that is created in /tmp. If this file is removed
6745 after a process has opened it but before that process has acquired a lock,
6746 there is the potential for a second process to recreate the file and also
6747 acquire a lock. This could lead to two Exim processes writing to the file
6748 at the same time. The added code performs the same test as UW imapd; it
6749 checks after acquiring the lock that its file descriptor still refers to
6750 the same named file.
6752 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6753 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6756 11. Added the smtp_active_hostname option. If used, this will typically be made
6757 to depend on the incoming interface address. Because $interface_address is
6758 not set up until the daemon has forked a reception process, error responses
6759 that can happen earlier (such as "too many connections") no longer contain
6762 12. If an expansion in a condition on a "warn" statement fails because a lookup
6763 defers, the "warn" statement is abandoned, and the next ACL statement is
6764 processed. Previously this caused the whole ACL to be aborted.
6766 13. Added the iplsearch lookup type.
6768 14. Added ident_timeout as a log selector.
6770 15. Added tls_certificate_verified as a log selector.
6772 16. Added a global option tls_require_ciphers (compare the smtp transport
6773 option of the same name). This controls incoming TLS connections.
6775 17. I finally figured out how to make tls_require_ciphers do a similar thing
6776 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6777 before starting the TLS session.
6779 18. Tabs are now shown as \t in -bP output.
6781 19. If the log selector return_path_on_delivery was set, Exim crashed when
6782 bouncing a message because it had too many Received: header lines.
6784 20. If two routers both had headers_remove settings, and the first one included
6785 a superfluous trailing colon, the final name in the first list and the
6786 first name in the second list were incorrectly joined into one item (with a
6787 colon in the middle).
6793 1. Added -C and -D options to the exinext utility, mainly to make it easier
6794 to include in the automated testing, but these could be helpful when
6795 multiple configurations are in use.
6797 2. The exinext utility was not formatting the output nicely when there was
6798 an alternate port involved in the retry record key, nor when there was a
6799 message id as well (for retries that were specific to a specific message
6800 and a specific host). It was also confused by IPv6 addresses, because of
6801 the additional colons they contain. I have fixed the IPv4 problem, and
6802 patched it up to do a reasonable job for IPv6.
6804 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6805 delivery, the log line now contains "pipelined" if PIPELINING was used.
6807 4. An SMTP transport process used to panic and die if the bind() call to set
6808 an explicit outgoing interface failed. This has been changed; it is now
6809 treated in the same way as a connect() failure.
6811 5. A reference to $sender_host_name in the part of a conditional expansion
6812 that was being skipped was still causing a DNS lookup. This no longer
6815 6. The def: expansion condition was not recognizing references to header lines
6816 that used bh_ and bheader_.
6818 7. Added the _cache feature to named lists.
6820 8. The code for checking quota_filecount in the appendfile transport was
6821 allowing one more file than it should have been.
6823 9. For compatibility with Sendmail, the command line option
6831 and sets the incoming protocol and host name (for trusted callers). The
6832 host name and its colon can be omitted when only the protocol is to be set.
6833 Note the Exim already has two private options, -pd and -ps, that refer to
6834 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6835 "s", but I don't think that's a major issue.
6837 10. A number of refactoring changes to the code, none of which should affect
6840 (a) The number of logging options was getting close to filling up the
6841 32-bit word that was used as a bit map. I have split them into two classes:
6842 those that are passed in the argument to log_write(), and those that are
6843 only ever tested independently outside of that function. These are now in
6844 separate 32-bit words, so there is plenty of room for expansion again.
6845 There is no change in the user interface or the logging behaviour.
6847 (b) When building, for example, log lines, the code previously used a
6848 macro that called string_cat() twice, in order to add two strings. This is
6849 not really sufficiently general. Furthermore, there was one instance where
6850 it was actually wrong because one of the argument was used twice, and in
6851 one call a function was used. (As it happened, calling the function twice
6852 did not affect the overall behaviour.) The macro has been replaced by a
6853 function that can join an arbitrary number of extra strings onto a growing
6856 (c) The code for expansion conditions now uses a table and a binary chop
6857 instead of a serial search (which was left over from when there were very
6858 few conditions). Also, it now recognizes conditions like "pam" even when
6859 the relevant support is not compiled in: a suitably worded error message is
6860 given if an attempt is made to use such a condition.
6862 11. Added ${time_interval:xxxxx}.
6864 12. A bug was causing one of the ddress fields not to be passed back correctly
6865 from remote delivery subprocesses. The field in question was not being
6866 subsequently used, so this caused to problems in practice.
6868 13. Added new log selectors queue_time and deliver_time.
6870 14. Might have fixed a bug in maildirsizefile handling that threw up
6871 "unexpected character" debug warnings, and recalculated the data
6872 unnecessarily. In any case, I expanded the warning message to give more
6875 15. Added the message "Restricted characters in address" to the statements in
6876 the default ACL that block characters like @ and % in local parts.
6878 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6879 Three changes have been made:
6881 (a) There was a serious bug; a negative response to MAIL caused the whole
6882 recipient domain to be cached as invalid, thereby blocking all messages
6883 to all local parts at the same domain, from all senders. This bug has
6884 been fixed. The domain is no longer cached after a negative response to
6885 MAIL if the sender used is not empty.
6887 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6890 (c) A new callout option, "use_sender" has been added for people who want
6891 the modified behaviour.
6897 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6900 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6901 indeed breaks things for older releases.
6903 3. Added additional logging to the case where there is a problem reading data
6904 from a filter that is running in a subprocess using a pipe, in order to
6905 try to track down a specific problem.
6907 4. Testing facility fudge: when running in the test harness and attempting
6908 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6909 getting "No route to host". Convert this to a timeout.
6911 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6914 6. Some OS don't have socklen_t but use size_t instead. This affects the
6915 fifth argument of getsockopt() amongst other things. This is now
6916 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6917 can be set for individual OS. I have set it for SunOS5, OSF1, and
6918 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6919 some earlier ones do not.
6921 7. Change 4.30/15 was not doing the test caselessly.
6923 8. The standard form for an IPv6 address literal was being rejected by address
6924 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6925 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6926 this, as well as the form without the "IPv6" on the front (but only when
6927 address literals are enabled, of course).
6929 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6931 10. Exim crashed if a message with an empty sender address specified by -f
6932 encountered a router with an errors_to setting. This could be provoked only
6933 by a command such as
6937 where an empty string was supplied; "<>" did not hit this bug.
6939 11. Installed PCRE release 4.5.
6941 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6942 remained set. It is now erased.
6944 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6945 times from message ids (which are base 36 rather than the normal 62).
6947 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6948 were being counted as actual protocol errors, and logged if the log
6949 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6950 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6951 and DATA following a set of rejected RCPTs do not count as protocol errors.
6952 In other words, Exim assumes they were pipelined, though this may not
6953 actually be the case. Of course, in all cases the client gets an
6954 appropriate error code.
6956 15. If a lookup fails in an ACL condition, a message about the failure may
6957 be available; it is used if testing the ACL cannot continue, because most
6958 such messages specify what the cause of the deferral is. However, some
6959 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6960 that caused an old message to be retained and used if a later statement
6961 caused a defer, replacing the real cause of the deferral.
6963 16. If an IP address had so many PTR records that the DNS lookup buffer
6964 was not large enough to hold them, Exim could crash while trying to process
6965 the truncated data. It now detects and logs this case.
6967 17. Further to 4.21/58, another change has been made: if (and only if) the
6968 first line of a message (the first header line) ends with CRLF, a bare LF
6969 in a subsequent header line has a space inserted after it, so as not to
6970 terminate the header.
6972 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6973 unnecessarily, used atoi() instead of strtol(), and didn't check the
6974 termination when getting file sizes from file names by regex.
6976 19. Completely re-implemented the support for maildirsize files, in the light
6977 of a number of problems with the previous contributed implementation
6978 (4.30/29). In particular:
6980 . If the quota is zero, the maildirsize file is maintained, but no quota is
6983 . If the maildir directory does not exist, it is created before any attempt
6984 to write a maildirsize file.
6986 . The quota value in the file is just a cache; if the quota is changed in
6987 the transport, the new value overrides.
6989 . A regular expression is available for excluding directories from the
6992 20. The autoreply transport checks the characters in options that define the
6993 message's headers; it allows continued headers, but it was checking with
6994 isspace() after an embedded newline instead of explicitly looking for a
6997 21. If all the "regular" hosts to which an address was routed had passed their
6998 expiry times, and had not reached their retry times, the address was
6999 bounced, even if fallback hosts were defined. Now Exim should go on to try
7002 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7003 equivalent code in the SMTP transport. Some hosts send humungous responses
7004 to HELO/EHLO, more than 1024 it seems.
7006 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7007 gives compiler warnings in some environments. I've now done it "properly",
7010 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7011 (because of problems with the built-in one) was declared to return uschar *
7012 instead of char *, causing compiler failure.
7014 25. Fixed a file descriptor leak when processing alias/forward files.
7016 26. Fixed a minor format string issue in dbfn.c.
7018 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7020 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7021 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7024 29. When a sender address is verified, it is cached, to save repeating the test
7025 when there is more than one recipient in a message. However, when the
7026 verification involves a callout, it is possible for different callout
7027 options to be set for different recipients. It is too complicated to keep
7028 track of this in the cache, so now Exim always runs a verification when a
7029 callout is required, relying on the callout cache for the optimization.
7030 The overhead is duplication of the address routing, but this should not be
7033 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7034 to be verified with callout=postmaster, and the main callout worked but the
7035 postmaster check failed, the verification correctly failed. However, if a
7036 subsequent RCPT command asked for sender verification *without* the
7037 postmaster check, incorrect caching caused this verification also to fail,
7040 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7041 it was not caching the DNS options (qualify_single, search_parents) that
7042 were used when the lookup failed. A subsequent lookup with different
7043 options therefore always gave the same answer, though there were cases
7044 where it should not have. (Example: a "domains = !$mx_any" option on a
7045 dnslookup router: the "domains" option is always processed without any
7046 widening, but the router might have qualify_single set.) Now Exim uses the
7047 cached value only when the same options are set.
7049 32. Added John Jetmore's "exipick" utility to the distribution.
7051 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7052 than a timeout (e.g. a certificate is required, and is not provided), an
7053 Exim server now closes the connection immediately. Previously it waited for
7054 the client to close - but if the client is SSL, it seems that they each
7055 wait for each other, leading to a delay before one of them times out.
7057 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7058 maintained 0.8.x compatibility because I don't think many are using it, and
7059 it is clearly obsolete.
7061 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7064 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7065 client certificate was expired. A simple patch fixes this, though I don't
7066 understand the full logic of why the verify callback is called multiple
7069 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7070 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7071 which causes problems with some clients (such as the Certicom SSL Plus
7072 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7073 disables the coutermeasure allowing Eudora to connect."
7075 38. Exim was not checking that a write() to a log file succeeded. This could
7076 lead to Bad Things if a log got too big, in particular if it hit a file
7077 size limit. Exim now panics and dies if it cannot write to a log file, just
7078 as it does if it cannot open a log file.
7080 39. Modified OS/Makefile-Linux so that it now contains
7082 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7084 The two -D definitions ensure that Exim is compiled with large file
7085 support, which makes it possible to handle log files that are bigger than
7088 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7089 instance) a domain was checked against a named list that involved a lookup,
7090 causing $domain_data to be set, then another domain was checked against the
7091 same list, then the first domain was re-checked, the value of $domain_data
7092 after the final check could be wrong. In particular, if the second check
7093 failed, it could be set empty. This bug probably also applied to
7096 41. The strip_trailing_dot option was not being applied to the address given
7097 with the -f command-line option.
7099 42. The code for reading a message's header from the spool was incrementing
7100 $received_count, but never initializing it. This meant that the value was
7101 incorrect (doubled) while delivering a message in the same process in which
7102 it was received. In the most common configuration of Exim, this never
7103 happens - a fresh exec is done - but it can happen when
7104 deliver_drop_privilege is set.
7106 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7107 it now includes up to 150 characters of the unexpected data in the log
7110 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7111 and building data strings. The size of both of these buffers was 10 000
7112 bytes - far larger than anybody would *ever* want, thought I. Needless to
7113 say, somebody hit the limit. I have increased the maximum line length to
7114 20 000 and the maximum data length of concatenated lines to 100 000. I have
7115 also fixed two bugs, because there was no checking on these buffers. Tsk,
7116 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7117 buffer is too small.
7119 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7120 lsearch lookups. Now it does.
7122 46. When parsing a route_list item in a manualroute router, a fixed-length
7123 buffer was used for the list of hosts. I made this 1024 bytes long,
7124 thinking that nobody would ever have a list of hosts that long. Wrong.
7125 Somebody had a whole pile of complicated expansion conditions, and the
7126 string was silently truncated, leading to an expansion error. It turns out
7127 that it is easier to change to an unlimited length (owing to other changes
7128 that have happened since this code was originally written) than to build
7129 structure for giving a limitation error. The length of the item that
7130 expands into the list of hosts is now unlimited.
7132 47. The lsearch lookup could not handle data where the length of text line was
7133 more than 4095 characters. Such lines were truncated, leading to shortened
7134 data being returned. It should now handle lines of any length.
7136 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7137 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7140 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7141 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7142 no longer happens in comment lines. A list of replacements is now placed
7143 at the head of all of the source files, except those whose only change is
7144 to replace PERL_COMMAND in the very first #! line.
7146 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7147 messages on the queue, with a bottom-up merge sort, using code contributed
7148 by Michael Haardt. This should make operations like -bp somewhat faster on
7149 large queues. It won't affect queue runners, except when queue_run_in_order
7152 51. Installed eximstats 1.31 in the distribution.
7154 52. Added support for SRV lookups to the dnslookup router.
7156 53. If an ACL referred to $message_body or $message_body_end, the value was not
7157 reset for any messages that followed in the same SMTP session.
7159 54. The store-handling optimization for building very long strings was not
7160 differentiating between the different store pools. I don't think this
7161 actually made any difference in practice, but I've tidied it.
7163 55. While running the routers to verify a sender address, $sender_address
7164 was still set to the sender address. This is wrong, because when routing to
7165 send a bounce to the sender, it would be empty. Therefore, I have changed
7166 it so that, while verifying a sender address, $sender_address is set to <>.
7167 (There is no change to what happens when verifying a recipient address.)
7169 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7170 target A or AAAA records (if not already returned) without resetting the
7171 qualify_single or search_parents options of the DNS resolver. These are
7172 inappropriate in this case because the targets of MX and SRV records must
7173 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7174 target that, when qualified, matched something in the local domain. These
7175 two options are now turned off when doing these lookups.
7177 57. It seems that at least some releases of Reiserfs (which does not have the
7178 concept of a fixed number of inodes) returns zero and not -1 for the
7179 number of available inodes. This interacted badly with check_spool_inodes,
7180 which assumed that -1 was the "no such thing" setting. What I have done is
7181 to check that the total number of inodes is greater than zero before doing
7182 the test of how many are available.
7184 58. When a "warn" ACL statement has a log_message modifier, the message is
7185 remembered, and not repeated. This is to avoid a lot of repetition when a
7186 message has many recipients that cause the same warning to be written.
7187 However, Exim was preserving the list of already written lines for an
7188 entire SMTP session, which doesn't seem right. The memory is now reset if a
7189 new message is started.
7191 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7192 debugging output unless log_rewrite was also set.
7194 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7195 of (int)(handle) when we know that handle contains (void *)(-1).
7197 61. The Exim daemon panic-logs an error return when it closes the incoming
7198 connection. However "connection reset by peer" seems to be common, and
7199 isn't really an error worthy of noting specially, so that particular error
7202 62. When Exim is trying to find all the local interfaces, it used to panic and
7203 die if the ioctl to get the interface flags failed. However, it seems that
7204 on at least one OS (Solaris 9) it is possible to have an interface that is
7205 included in the list of interfaces, but for which you get a failure error
7206 for this call. This happens when the interface is not "plumbed" into a
7207 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7208 failure of the "get flags" call assumes that the interface is down.
7210 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7211 makes life easier for people who are doing arithmetic on fields extracted
7212 from dates, where you often get leading zeros that should not be
7213 interpreted as octal.
7215 64. Added qualify_domain to the redirect router, to override the global
7218 65. If a pathologically long header line contained very many addresses (the
7219 report of this problem mentioned 10 000) and each of them was rewritten,
7220 Exim could use up a very large amount of memory. (It kept on making new
7221 copies of the header line as it rewrote, and never released the old ones.)
7222 At the expense of a bit more processing, the header rewriting function has
7223 been changed so that it no longer eats memory in this way.
7225 66. The generation of the Received: header has been moved from the time that a
7226 message starts to be received, to the time that it finishes. The timestamp
7227 in the Received: header should now be very close to that of the <= log
7228 line. There are two side-effects of this change:
7230 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7231 logged header lines no longer include the local Received: line, because
7232 it has not yet been created. The same applies to a copy of the message
7233 that is returned to a non-SMTP sender when a message is rejected.
7235 (b) When a filter file is tested using -bf, no additional Received: header
7236 is added to the test message. After some thought, I decided that this
7239 This change does not affect the value of $received_for. It is still set
7240 after address rewriting, but before local_scan() is called.
7242 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7244 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7245 gave an unhelpful panic error message, and a defer error. I have managed to
7246 change this behaviour so that it now rejects any supplied certificate,
7247 which seems right, as the list of acceptable certificates is empty.
7249 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7250 gave an unhelpful defer error. I have not managed to make this reject any
7251 supplied certificates, but the error message it gives is "no certificate
7252 supplied", which is not helpful.
7254 70. exigrep's output now also includes lines that are not associated with any
7255 message, but which match the given pattern. Implemented by a patch from
7256 Martin Sluka, which also tidied up the Perl a bit.
7258 71. Recipient callout verification, like sender verification, was using <> in
7259 the MAIL FROM command. This isn't really the right thing, since the actual
7260 sender may affect whether the remote host accepts the recipient or not. I
7261 have changed it to use the actual sender in the callout; this means that
7262 the cache record is now keyed on a recipient/sender pair, not just the
7263 recipient address. There doesn't seem to be a real danger of callout loops,
7264 since a callout by the remote host to check the sender would use <>.
7265 [SEE ABOVE: changed after hitting problems.]
7267 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7268 temporary errors. However, in the case of such a code being given after
7269 the end of a data transmission (i.e. after ".") Exim was failing to write
7270 a retry record for the message. (Yes, there was some broken host that was
7271 actually sending 8xx at this point.)
7273 73. An unknown lookup type in a host list could cause Exim to panic-die when
7274 the list was checked. (An example that provoked this was putting <; in the
7275 middle of a list instead of at the start.) If this happened during a DATA
7276 ACL check, a -D file could be left lying around. This kind of configuration
7277 error no longer causes Exim to die; instead it causes a defer error. The
7278 incident is still logged to the main and panic logs.
7280 74. Buglet left over from Exim 3 conversion. The message "too many messages
7281 in one connection" was written to the rejectlog but not the mainlog, except
7282 when address rewriting (yes!) was being logged.
7284 75. Added write_rejectlog option.
7286 76. When a system filter was run not as root (that is, when system_filter_user
7287 was set), the values of the $n variables were not being returned to the
7288 main process; thus, they were not subsequently available in the $sn
7291 77. Added +return_path_on_delivery log selector.
7293 78. A connection timeout was being treated differently from recipients deferred
7294 when testing hosts_max_try with a message that was older than the host's
7295 retry timeout. (The host should not be counted, thus allowing all hosts to
7296 be tried at least once before bouncing.) This may have been the cause of an
7297 occasionally reported bug whereby a message would remain on the queue
7298 longer than the retry timeout, but would be bounced if a delivery was
7299 forced. I say "may" because I never totally pinned down the problem;
7300 setting up timeout/retry tests is difficult. See also the next item.
7302 79. The ultimate address timeout was not being applied to errors that involved
7303 a combination of host plus message (for example, a timeout on a MAIL
7304 command). When an address resolved to a number of possible hosts, and they
7305 were not all tried for each delivery (e.g. because of hosts_max_try), a
7306 message could remain on the queue longer than the retry timeout.
7308 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7311 81. Fixed an obscure SMTP outgoing bug which required at least the following
7312 conditions: (a) there was another message waiting for the same server;
7313 (b) the server returned 5xx to all RCPT commands in the first message so
7314 that the message was not completed; (c) the server dropped the connection
7315 or gave a negative response to the RSET that Exim sends to abort the
7316 transaction. The observed case was a dropped connection after DATA that had
7317 been sent in pipelining mode. That is, the server had advertised PIPELINING
7318 but was not implementing it correctly. The effect of the bug was incorrect
7319 behaviour, such as trying another host, and this could lead to a crash.
7325 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7326 and daemon.c were passed as pointers to ints; they should have been
7327 pointers to socklen_t variables (which are typically unsigned ints).
7329 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7332 3. Fixed a really odd bug that affected only the testing scheme; patching a
7333 certain fixed string in the binary changed the value of another string that
7334 happened to be identical to the end of the original first string.
7336 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7337 name", it returns that address as the IP address. On some operating
7338 systems (e.g. Solaris), it also passes back the IP address string as the
7339 "host name". However, on others (e.g. Linux), it passes back an empty
7340 string. Exim wasn't checking for this, and was changing the host name to an
7341 empty string, assuming it had been canonicalized.
7343 5. Although rare, it is permitted to have more than one PTR record for a given
7344 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7345 all the names associated with an address, because they do in Solaris.
7346 However, it seems that they do not in Linux for data that comes from the
7347 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7348 I found this out when I moved to a new Linux workstation and tried to run
7349 the Exim test suite.
7351 To get round this problem I have changed the code so that it now does its
7352 own call to the DNS to look up PTR records when searching for a host name.
7353 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7354 addresses that are only in /etc/hosts are still found.
7356 This behaviour is, however, controlled by an option called host_lookup_
7357 order, which defaults to "bydns:byaddr". If people want to use the other
7358 order, or indeed, just use one or the other means of lookup, they can
7359 specify it in this variable.
7361 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7362 some operating systems, this comes back from gethostbyaddr() as an empty
7363 string, and this is what Exim used to test for. However, it seems that in
7364 other systems, "." is yielded. Exim now tests for this case too.
7366 7. The values of check_spool_space and check_log_space are now held internally
7367 as a number of kilobytes instead of an absolute number of bytes. If a
7368 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7369 kilobyte. This means that much larger values can be stored.
7371 8. Exim monitor: an attempt to get the action menu when not actually pointing
7372 at a message produces an empty menu entitled "No message selected". This
7373 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7374 no entries in it ("Shell widget menu has zero width and/or height"). So I
7375 have added a single, blank menu entry in this case.
7377 9. Added ${quote_local_part.
7379 10. MIME decoding is now applied to the contents of Subject: header lines when
7382 11. Now that a reference to $sender_host_address automatically causes a reverse
7383 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7384 host lookup before query-style lookups in lists that might use this
7385 variable. This has therefore been abolished, and the "net-" prefix is no
7386 longer necessary for query-style lookups.
7388 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7389 have been a typo for LFLAGS, so it has been changed.
7391 13. The install script calls Exim with "-C /dev/null" in order to find the
7392 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7393 to be output. However, since Exim outputs its version number before the
7394 error, it didn't break the script. It just looked ugly. I fixed this by
7395 always allowing "-C /dev/null" if the caller is root.
7397 14. Ignore overlarge ACL variable number when reading spool file - insurance
7398 against a later release with more variables having written the file.
7400 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7401 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7402 this, as well as the form without the "IPv6" on the front.
7404 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7405 OS/Makefile-Darwin file.
7407 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7408 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7410 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7411 variable. However, some OS use ints for inodes. I've added cast to long int
7412 to get rid of the compiler warning.
7414 19. I had forgotten to lock out "/../" in configuration file names when
7415 ALT_CONFIG_PREFIX was set.
7417 20. Routers used for verification do not need to specify transports. However,
7418 if such a router generated a host list, and callout was configured, Exim
7419 crashed, because it could not find a port number from the (non-existent)
7420 transport. It now assumes port 25 in this circumstance.
7422 21. Added the -t option to exigrep.
7424 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7425 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7426 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7427 them is set, LOOKUP_LSEARCH is forced.
7429 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7431 24. Added sender and host information to the "rejected by local_scan()" log
7432 line; previously there was no indication of these.
7434 25. Added .include_if_exists.
7436 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7437 Linux. It turns out that not all file systems support this. Apparently some
7438 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7439 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7440 not supported on the file descriptor, is now ignored when Exim is trying to
7441 sync a directory. This applies only to Linux.
7443 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7445 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7446 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7449 29. Added support for maildirsize files from supplied patch (modified a bit).
7451 30. The use of :fail: followed by an empty string could lead Exim to respond to
7452 sender verification failures with (e.g.):
7454 550 Verification failed for <xxx>
7455 550 Sender verify failed
7457 where the first response line was missing the '-' that indicates it is not
7458 the final line of the response.
7460 31. The loop for finding the name of the user that called Exim had a hardwired
7461 limit of 10; it now uses the value of finduser_retries, which is used for
7462 all other user lookups.
7464 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7467 33. Exim was neglecting to zero errno before one call of strtol() when
7468 expanding a string and expecting an integer value. On some systems this
7469 resulted in spurious "integer overflow" errors. Also, it was casting the
7470 result into an int without checking.
7472 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7473 did not work. The code looks as if it has *never* worked, though it appears
7474 to have been documented since at least release 1.62. I have made it work.
7476 35. The "timeout_DNS" error in retry rules, also documented since at least
7477 1.62, also never worked. As it isn't clear exactly what this means, and
7478 clearly it isn't a major issue, I have abolished the feature by treating it
7479 as "timeout", and writing a warning to the main and panic logs.
7481 36. The display of retry rules for -brt wasn't always showing the error code
7484 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7485 timeout_connect_A, timeout_connect_MX.
7487 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7488 to the empty sender.
7490 39. The daemon was not analysing the content of -oX till after it had closed
7491 stderr and disconnected from the controlling terminal. This meant that any
7492 syntax errors were only noted on the panic log, and the return code from
7493 the command was 0. By re-arranging the code a little, I've made the
7494 decoding happen first, so such errors now appear on stderr, and the return
7495 code is 1. However, the actual setting up of the sockets still happens in
7496 the disconnected process, so errors there are still only recorded on the
7499 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7500 connections (as happens on some IP stacks) was logged at start up time as
7501 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7502 it from "IPv6 and IPv4", which means that two separate sockets are being
7505 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7506 says whether AF_INET or AF_INET6 was passed as an argument.
7508 42. Exiwhat output was messed up when time zones were included in log
7511 43. Exiwhat now gives more information about the daemon's listening ports,
7512 and whether -tls-on-connect was used.
7514 44. The "port" option of the smtp transport is now expanded.
7516 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7517 silently ignored. Now an error message is written to the main and panic
7520 46. There's a new ACL modifier called "logwrite" which writes to a log file
7521 as soon as it is encountered.
7523 47. Added $local_user_uid and $local_user_gid at routing time.
7525 48. Exim crashed when trying to verify a sender address that was being
7528 49. Exim was recognizing only a space character after ".include". It now also
7529 recognizes a tab character.
7531 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7532 extracting the relevant information from the specification. The man page no
7533 longer contains scrambled data for the -d option, and I've added a section
7534 at the front about calling Exim under different names.
7536 51. Added "extra_headers" argument to the "mail" command in filter files.
7538 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7541 53. Installed eximstats 1.29.
7543 54. Added transport_filter_timeout as a generic transport option.
7545 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7546 Cc: header lines. This was required by RFC 822, but it not required by RFC
7549 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7550 incoming messages that did not have them. Now it does so only if the
7551 message originates locally, that is, if there is no associated remote host
7552 address. When Resent- header lines are present, this applies to the Resent-
7553 lines rather than the non-Resent- lines.
7555 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7556 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7558 58. Messages for configuration errors now include the name of the main
7559 configuration file - useful now that there may be more than one file in a
7560 list (.included file names were always shown).
7562 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7563 for those rare installations that do not start the daemon as root or run it
7564 setuid root. I've cut out the call to initgroups() if the daemon is not
7567 60. The Exim user and group can now be bound into the binary as text strings
7568 that are looked up at the start of Exim's processing.
7570 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7572 62. Added $mailstore_basename variable.
7574 63. Installed patch to sieve.c from Michael Haardt.
7576 64. When Exim failed to open the panic log after failing to open the main log,
7577 the original message it was trying to log was written to stderr and debug
7578 output, but if they were not available (the usual case in production), it
7579 was lost. Now it is written to syslog before the two lines that record the
7580 failures to open the logs.
7582 65. Users' Exim filters run in subprocesses under the user's uid. It is
7583 possible for a "deliver" command or an alias in a "personal" command to
7584 provoke an address rewrite. If logging of address rewriting is configured,
7585 this fails because the process is not running as root or exim. There may be
7586 a better way of dealing with this, but for the moment (because 4.30 needs
7587 to be released), I have disabled address rewrite logging when running a
7588 filter in a non-root, non-exim process.
7594 1. The buildconfig auxiliary program wasn't quoting the value set for
7595 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7596 not defined. This bug was masked in 4.22 by the effect that was fixed in
7599 2. Some messages that were rejected after a message id was allocated were
7600 shown as "incomplete" by exigrep. It no longer does this for messages that
7601 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7603 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7604 have allow_domain_literals set, the ID did not get logged in the <= line.
7605 Domain literals are now always recognized in Message-ID: header lines.
7607 4. The first argument for a ${extract expansion item is the key name or field
7608 number. Leading and trailing spaces in this item were not being ignored,
7609 causing some misleading effects.
7611 5. When deliver_drop_privilege was set, single queue runner processes started
7612 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7613 same command in the process it spins off) were not dropping privilege.
7615 6. When the daemon running as "exim" started a queue runner, it always
7616 re-executed Exim in the spun-off process. This is a waste of effort when
7617 deliver_drop_privilege is set. The new process now just calls the
7618 queue-runner function directly.
7624 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7627 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7628 ignored. Though the use of -oP was forcing the writing of a pid file, it
7629 was always written to the default place.
7631 3. If the message "no IP address found for host xxxx" is generated during
7632 incoming verification, it is now followed by identification of the incoming
7633 connection (so you can more easily find what provoked it).
7635 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7637 5. Added some features to "harden" Exim a bit more against certain attacks:
7639 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7640 be put in Local/Makefile. This is like the never_users runtime option,
7641 but it cannot be overridden. The default setting is "root".
7643 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7644 prefix string with which any file named in a -C command line option
7647 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7648 is retained for -C and -D only if the caller of Exim is root. Without
7649 it, the exim user may also use -C and -D and retain privilege.
7651 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7652 command line option is disabled.
7654 6. Macro names set by the -D option must start with an upper case letter, just
7655 like macro names defined in the configuration file.
7657 7. Added "dereference=" facility to LDAP.
7659 8. Two instances of the typo "uknown" in the source files are fixed.
7661 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7662 the Configure-Makefile script screwed up while processing it.
7664 10. Incorporated PCRE 4.4.
7666 11. The SMTP synchronization check was not operating right at the start of an
7667 SMTP session. For example, it could not catch a HELO sent before the client
7668 waited for the greeting. There is now a check for outstanding input at the
7669 point when the greeting is written. Because of the duplex, asynchronous
7670 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7671 way, but not yet received, when the check is performed.
7673 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7674 on TCP/IP sockets, because this apparently causes some broken clients to
7677 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7678 unchanged) from the Cygwin maintainer.
7680 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7681 was supported instead of testing for mbx. Effectively a typo.
7683 15. The spa authenticator server code was not checking that the input it
7684 received was valid base64.
7686 16. The debug output line for the "set" modifier in ACLs was not showing the
7687 name of the variable that was being set.
7689 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7691 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7692 lookup on its own. Something else was needed to trigger the lookup. For
7693 example, a match in host_lookup or the need for a host name in a host list.
7694 Now, if $sender_host_name is referenced and the host name has not yet been
7695 looked up, a lookup is performed. If the lookup fails, the variable remains
7696 empty, and $host_lookup_failed is set to "1".
7698 19. Added "eqi" as a case-independent comparison operator.
7700 20. The saslauthd authentication condition could segfault if neither service
7701 nor realm was specified.
7703 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7704 error message that was logged was misleading, and incoming SMTP
7705 connections were dropped. The message is now more accurate, and temporary
7706 errors are given to SMTP connections.
7708 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7709 (except RSET) with a 421 error, until QUIT is received. However, it was
7710 failing to send a response to QUIT.
7712 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7713 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7714 ACL is now run after the helo_try_verify_hosts code.
7716 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7717 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7718 case-independent, so other case variants are also recognized). Apparently
7719 some systems use these upper case variants.
7721 25. If more than two messages were waiting for the same host, and a transport
7722 filter was specified for the transport, Exim sent two messages over the
7723 same TCP/IP connection, and then failed with "socket operation on non-
7724 socket" when it tried to send the third.
7726 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7728 27. The extern definition of crypt16() in expand.c was not being excluded when
7729 the OS had its own crypt16() function.
7731 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7732 as a preferred synonym for return_size_limit, both as an option and as an
7735 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7737 30. Changed the default configuration ACL to relax the local part checking rule
7738 for addresses that are not in any local domains. For these addresses,
7739 slashes and pipe symbols are allowed within local parts, but the sequence
7740 /../ is explicitly forbidden.
7742 31. SPA server authentication was not clearing the challenge buffer before
7745 32. log_message in a "warn" ACL statement was writing to the reject log as
7746 well as to the main log, which contradicts the documentation and doesn't
7747 seem right (because no rejection is happening). So I have stopped it.
7749 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7750 However, I am unable to do any testing of this.
7752 34. Fixed an infelicity in the appendfile transport. When checking directories
7753 for a mailbox, to see if any needed to be created, it was accidentally
7754 using path names with one or more superfluous leading slashes; tracing
7755 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7757 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7758 contents are added to the log line that is written for every discarded
7759 recipient. (Previously a log_message setting was ignored.)
7761 36. The ${quote: operator now quotes the string if it is empty.
7763 37. The install script runs exim in order to find its version number. If for
7764 some reason other than non-existence or emptiness, which it checks, it
7765 could not run './exim', it was installing it with an empty version number,
7766 i.e. as "exim-". This error state is now caught, and the installation is
7769 38. An argument was missing from the function that creates an error message
7770 when Exim fails to connect to the socket for saslauthd authentication.
7771 This could cause Exim to crash, or give a corrupted message.
7773 39. Added isip, isip4, and isip6 to ${if conditions.
7775 40. The ACL variables $acl_xx are now saved with the message, and can be
7776 accessed later in routers, transports, and filters.
7778 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7779 strings in the file are not string-expanded.
7781 42. If a MAIL command specified a SIZE value that was too large to fit into an
7782 int variable, the check against message_size_limit failed. Such values are
7783 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7784 day this will have to be increased, but I don't think I want to be around
7785 when emails are that large.
7792 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7793 iconv() is not standard in FreeBSD.
7795 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7796 IPv6 enabled. The observed symptom was a segmentation fault on return from
7797 the function os_common_find_running_interfaces() in src/os.c.
7799 3. In the check_special_case() function in daemon.c I had used "errno" as an
7800 argument name, which causes warnings on some systems. This was basically a
7801 typo, since it was named "eno" in the comments!
7803 4. The code that waits for the clock to tick (at a resolution of some fraction
7804 of a second) so as to ensure message-id uniqueness was always waiting for
7805 at least one whole tick, when it could have waited for less. [This is
7806 almost certainly not relevant at current processor speeds, where it is
7807 unlikely to ever wait at all. But we try to future-proof.]
7809 5. The function that sleeps for a time interval that includes fractions of a
7810 second contained a race. It did not block SIGALRM between setting the
7811 timer, and suspending (a couple of lines later). If the interval was short
7812 and the sigsuspend() was delayed until after it had expired, the suspension
7813 never ended. On busy systems this could lead to processes getting stuck for
7816 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7817 process, before it forks any delivery processes. The open lookup caching
7818 mechanism meant that the open file or database connection was passed into
7819 the delivery process. The problem was that delivery processes always tidy
7820 up cached lookup data. This could cause a problem for the next delivery
7821 process started by the queue runner, because the external queue runner
7822 process does not know about the closure. So the next delivery process
7823 still has data in the lookup cache. In the case of a file lookup, there was
7824 no problem because closing a file descriptor in a subprocess doesn't affect
7825 the parent. However, if the lookup was caching a connection to a database,
7826 the connection was closed, and the second delivery process was likely to
7827 see errors such as "PGSQL: query failed: server closed the connection
7828 unexpectedly". The problem has been fixed by closing all cached lookups
7829 in a queue runner before running a delivery process.
7831 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7832 seem to have the "const" qualifier which it has on other OS. I've
7835 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7836 *of the same type* (client or server) with the same public name that an
7837 error should be diagnosed.
7839 9. When Exim looked up a host name for an IP address, but failed to find the
7840 original IP address when looking up the host name (a safety check), it
7841 output the message "<ip address> does not match any IP for NULL", which was
7842 confusing, to say the least. The bug was that the host name should have
7843 appeared instead of "NULL".
7845 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7846 user that is built into the binary, and the -C or -D options is used, root
7847 privilege is dropped before the configuration file is read. In addition,
7848 logging is switched to stderr instead of the normal log files. If the
7849 configuration then re-defines the Exim user, the unprivileged environment
7850 is probably not what is expected, so Exim logs a panic warning message (but
7853 However, if deliver_drop_privilege is set, the unprivileged state may well
7854 be exactly what is intended, so the warning has been cut out in that case,
7855 and Exim is allowed to try to write to its normal log files.
7861 1. smtp_return_error_details was not giving details for temporary sender
7862 or receiver verification errors.
7864 2. Diagnose a configuration error if two authenticators have the same public
7867 3. Exim used not to create the message log file for a message until the first
7868 delivery attempt. This could be confusing when incoming messages were held
7869 for policy or load reasons. The message log file is now created at the time
7870 the message is received, and an initial "Received" line is written to it.
7872 4. The automatically generated man page for command line options had a minor
7873 bug that caused no ill effects; however, a more serious problem was that
7874 the procedure for building the man page automatically didn't always
7875 operate. Consequently, release 4.20 contains an out-of-date version. This
7876 shouldn't happen again.
7878 5. When building Exim with embedded Perl support, the script that builds the
7879 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7880 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7882 6. The freeze_tell option was not being used for messages that were frozen on
7883 arrival, either by an ACL or by local_scan().
7885 7. Added the smtp_incomplete_transaction log selector.
7887 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7888 was accepting AUTH without a new EHLO.
7890 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7891 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7894 10. Logging of TCP/IP connections (when configured) now happens in the main
7895 daemon process instead of the child process, so that the TCP/IP connection
7896 count is more accurate (but it can never be perfect).
7898 11. The use of "drop" in a nested ACL was not being handled correctly in the
7899 outer ACL. Now, if condition failure induced by the nested "drop" causes
7900 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7901 or "require"), the connection is dropped.
7903 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7904 that yield "discard" can now be used with an "accept" or a "discard" verb,
7905 but an error is generated for any others (because I can't see a useful way
7906 to define what should happen).
7908 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7909 are now processed in the same way as lines in the Exim configuration file.
7910 In particular, continuation lines are supported.
7912 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7914 15. Added -ti meaning -t -i.
7916 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7917 domains, and warn by logging if others are found.
7919 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7920 returned from GIFCONF when Exim is trying to find the list of interfaces on
7921 a host. The code in os.c has been modified to copy each ifreq to an aligned
7922 structure in all cases.
7924 Also, in some cases, the returned ifreq's were being copied to a 'struct
7925 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7926 means the last couple of bytes of an IPv6 address could be chopped if the
7927 ifreq contained only a normal sockaddr (14 bytes storage).
7929 18. Named domain lists were not supported in the hosts_treat_as_local option.
7930 An entry such as +xxxx was not recognized, and was treated as a literal
7933 19. Ensure that header lines added by a DATA ACL are included in the reject log
7934 if the ACL subsequently rejects the message.
7936 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7937 MD5 (which is deprecated).
7939 21. When testing a filter file using -bf, Exim was writing a message when it
7940 took the sender from a "From " line in the message, but it was not doing so
7941 when it took $return_path from a Return-Path: header line. It now does.
7943 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7944 with a valid header line field name (a series of printing characters
7945 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7947 23. Changed "disc" in the source to "disk" to conform to the documentation and
7948 the book and for uniformity.
7950 24. Ignore Sendmail's -Ooption=value command line item.
7952 25. When execve() failed while trying to run a command in a pipe transport,
7953 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7954 could be confused with a return value of 69 from the command itself. This
7955 has been changed to 127, the value the shell returns if it is asked to run
7956 a non-existent command. The wording for the related log line suggests a
7957 non-existent command as the problem.
7959 26. If received_header_text expands to an empty string, do not add a Received:
7960 header line to the message. (Well, it adds a token one on the spool, but
7961 marks it "old" so that it doesn't get used or transmitted.)
7963 27. Installed eximstats 1.28 (addition of -nt option).
7965 28. There was no check for failure on the call to getsockname() in the daemon
7966 code. This can fail if there is a shortage of resources on the system, with
7967 ENOMEM, for example. A temporary error is now given on failure.
7969 29. Contrary to the C standard, it seems that in some environments, the
7970 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7971 program. Exim now does this explicitly; it affects the formatting of
7972 timestamps using strftime().
7974 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7975 complaints. I've now initialized all the variables, to avoid this.
7977 32. Header lines added by a system filter were not being "seen" during
7978 transport-time rewrites.
7980 33. The info_callback() function passed to OpenSSL is set up with type void
7981 (*)(SSL *, int, int), as described somewhere. However, when calling the
7982 function (actually a macro) that sets it up, the type void(*)() is
7983 expected. I've put in a cast to prevent warnings from picky compilers.
7985 34. If a DNS black list lookup found a CNAME record, but there were no A
7986 records associated with the domain it pointed at, Exim crashed.
7988 35. If a DNS black list lookup returned more than one A record, Exim ignored
7989 all but the first. It now scans all returned addresses if a particular IP
7990 value is being sought. In this situation, the contents of the
7991 $dnslist_value variable are a list of all the addresses, separated by a
7994 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7995 to accept a host name and all its aliases if the forward lookup for any of
7996 them yielded the IP address of the incoming connection. Now it accepts only
7997 those names whose forward lookup yields the correct IP address. Any other
7998 names are discarded. This closes a loophole whereby a rogue DNS
7999 administrator could create reverse DNS records to break through a
8000 wildcarded host restriction in an ACL.
8002 37. If a user filter or a system filter that ran in a subprocess used any of
8003 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8004 the wrong values were passed to the pipe command ($thisaddress had the
8005 value of $0, $0 had the value of $1, etc). This bug was introduced by
8006 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8008 38. Improved the line breaking for long SMTP error messages from ACLs.
8009 Previously, if there was no break point between 40 and 75 characters, Exim
8010 left the rest of the message alone. Two changes have been made: (a) I've
8011 reduced the minimum length to 35 characters; (b) if it can't find a break
8012 point between 35 and 75 characters, it looks ahead and uses the first one
8013 that it finds. This may give the occasional overlong line, but at least the
8014 remaining text gets split now.
8016 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8017 file descriptors might be low, and that setting 1000 would always raise it.
8018 It turns out that in some environments, the limit is already over 1000 and
8019 that lowering it causes trouble. So now Exim takes care not to decrease it.
8021 40. When delivering a message, the value of $return_path is set to $sender_
8022 address at the start of routing (routers may change the value). By an
8023 oversight, this default was not being set up when an address was tested by
8024 -bt or -bv, which affected the outcome if any router or filter referred to
8027 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8028 log only when "message" or "log_message" are set. However, if one of the
8029 conditions was an address verification, or a call to a nested ACL, the
8030 messages generated by the underlying test were being passed through. This
8031 no longer happens. The underlying message is available in $acl_verify_
8032 message for both "message" and "log_message" expansions, so it can be
8033 passed through if needed.
8035 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8036 new expansion $bh_ to give the encoded byte string without charset
8037 translation. Translation happens only if iconv() is available; HAVE_ICONV
8038 indicates this at build time. HEADERS_CHARSET gives the charset to
8039 translate to; headers_charset can change it in the configuration, and
8040 "headers charset" can change it in an individual filter file.
8042 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8043 that creates RFC 2047 encoded "words" labels them as that charset instead
8044 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8045 expansion operator; (ii) when Exim creates a From: line for a local
8046 message; (iii) when a header line is rewritten to include a "phrase" part.
8048 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8049 buggy, causing it to skip the first lines of messages whose message ID
8050 ended in 'D'. This would not have bitten before Exim release 4.14, because
8051 message IDs were unlikely to end in 'D' before then. The effect was to have
8052 incorrect size information for certain domains.
8054 45. #include "config.h" was missing at the start of the crypt16.c module. This
8055 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8058 46. If there was a timeout during a "random" callout check, Exim treated it as
8059 a failure of the random address, and carried on sending RSET and the real
8060 address. If the delay was just some slowness somewhere, the response to the
8061 original RCPT would be taken as a response to RSET and so on, causing
8062 mayhem of various kinds.
8064 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8065 when I implemented it. It didn't allow for the fact that some option values
8066 may legitimately be negative (e.g. size_addition), and it didn't even do
8067 the right test for positive values.
8069 48. Domain names in DNS records are case-independent. Exim always looks them up
8070 in lower case. Some resolvers return domain names in exactly the case they
8071 appear in the zone file, that is, they may contain uppercase letters. Not
8072 all resolvers do this - some return always lower case. Exim was treating a
8073 change of case by a resolver as a change of domain, similar to a widening
8074 of a domain abbreviation. This triggered its re-routing code and so it was
8075 trying to route what was effectively the same domain again. This normally
8076 caused routing to fail (because the router wouldn't handle the domain
8077 twice). Now Exim checks for this case specially, and just changes the
8078 casing of the domain that it ultimately uses when it transmits the message
8081 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8084 50. If a filter generated a file delivery with a non-absolute name (possible if
8085 no home directory exists for the router), the forbid_file option was not
8088 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8089 the existing equality matching.
8091 52. Exim was using ints instead of ino_t variables in some places where it was
8092 dealing with inode numbers.
8094 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8095 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8096 called TMPDIR, and if it finds it is different, it changes its value.
8098 54. The smtp_printf() function is now made available to local_scan() so
8099 additional output lines can be written before returning. There is also an
8100 smtp_fflush() function to enable the detection of a dropped connection.
8101 The variables smtp_input and smtp_batched_input are exported to
8104 55. Changed the default runtime configuration: the message "Unknown user"
8105 has been removed from the ACL, and instead placed on the localuser router,
8106 using the cannot_route_message feature. This means that any verification
8107 failures that generate their own messages won't get overridden. Similarly,
8108 the "Unrouteable address" message that was in the ACL for unverifiable
8109 relay addresses has also been removed.
8111 56. Added hosts_avoid_esmtp to the smtp transport.
8113 57. The exicyclog script was not checking for the esoteric option
8114 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8115 will work only if exicyclog is run under the appropriate euid.
8117 58. Following a discussion on the list, the rules by which Exim recognises line
8118 endings on incoming messages have been changed. The -dropcr and drop_cr
8119 options are now no-ops, retained only for backwards compatibility. The
8120 following line terminators are recognized: LF CRLF CR. However, special
8121 processing applies to CR:
8123 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8124 nor a local message in the state where . is a terminator.
8126 (ii) If a bare CR is encountered in a header line, an extra space is added
8127 after the line terminator so as not to end the header. The reasoning
8128 behind this is that bare CRs in header lines are most likely either
8129 to be mistakes, or people trying to play silly games.
8131 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8132 was being incorrectly given as 18 bytes larger than it should have been.
8133 This is a VOB (very old bug).
8135 60. This may never have affected anything current, but just in case it has:
8136 When the local host is found other than at the start of a list of hosts,
8137 the local host, those with the same MX, and any that follow, are discarded.
8138 When the list in question was part of a longer list of hosts, the following
8139 hosts (not currently being processed) were also being discarded. This no
8140 longer happens. I'm not sure if this situation could ever has previously
8143 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8146 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8147 header. This is something that is recommended in a new Internet Draft, and
8148 is something that is documented as being done by Sendmail. There are two
8149 possible values. For messages generated by the autoreply transport, Exim
8152 Auto-Submitted: auto-replied
8154 whereas for all other generated messages (e.g. bounces) it adds
8156 Auto-Submitted: auto-generated
8158 63. The "personal" condition in filters now includes a test for the
8159 Auto-Submitted: header. If it contains the string "auto-" the message it
8160 not considered personal.
8162 64. Added rcpt_include_affixes as a generic transport option.
8164 65. Added queue_only_override (default true).
8166 66. Added the syslog_duplication option.
8168 67. If what should have been the first header line of a message consisted of
8169 a space followed by a colon, Exim was mis-interpreting it as a header line.
8170 It isn't of course - it is syntactically invalid and should therefore be
8171 treated as the start of the message body. The misbehaviour could have
8172 caused a number of strange effects, including loss of data in subsequent
8173 header lines, and spool format errors.
8175 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8176 client host had authenticated. This control can now be exercised by an ACL
8177 for more flexibility.
8179 69. By default, callouts do not happen when testing with -bh. There is now a
8180 variant, -bhc, which does actually run the callout code, including
8181 consulting and updating the callout cache.
8183 70. Added support for saslauthd authentication, courtesy of Alexander
8186 71. If statvfs() failed on the spool or log directories while checking their
8187 size for availability, Exim confusingly gave the error "space shortage".
8188 Furthermore, in debugging mode it crashed with a floating point exception.
8189 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8190 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8191 serious problem, Exim now writes to the main and panic logs when this
8192 happens, with details of the failure. It then refuses to accept the
8193 incoming message, giving the message "spool directory problem" or "log
8194 directory problem" with a 421 code for SMTP messages.
8196 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8197 0, 1, and 2 exist, because some OS complain for execs without them (see
8198 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8199 descriptors. However, the code omitted to check that the open succeeded,
8200 causing mysterious errors if for some reason the permissions on /dev/null
8201 got screwed. Now Exim writes a message to the main and panic logs, and
8202 bombs out if it can't open /dev/null.
8204 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8205 interact so that it is all more flexible. It is supposed to remain
8206 backwards compatible. Also added extra_local_interfaces.
8208 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8209 to bomb out with an assertion failure - to the client this appears as a
8210 connection drop. This problem occurs in the part of the code that was taken
8211 from the Samba project. Fortunately, the assertion is in a very simple
8212 function, so I have fixed this by reproducing the function inline in the
8213 one place where it is called, and arranging for authentication to fail
8214 instead of killing the process with assert().
8216 75. The SPA client code was not working when the server requested OEM rather
8217 than Unicode encoding.
8219 76. Added code to make require_files with a specific uid setting more usable in
8220 the case where statting the file as root fails - usually a non-root-mounted
8221 NFS file system. When this happens and the failure is EACCES, Exim now
8222 forks a subprocess and does the per-uid checking as the relevant uid.
8224 77. Added process_log_path.
8226 78. If log_file_path was not explicitly set, a setting of check_log_space or
8227 check_log_inodes was ignored.
8229 79. If a space check for the spool or log partitions fails, the incident is now
8230 logged. Of course, in the latter case the data may get lost...
8232 80. Added the %p formatting code to string_format() so that it can be used to
8233 print addresses in debug_print(). Adjusted all the address printing in the
8234 debugging in store.c to use %p rather than %d.
8236 81. There was a concern that a line of code in smtp_in.c could overflow a
8237 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8238 initially expressed, the concern was not well-founded, because trailing
8239 spaces are removed early. However, if the trailing spaces were followed by
8240 a NULL, they did not get removed, so the overflow was possible. Two fixes
8243 (a) I re-wrote the offending code in a cleaner fashion.
8244 (b) If an incoming SMTP command contains a NULL character, it is rejected
8247 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8248 runs initgroups(), so that if the Exim user is in any additional groups,
8249 they will be used during message reception.
8255 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob