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 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
43 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
44 Find and fix by Jasen Betts.
46 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
47 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
48 the systemwide configuration to override the Exim config.
50 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
51 API changes in libopendmarc.
53 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
54 pid 1, in the normal "background daemon" mode, having to drop process-
55 group leadership also lost track of needing to create listener sockets.
57 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
58 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
60 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
61 a capture group which obtained no text (eg. "(abc)*" matching zero
62 occurrences) could cause a segfault if the corresponding $<n> was
65 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
66 included a close-brace character (eg. it itself used an expansion) an
69 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
70 starting TLS. Previously it was after, meaning that attackers on such
71 ports had to be screened using the host_reject_connection main config
72 option. The new sequence aligns better with the STARTTLS behaviour, and
73 permits defences against crypto-processing load attacks, even though it
74 is strictly an incompatible change.
75 Also, avoid sending any SMTP fail response for either the connect ACL
76 or host_reject_connection, for TLS-on-connect ports.
78 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
79 Previously this was not permitted, but it makes reasonable sense.
80 While there, restore a restriction on using it from a connect ACL; given
81 the change JH/16 it could only return false (and before 4.91 was not
84 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
85 was exactly sized compared to the log buffer, a crash occurred with the
86 misleading message "bad memory reference; pool not found".
87 Found and traced by Jasen Betts.
89 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
90 dns_again_means_nonexist included an element causing a DNS lookup which
91 iteslf returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/20 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these looks.
109 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
110 after reception to before a subsequent reception. This should
111 mean slightly faster delivery, and also confirmation of reception
114 JH/02 Move from using the pcre library to pcre2. The former is no longer
115 being developed or supported (by the original developer).
117 JH/03 Constification work in the filters module required a major version
118 bump for the local-scan API. Specifically, the "headers_charset"
119 global which is visible via the API is now const and may therefore
120 not be modified by local-scan code.
122 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
123 sendfile() didi not account for the way the ClamAV driver code called it.
125 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
126 time check was being done for every character; replace that with one
129 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
130 sent was prefixed with a length byte.
132 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
133 RFC 5321. Previously Dovecot (at least) would log errors during
136 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
137 in 4.61 without a reason logged. Binaries will be bigger, which might
138 matter on diskspace-constrained systems, but debug is easier.
140 JH/09 Fix macro-definition during "-be" expansion testing. The move to
141 write-protected store for macros had not accounted for these runtime
142 additions; fix by removing this protection for "-be" mode.
144 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
145 handing out large-numbered file descriptors, violating the usual Unix
146 assumption (and required by Posix) that the lowest possible number will be
147 allocated by the kernel when a new one is needed. In the daemon, and any
148 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
149 are not useable for FD_SET() [and hence select()] and overwrite the stack.
150 Assorted crashes happen.
152 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
153 main-section options or in a connect ACL, the value from the first ever
154 connection was never replaced for subsequent connections. Found by
157 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
158 though only once PCRE2 was introduced: the memory accounting used under
159 debug offset allocations by an int, giving a hard trap in early startup.
160 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
162 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
163 with underbars is given. The write-protection of configuration introduced
164 in 4.95 trapped when normalisation was applied to an option not needing
167 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
169 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
170 failing to close the certificates directory, every hour or any time it
173 JH/16 Debugging initiated by an ACL control now continues through into routing
174 and transport processes. Previously debugging stopped any time Exim
175 re-execs, or for processing a queued message.
177 JH/17 The "expand" debug selector now gives more detail, specifically on the
178 result of expansion operators and items.
180 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
181 bad comparison between the option value and the name of the file to
182 be included was done, and a mismatch was wrongly identified.
183 4.88 to 4.95 are affected.
185 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
187 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
188 supplied as the db file-pair basename. Previously, if a directory
189 path was given, for example via the autoreply "once" option, the DB
190 file.pag and file.dir files would be created in that directory's
193 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
194 "taint" log_selector. These were previously deprecated.
196 JH/22 Fix static address-list lookups to properly return the matched item.
197 Previously only the domain part was returned.
199 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
200 the call into OpenSSL to send a TLS Close was being repeated; this
201 resulted in the library waiting for the peer's Close. If that was never
202 sent we waited forever. Fix by tracking send calls.
204 JH/24 The ${run} expansion item now expands its command string elements after
205 splitting. Previously it was before; the new ordering makes handling
206 zero-length arguments simpler. The old ordering can be obtained by
207 appending a new option "preexpand", after a comma, to the "run".
209 JH/25 Taint-check exec arguments for transport-initiated external processes.
210 Previously, tainted values could be used. This affects "pipe", "lmtp" and
211 "queryprogram" transport, transport-filter, and ETRN commands.
212 The ${run} expansion is also affected: in "preexpand" mode no part of
213 the command line may be tainted, in default mode the executable name
216 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
217 the the facility was not passed across execs, and only the first message
218 passed over a connection could use BDAT; any further ones using DATA.
220 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
221 uses $sending_ip_address and an interface is specified.
222 Previously any use of the local address in the EHLO name disabled
223 PIPECONNECT, the common case being to use the rDNS of it.
225 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
226 resumption. Previously verify failed because no certificate status is
227 passed on the wire for the restarted session. Fix by using the recorded
228 ocsp status of the stored session for the new connection.
230 JH/29 TLS resumption: the key for session lookup in the client now includes
231 more info that a server could potentially use in configuring a TLS
232 session, avoiding oferring mismatching sessions to such a server.
233 Previously only the server IP was used.
235 JH/30 Fix string_copyn() for limit greater than actual string length.
236 Previously the copied amount was the limit, which could result in a
237 overlapping memcpy for newly allocated destination soon after a
238 source string shorter than the limit. Found/investigated by KM.
240 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
241 close; it may be needed for a subsequent connection. This caused a
242 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
244 JH/32 Fix CHUNKING for a second message on a connection when the first was
245 rejected. Previously we did not reset the chunking-offered state, and
246 erroneously rejected the BDAT command. Investigation help from
249 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
250 an empty address. Previously the expansion returned an error.
252 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
253 proxy. Previously these were misparsed, leading to paniclog entries.
259 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
260 related applications. Previously an "H" was used where available info
261 says that "M" should be, so change to match.
263 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
264 as arguments, so an implementation trying to copy these into a local
265 buffer was taking a taint-enforcement trap. Fix by using dynamically
266 created buffers. Similar fix for radius expansion condition.
268 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
269 reasonable, eg. to count headers. Fix by using dynamically created
270 buffers rather than a local. Do similar fixes for ACL actions "dcc",
271 "log_reject_target", "malware" and "spam"; the arguments are expanded
272 so could be handling tainted values.
274 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
275 broken the (no-op) support for this sendmail command. Restore it
276 to doing nothing, silently, and returning good status.
278 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
279 record path was given (or the default used) without a leading directory
280 path, an error occurred on trying to open it. Use the transport's working
283 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
284 transport. Previously it was the name on the DNS A-record; use instead
285 the head of the CNAME chain leading there (if there is one). This seems
286 to align better with RFC 6125.
288 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
289 smtp_accept_max_per_host allocated resources which were not released
290 when the limit was exceeded. This eventually crashed the daemon. Fix
291 by adding a release action in that path.
293 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
294 expanded; previously using tainted values was rejected. Fix by using
295 dynamically-created buffers.
297 JH/09 Relax restrictions on ACL verify condition needing access to message
298 headers. Previously they were only permitted in data and non-smtp ACLs;
299 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
300 not_blind, header_sender and header_names_ascii verification.
302 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
303 Previously a macro used one argument twice; when called with the
304 argument as an expression having side-effects, incorrect operation
305 resulted. Use an inlineable function.
307 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
308 held open for a verify callout. Previously this wan not accounted for
309 and a corrupt onward SMTP conversation resulted.
311 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
312 Previously they were embedded naively in the constructed address; when
313 needed, strip the quoting and quote the entire local_part.
314 Also make the inbound_srs expansion condition handle quoting.
316 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
317 excluded, not matching the documentation.
319 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
320 was given for the sqlite_dbfile a trap resulted.
322 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
323 "name" argument resulted in a trap. There is no reason to disallow such;
324 this was a coding error.
326 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
327 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
328 spent suspended, ignoring the POSIX definition. Previously we assumed
329 it did and a constant offset from real time could be used as a correction.
330 Change to using the same clock source for the start-of-message and the
331 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
332 exists, just to get a clock slightly more aligned to reality.
334 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
335 RFC says it is optional some validators care. The missing char was not
336 intended but triggered by a line-wrap alignment. Discovery and fix by
337 Guillaume Outters, hacked on by JH.
339 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
340 name being quoted was tainted a trap would be taken. Fix by using
341 dynamically created buffers. The routine could have been called by a
342 rewrite with the "h" flag, by using the "-F" command-line option, or
343 by using a "name=" option on a control=submission ACL modifier.
345 JH/19 SPF: change the Authentication-Results expansion component to give
346 smtp.helo when the sender domain is empty. Previously it gave
349 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
350 not in 127.0.0.0/8 to help in spotting list domains taken over by a
351 domain-parking registrar.
353 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
354 Previously when a whitespace character was specified it was not inserted
355 after removing the newline.
357 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
358 the domain part of the recipient address. This overrides any tls_sni
359 option set, which was previously used.
361 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
364 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
365 is_tainted() had an off-by-one error in the overenthusiastic direction.
366 Find and fix by Gavan. Although NetBSD is not a supported platform for
367 4.94 this bug could affect other platforms.
369 PP/01 Fix default prime selection to be consistent.
370 One path used ike23 still, instead of exim.dev.20160529.3; now both
371 execution flows will use the same DH primes (currently
372 exim.dev.20160529.3).
374 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
375 option in smtp transport, to match the documentation. Previously
376 verification was not being done in some cases where it should have been.
378 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
379 than one server was defined and depending on the platform memory layout
380 details, an internal consistency trap could be hit while walking the list
383 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
384 files. The value is used by the authresults expansion item. Previously
385 if this was used in a router or transport, a crash could result.
387 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
388 have a sufficient incidence of EINTR returns from select that an
389 interaction with other operations done by the main daemon loop exposed
390 a bug in the error-handling. This was benign apart from the log
393 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
394 for consistency with delivered "=>" and failed "**" lines. While we're
395 there, handle PRX and TFO.
397 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
398 applied. This resulted, if any header-line rewrite rules were configured,
399 in a panic-log triggerable by sending a message with a long address in
400 a header. Fix by increasing the arbitrary limit to larger than a single
401 (dewrapped) 5322 header line maximum size.
403 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
404 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
405 RFC 6648 which deprecates X- options in protocols as a general practice.
406 Changeover between the implementations is handled by the mechanisms
409 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
410 a remote which uses callout/hold. Previously the local was queued.
412 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
415 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
416 log line. The asterisk is now consistently appended to the remote IP
417 (and port, if given), and will also be provided on defer and fail log
418 lines. Previously it could be placed on the local IP if that was being
419 logged, and was only provided on delivery lines.
421 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
423 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
424 authenticator client_send option. Previously the next char, after a pair
425 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
426 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
427 documentation. There is still no way to get a leading ^ immediately
428 after a NUL (ie. for the password of a PLAIN method authenticator.
430 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
431 files. For bad sizes read, delete the record and whine to paniclog.
433 JH/38 When logging an AUTH failure, as server, do not include sensitive
434 information. Previously, the credentials would be included if given
435 as part of the AUTH command line and an ACL denied authentication.
437 JH/39 Bug 2691: fix $local_part_data. When the matching list element
438 referred to a file, bad data was returned. This likely also affected
441 JH/40 The gsasl authenticator now supports caching of the salted password
442 generated by the client-side implementation. This required the addition
443 of a new variable: $auth4.
445 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
446 left undeleted; the attempt to re-create it then failed - resulting in
447 the usual "SIGHUP tp have daemon reload configuration" to not work.
448 This affected any platform not supporting "abstract" Unix-domain
449 sockets (i.e. not Linux).
451 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
452 recipients" response to RCPT in a later response, with a 250. The
453 previous coding assumed this would not happen, and under PIPELINING
454 would result in both lost and duplicate recipients for a message.
456 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
457 Previously the weighting was incorrectly applied. Similar fix for socks
458 proxies. Found and fixed by Heiko Schlichting.
460 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
461 not handle sub-lists included using the +namedlist syntax. While
462 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
463 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
466 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
467 Previously the permanent pool was used, so the sore could not be freed.
468 This meant a connection with many messages would use continually-growing
471 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
472 per-pool so as not to waste too much space. Previously a constant size
473 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
474 DOS attacks harder. The cost is wasted memory use in the larger blocks.
476 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
477 same space repeatedly, and a smaller process footprint.
479 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
480 for GnuTLS. Previously the name started "%s" which, while not a bug,
481 looked as if if might be one.
483 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
484 "continued connection") make the $proxy_* variables available. Previously
485 the information was not passed across the exec() call for subsequent
486 transport executions. This also mean that the log lines for the
487 messages can show the proxy information.
489 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
490 receive time. With modern systems the difference is significant.
491 The historical behaviour can be restored by disabling (a new) log_selector
492 "queue_time_exclusive".
494 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
495 content) were specifically tested for. Now, also cover expansions
496 resulting in ACL names and inline ACL content.
498 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
499 making it unusable in complex expressions.
501 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
502 fail with a taint-check on the spoolfile name, and leave the message
505 HS/01 Enforce absolute PID file path name.
507 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
509 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
511 PP/02 Bug 2643: Correct TLS DH constants.
512 A missing NUL termination in our code-generation tool had led to some
513 incorrect Diffie-Hellman constants in the Exim source.
514 Reported by kylon94, code-gen tool fix by Simon Arlott.
516 PP/03 Impose security length checks on various command-line options.
517 Fixes CVE-2020-SPRSS reported by Qualys.
519 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
520 better. Reported by Qualys.
522 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
523 providing a particularly obnoxious sender full name.
526 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
528 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
531 PP/08 Change default for recipients_max from unlimited to 50,000.
533 PP/09 Fix security issue with too many recipients on a message (to remove a
534 known security problem if someone does set recipients_max to unlimited,
535 or if local additions add to the recipient list).
536 Fixes CVE-2020-RCPTL reported by Qualys.
538 PP/10 Fix security issue in SMTP verb option parsing
539 Fixes CVE-2020-EXOPT reported by Qualys.
541 PP/11 Fix security issue in BDAT state confusion.
542 Ensure we reset known-good where we know we need to not be reading BDAT
543 data, as a general case fix, and move the places where we switch to BDAT
544 mode until after various protocol state checks.
545 Fixes CVE-2020-BDATA reported by Qualys.
547 HS/03 Die on "/../" in msglog file names
549 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
550 the Exim runtime user are allowed to create files.
552 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
555 QS/03 When reading the output from interpreted forward files we do not
556 pass the pipe between the parent and the interpreting process to
557 executed child processes (if any).
559 QS/04 Always die if requested from internal logging, even is logging is
562 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
563 the API; compilation noo longer completes with DMARC support included.
564 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
565 been reported on other platforms.
567 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
569 JH/56 Make the majority of info read from config files readonly, for defence-in-
570 depth against exploits. Suggestion by Qualys.
571 Not supported on Solaris 10.
573 JH/57 Fix control=fakreject for a custom message containing tainted data.
574 Previously this resulted in a log complaint, due to a re-expansion present
575 since fakereject was originally introduced.
577 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
578 modifier was given, a loop resulted.
580 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
581 pipelined SMTP command followed the BDAT LAST then it would be
582 incorrectly treated as part of the message body, causing a verification
585 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
586 require looser validation rules than those for 821-level addresses,
587 which only permit IP addresses.
593 JH/01 Avoid costly startup code when not strictly needed. This reduces time
594 for some exim process initialisations. It does mean that the logging
595 of TLS configuration problems is only done for the daemon startup.
597 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
599 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
600 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
603 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
604 DKIM signing is being done. Previously a transport_filter would always
605 disable CHUNKING, falling back to traditional DATA.
607 JH/05 Regard command-line recipients as tainted.
609 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
611 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
612 PAM library frees one of the arguments given to it, despite the
613 documentation. Therefore a plain malloc must be used.
615 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
616 on-stack buffers were used, resulting in a taint trap when DSN information
617 copied from a received message was written into the buffer.
619 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
620 the ordering of its ARC headers. This caused a crash.
622 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
623 a new record was being constructed with information from the peer, a trap
626 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
627 installation would get error messages from DMARC verify, when it hit the
628 nonexistent file indicated by the default. Distros wanting DMARC enabled
629 should both provide the file and set the option.
630 Also enforce no DMARC verification for command-line sourced messages.
632 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
633 could, depending on the platform, hang at the STARTTLS response.
635 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
636 message on a connection. Previously if one message had ARC headers and
637 the following one did not, a crash could result when adding an
638 Authentication-Results: header.
640 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
641 Exim main code and Exim-related utities. The introduction of taint
642 tracking also did many adjustments to string handling. Since then, eximon
643 frequently terminated with an assert failure.
645 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
646 check for 452 responses. This slightly helps the inefficieny of doing
647 a large alias-expansion into a recipient-limited target. The max_rcpt
648 transport option still applies (and at the current default, will override
649 the new feature). The check is done for either cause of synch, and forces
650 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
651 connection. The new facility is not tunable at this time.
653 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
654 library live data was being used, so the results became garbage. Make
655 copies while it is still usable.
657 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
658 on delivery deferred (==) and failed (**) lines (if a delivery was
659 attemtped). Previously it was only on completion (=>) lines.
661 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
662 for the expansion of the server_scram_iter and server_scram_salt options.
664 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
665 are now specifically given a NO_DATA response without hitting the system
666 resolver. The library goes on to do the now-standard TXT lookup.
667 Use of dnsdb lookups is not affected.
669 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
670 only retrieve the errormessage once. Previously two calls to dlerror()
671 were used, and the second one (for mainlog/paniclog) retrieved null
674 JH/20 Taint checking: disallow use of tainted data for
675 - the appendfile transport file and directory options
676 - the pipe transport command
677 - the autoreply transport file, log and once options
678 - file names used by the redirect router (including filter files)
680 - paths used by single-key lookups
681 Previously this was permitted.
683 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
684 adjusted the size of a major service buffer; this failed because the
685 buffer was in use at the time. Change to a compile-time increase in the
686 buffer size, when this authenticator is compiled into exim.
688 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
689 previous fast-mode was untenable in the face of glibs using mmap to
690 support larger malloc requests.
692 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
693 New values supported, if defined on system where compiled:
694 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
695 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
697 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
698 running a limited number of proceses in parallel, a benefit is gained. The
699 amount varies with the platform hardware and load. The use of the option
700 queue_run_in_order means we cannot do this, as ordering becomes
703 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
704 had introduced a string-copy (for ensuring NUL-termination) which was not
705 appropriate for that case, which can include embedded NUL bytes in the
706 block of data. Investigation showed the copy to actually be needless, the
707 data being length-specified.
709 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
710 done during a receiving connection, and both used TLS, global info was
711 used rather than per-connection info for tracking the state of data
712 queued for transmission. This could result in a connection hang.
714 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
715 Previously, when delivering serveral messages down a single connection
716 only the first would provide a SIZE. This was due to the size information
717 not being properly tracked.
719 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
720 TAI (at 37 seconds currently), pretend to be in UTC for time-related
721 expansion and logging. Previously, spurious values such as a future
722 minute could be seen.
724 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
725 it could crash from a null-deref. This could also affect the
726 ${addresses: } operator and ${readsock } item.
728 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
729 message following a mime one, the variable was not reset.
731 JH/30 When an pipelined-connect fails at the first response, assume incorrect
732 cached capability (perhaps the peer reneged?) and immediately retry in
735 JH/31 Fix spurious detection of timeout while writing to transport filter.
737 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
738 an attempt to copy the string was made before checking it.
740 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
741 taint of the lookup key was maintained; we now regard the presence in the
742 filesystem as sufficient validation.
744 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
745 argument is supplied.
747 JH/35 The dsearch lookup now requires that the directory is an absolute path.
748 Previously this was not checked, and nonempty relative paths made an
749 access under Exim's current working directory.
751 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
752 Previously no event was raised.
754 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
755 parameter supplied by the sender MAIL FROM command. Previously it was
756 ignored, and only the check_spool_space option value for the required
759 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
760 the size of the signing public-key. Previously it was instead giving
761 the size of the signature hash.
763 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
764 the default. See the (new) dkim_verify_min_keysizes option.
766 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
767 and an ACL use a lookup for checking either the local_part or domain,
768 stale data could be accessed. Ensure that variable references are
769 dropped between messages.
771 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
772 by the client was not checked as pointing within response data before
773 being used. A malicious client could thus cause an out-of-bounds read and
774 possibly gain authentication. Fix by adding the check.
776 JH/42 Internationalisation: change the default for downconversion in the smtp
777 transport to be "if needed". Previously it was "as previously set" for
778 the message, which usually meant "if needed" for message-submission but
779 "no" for everything else. However, MTAs have been seen using SMTPUTF8
780 even when the envelope addresses did not need it, resulting in forwarding
781 failures to non-supporting MTAs. A downconvert in such cases will be
782 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
783 transport. The change does mean that addresses needing conversion will
784 be converted when previously a delivery failure would occur.
786 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
787 response was received it would be used unchecked in a fail-DSN, violating
788 standards on line-length limits. Truncate if needed.
790 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
791 doesn't work. (Additionally add a new main config option to configure the
798 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
799 side, to decode a TLS 1.3 packet capture.
801 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
802 Previously the default library behaviour applied, sending two, each in
805 JH/03 Debug output for ACL now gives the config file name and line number for
808 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
810 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
812 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
813 buffer overrun for (non-chunking) other transports.
815 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
816 TLS1.3, means that a server rejecting a client certificate is not visible
817 to the client until the first read of encrypted data (typically the
818 response to EHLO). Add detection for that case and treat it as a failed
819 TLS connection attempt, so that the normal retry-in-clear can work (if
820 suitably configured).
822 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
823 and/or domain. Found and fixed by Jason Betts.
825 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
826 configuration). If a CNAME target was not a wellformed name pattern, a
829 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
830 the OS reports them interleaved with other addresses.
832 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
833 used both for input and for a verify callout, both encrypted, SMTP
834 responses being sent by the server could be lost. This resulted in
835 dropped connections and sometimes bounces generated by a peer sending
838 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
839 string. Previously it was possible to cause undefined behaviour in a
840 library routine (usually a crash). Found by "zerons".
842 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
845 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
846 API was removed, so update to use the newer ones.
848 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
849 any timeout set, is taking a long time. Previously we would hang on to a
850 rotated logfile "forever" if the input was arriving with long gaps
851 (a previous attempt to fix addressed lack, for a long time, of initial
854 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
855 shared (NFS) environment. The length of the tempfile name is now
856 4 + 16 ("hdr.$message_exim_id") which might break on file
857 systems which restrict the file name length to lower values.
860 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
861 shared (NFS) environment.
863 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
864 did for all versions <4.90). Notably -M, -m, --invert, -I may be
867 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
868 on some platforms for bit 31.
870 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
871 to changes apparently associated with TLS1.3 handling some of the APIs
872 previously used were either nonfunctional or inappropriate. Strings
873 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
874 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
875 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
876 This affects log line X= elements, the $tls_{in,out}_cipher variables,
877 and the use of specific cipher names in the encrypted= ACL condition.
879 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
881 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
882 verification result was not updated unless hosts_require_ocsp applied.
884 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
885 queue_list_requires_admin set to false, non-admin users were denied the
888 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
889 directory-of-certs mode. Previously they were advertised despite the
892 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
893 A single TCP connection by a client will now hold a TLS connection open
894 for multiple message deliveries, by default. Previously the default was to
897 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
898 default. If built with the facility, DANE will be used. The facility
899 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
901 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
902 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
903 must be defined and you must still, unless you define DISABLE_TLS, manage
904 the the include-dir and library-file requirements that go with that
905 choice. Non-TLS builds are still supported.
907 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
910 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
911 default. If the platform supports and has the facility enabled, it will
912 be requested on all coneections.
914 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
915 controlled by the build-time option SUPPORT_PIPE_CONNECT.
917 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
919 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
920 success-DSN messages. Previously the From: header was always the default
921 one for these; the option was ignored.
923 JH/28 Fix the timeout on smtp response to apply to the whole response.
924 Previously it was reset for every read, so a teergrubing peer sending
925 single bytes within the time limit could extend the connection for a
926 long time. Credit to Qualsys Security Advisory Team for the discovery.
928 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
929 delivery address, which leaked information of the results of local
930 forwarding. Change to the original envelope recipient address, per
933 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
934 requested. Previously not bounce was generated and a log entry of
935 error ignored was made.
937 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
939 JH/32 Introduce a general tainting mechanism for values read from the input
940 channel, and values derived from them. Refuse to expand any tainted
941 values, to catch one form of exploit.
943 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
944 was unused and the unexpanded text used for the test. Found and
945 fixed by Ruben Jenster.
947 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
948 an attempt to use a TLS library read routine dereffed a nul pointer,
951 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
952 them in our smtp response.
954 JH/36 Have the general router option retry_use_local_part default to true when
955 any of the restrictive preconditions are set (to anything). Previously it
956 was only for check_local user. The change removes one item of manual
957 configuration which is required for proper retries when a remote router
958 handles a subset of addresses for a domain.
960 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
961 link count into consideration.
963 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
964 caused the extension of big_buffer, the following lines were ignored.
966 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
967 accordance with RFC 2308. Previously there was no expiry, so a longlived
968 receive process (eg. due to ACL delays) versus a short SOA value could
971 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
973 JH/39 Promote DMARC support to mainline.
975 JH/40 Bug 2452: Add a References: header to DSNs.
977 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
978 parameters. The relevant library call is documented as "Deprecated: This
979 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
980 3.6.0, DH parameters are negotiated following RFC7919."
982 HS/06 Change the default of dnssec_request_domains to "*"
984 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
985 carried on and emitted a BDAT command, even when PIPELINING was not
988 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
989 buffer was used for the filename, resulting in a trap when tainted
990 arguments (eg. $domain) were used.
992 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
993 recommended to avoid a possible server-load attack. The feature can be
994 re-enabled via the openssl_options main cofiguration option.
996 JH/45 local_scan API: documented the current smtp_printf() call. This changed
997 for version 4.90 - adding a "more data" boolean to the arguments.
998 Bumped the ABI version number also, this having been missed previously;
999 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1000 respect of usage of smtp_printf() by either local_scan code or libraries
1001 accessed via the ${dlfunc } expansion item. Both will need coding
1002 adjustment for any calls to smtp_printf() to match the new function
1003 signature; a FALSE value for the new argument is always safe.
1005 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1006 the file-offset (which the Linux syscall does, and exim expects); this
1007 resulted in an indefinite loop.
1009 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1010 to do ARC verification. The Authentication-Results: header line added
1011 by the configuration then had no ARC item.
1013 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1014 an incorrect mode was used for reading the result, resulting in it being
1021 JH/01 Remove code calling the customisable local_scan function, unless a new
1022 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1024 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1025 non-signal-safe functions being used.
1027 JH/03 Bug 2269: When presented with a received message having a stupidly large
1028 number of DKIM-Signature headers, disable DKIM verification to avoid
1029 a resource-consumption attack. The limit is set at twenty.
1031 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1032 report of oldest_pass in ${authres } in consequence, and separate out
1033 some descriptions of reasons for verification fail.
1035 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1036 files in the spool were present and unlocked. A queue-runner could spot
1037 them, resulting in a duplicate delivery. Fix that by doing the unlock
1038 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1039 add more error-checking on spoolfile handling while that code is being
1042 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1043 No known attacks, no CVE, this is defensive hardening.
1045 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1046 a queue-runner could start a delivery while other operations were ongoing.
1047 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1048 Found and investigated by Tim Stewart. Fix by using the open message data
1049 file handle rather than opening another, and not locally closing it (which
1050 releases a lock) for that case, while creating the temporary .eml format
1051 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1053 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1054 $sender_verify_failure/$recipient_verify_failure to "random".
1056 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1059 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1060 Previously this would segfault.
1062 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1065 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1066 like zero, since the resolver should be doing this for us, But we need one
1067 as a CNAME but no MX presence gets the CNAME returned; we need to check
1068 that doesn't point to an MX to declare it "no MX returned" rather than
1069 "error, loop". A new main option is added so the older capability of
1070 following some limited number of chain links is maintained.
1072 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1074 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1075 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1076 now-deprecated earlier definitions used only specified the range up to TLS
1077 1.2 (in the older-version library docs).
1079 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1081 JH/15 Rework TLS client-side context management. Stop using a global, and
1082 explicitly pass a context around. This enables future use of TLS for
1083 connections to service-daemons (eg. malware scanning) while a client smtp
1084 connection is using TLS; with cutthrough connections this is quite likely.
1086 JH/16 Fix ARC verification to do AS checks in reverse order.
1088 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1090 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1091 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1092 Previously the "utf8" would be re-prepended for every additional message.
1094 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1095 Previously thery were accepted, resulting in issues when attempting to
1096 forward messages to a non-supporting MTA.
1098 PP/02 Let -n work with printing macros too, not just options.
1100 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1101 one parent address was copied, and bogus data was used at delivery-logging
1102 time. Either a crash (after delivery) or bogus log data could result.
1103 Discovery and analysis by Tim Stewart.
1105 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1106 Previously if the string ended mid-character, we did not insert the
1107 promised '?' replacement.
1109 PP/04 Documentation: current string operators work on bytes, not codepoints.
1111 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1112 should pack well giving a smaller memory footprint so better caching and
1113 therefore performance. Group the declarations where this can't be done so
1114 that the byte-sized flag variables are not interspersed among pointer
1115 variables, giving a better chance of good packing by the compiler.
1117 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1118 non-null, to avoid issues with sites running BATV. Previously reports were
1119 sent with an empty envelope sender so looked like bounces.
1121 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1122 The ignore_error flag wasn't being returned from the filter subprocess so
1123 was not set for later routers. Investigation and fix by Matthias Kurz.
1125 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1126 and a msg:complete for the whole, when a message is manually removed using
1127 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1129 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1130 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1131 hope that is portable enough.
1133 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1134 requiring a known-CA anchor certificate; make it now rely entirely on the
1135 TLSA as an anchor. Checking the name on the leaf cert against the name
1136 on the A-record for the host is still done for TA (but not for EE mode).
1138 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1139 included in delivery lines for non-proxied connections, when compiled with
1140 SUPPORT_SOCKS and running with proxy logging enabled.
1142 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1143 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1144 move the existing event to fire before the normal logging of message
1145 failure so that custom logging is bracketed by normal logging.
1147 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1148 msg:fail:internal event. Developement by Matthias Kurz.
1150 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1151 far too small for todays use of crypto signatures stored there. Go all
1152 the way to the max DNS message size of 64kB, even though this might be
1153 overmuch for IOT constrained device use.
1155 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1156 copy a string over itself. The library routine is documented as not
1157 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1159 JH/32 For main options check_spool_space and check_inode_space, where the
1160 platform supports 64b integers, support more than the previous 2^31 kB
1161 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1162 the previous G, M, k.
1164 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1165 $authenticated_fail_id variable on authentication failure. Previously
1168 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1169 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1170 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1171 GNUTLS_SEC_PARAM_MEDIUM.
1173 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1174 side. Previously we would continue as if no SNI had been received.
1176 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1177 "<" character, which should be a list-separator specification, we walked
1178 off past the nul-terimation.
1180 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1181 causes) even when the retry time is not yet met. Previously they were
1182 not, meaning that when (say) an account was over-quota and temp-rejecting,
1183 and multiple senders' messages were queued, only one sender would get
1184 notified on each configured delay_warning cycle.
1186 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1188 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1189 cipher-suites, an error can be left on the stack even for a succeeding
1190 accept; this results in impossible error messages when a later operation
1193 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1194 return error codes indicating retry. Under TLS1.3 this becomes required.
1196 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1197 it only wrote the new authenticators, resulting in a lack of tracking of
1198 peer changes of ESMTP extensions until the next cache flush.
1200 JH/41 Fix the loop reading a message header line to check for integer overflow,
1201 and more-often against header_maxsize. Previously a crafted message could
1202 induce a crash of the recive process; now the message is cleanly rejected.
1204 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1205 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1211 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1212 When redis_servers is set to a list of > 1 element, and the Redis servers
1213 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1214 case of MOVED into a DEFER case instead, thus moving the query onto the
1215 next server in the list. For a cluster of N elements, all N servers must
1216 be defined in redis_servers.
1218 GF/02 Catch and remove uninitialized value warning in exiqsumm
1219 Check for existence of @ARGV before looking at $ARGV[0]
1221 JH/01 Replace the store_release() internal interface with store_newblock(),
1222 which internalises the check required to safely use the old one, plus
1223 the allocate and data copy operations duplicated in both (!) of the
1224 extant use locations.
1226 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1227 modifier. This matches the restriction on the commandline.
1229 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1230 Previously only the last row was returned.
1232 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1233 we assumed that tags in the header were well-formed, and parsed the
1234 element content after inspecting only the first char of the tag.
1235 Assumptions at that stage could crash the receive process on malformed
1238 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1239 While running the DKIM ACL we operate on the Permanent memory pool so that
1240 variables created with "set" persist to the DATA ACL. Also (at any time)
1241 DNS lookups that fail create cache records using the Permanent pool. But
1242 expansions release any allocations made on the current pool - so a dnsdb
1243 lookup expansion done in the DKIM ACL releases the memory used for the
1244 DNS negative-cache, and bad things result. Solution is to switch to the
1245 Main pool for expansions.
1246 While we're in that code, add checks on the DNS cache during store_reset,
1247 active in the testsuite.
1248 Problem spotted, and debugging aided, by Wolfgang Breyha.
1250 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1251 When none of the hosts presented to a transport match an already-open
1252 connection, close it and proceed with the list. Previously we would
1253 queue the message. Spotted by Lena with Yahoo, probably involving
1256 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1257 Previously a spurious "250 OK id=" response was appended to the proper
1260 JH/08 The "support for" informational output now, which built with Content
1261 Scanning support, has a line for the malware scanner interfaces compiled
1262 in. Interface can be individually included or not at build time.
1264 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1265 by the template makefile "src/EDITME". The "STREAM" support for an older
1266 ClamAV interface method is removed.
1268 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1269 rows affected is given instead).
1271 JH/11 The runtime Berkeley DB library version is now additionally output by
1272 "exim -d -bV". Previously only the compile-time version was shown.
1274 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1275 SMTP connection. Previously, when one had more recipients than the
1276 first, an abortive onward connection was made. Move to full support for
1277 multiple onward connections in sequence, handling cutthrough connection
1278 for all multi-message initiating connections.
1280 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1281 routers. Previously, a multi-recipient message would fail to match the
1282 onward-connection opened for the first recipient, and cause its closure.
1284 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1285 a timeout on read on a GnuTLS initiating connection, resulting in the
1286 initiating connection being dropped. This mattered most when the callout
1287 was marked defer_ok. Fix to keep the two timeout-detection methods
1290 JH/15 Relax results from ACL control request to enable cutthrough, in
1291 unsupported situations, from error to silently (except under debug)
1292 ignoring. This covers use with PRDR, frozen messages, queue-only and
1295 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1297 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1298 metadata, resulting in a crash in free().
1300 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1301 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1302 Broken also in d185889f4, with init system revamp.
1304 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1305 Previously we abruptly closed the connection after reading a malware-
1306 found indication; now we go on to read the "scan ok" response line,
1309 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1310 ACL. Previously, a crash would result.
1312 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1313 macro text after a replacement (previously it was only once per line) and
1314 by skipping builtin macros when searching for an uppercase lead character.
1316 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1317 for the build is renamed.
1319 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1320 was allocated for every new TLS startup, meaning one per message. Fix
1321 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1323 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1324 reported the original. Fix to report (as far as possible) the ACL
1325 result replacing the original.
1327 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1328 OpenSSL. Certificate information is loaded for every new TLS startup,
1329 and the resources needed to be freed.
1331 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1333 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1336 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1337 DATA response info to the (existing) per-recipient response info for
1338 the "C=" log element. It can have useful tracking info from the
1339 destination system. Patch from Simon Arlott.
1341 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1342 length value. Previously this would segfault.
1344 HS/02 Support Avast multiline protoocol, this allows passing flags to
1345 newer versions of the scanner.
1347 JH/28 Ensure that variables possibly set during message acceptance are marked
1348 dead before release of memory in the daemon loop. This stops complaints
1349 about them when the debug_store option is enabled. Discovered specifically
1350 for sender_rate_period, but applies to a whole set of variables.
1351 Do the same for the queue-runner and queue-list loops, for variables set
1352 from spool message files. Do the same for the SMTP per-message loop, for
1353 certain variables indirectly set in ACL operations.
1355 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1356 as a multi-recipient message from a mailinglist manager). The coding had
1357 an arbitrary cutoff number of characters while checking for more input;
1358 enforced by writing a NUL into the buffer. This corrupted long / fast
1359 input. The problem was exposed more widely when more pipelineing of SMTP
1360 responses was introduced, and one Exim system was feeding another.
1361 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1362 receiving system, and refused recipients seen by the sending system
1363 (propating to people being dropped from mailing lists).
1364 Discovered and pinpointed by David Carter.
1366 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1367 replaced by the ${authresults } expansion.
1369 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1371 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1372 allows proper process termination in container environments.
1374 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1375 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1377 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1378 and "err_perm", deprecated since 4.83 when the RFC-defined words
1379 "temperror" and "permerror" were introduced.
1381 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1382 transport-filters or DKIM-signing. The restriction was lost in the
1383 consolidation of verify-callout and delivery SMTP handling.
1384 Extend the restriction to also cover ARC-signing.
1386 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1387 in defer=pass mode supply a 450 to the initiator. Previously the message
1390 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1391 tls_require_ciphers is used as before.
1393 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1394 "pass_unscanned". Only tmpfails from the scanner are written to
1395 the paniclog, as they may require admin intervention (permission
1396 denied, license issues). Other scanner errors (like decompression
1397 bombs) do not cause a paniclog entry.
1399 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1400 Previously it was possible to log spurious information in receive log
1403 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1404 triggered odd behaviour from Outlook Express clients.
1406 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1409 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1410 since the IETF WG has not yet settled on that versus the original
1411 "bare" representation.
1413 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1414 Previously the millisecond value corrupted the output.
1415 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1416 corrupted the output.
1422 JH/01 Rework error string handling in TLS interface so that the caller in
1423 more cases is responsible for logging. This permits library-sourced
1424 string to be attached to addresses during delivery, and collapses
1425 pairs of long lines into single ones.
1427 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1428 during configuration. Wildcards are allowed and expanded.
1430 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1431 This permits better logging.
1433 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1434 do not close it down and have the child transport start it up again on
1435 the passed-on TCP connection. Instead, proxy the child (and any
1436 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1437 affected: the continued delivery log lines do not have any DNSSEC, TLS
1438 Certificate or OCSP information. TLS cipher information is still logged.
1440 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1441 identical IP addresses on different listening ports. Will also affect
1444 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1445 add noisy ifdef guards to special-case this sillyness.
1446 Patch from Bernd Kuhls.
1448 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1449 than 255 are no longer allowed.
1451 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1452 Disable the session-cache too, which might reduce our load. Since we
1453 currrectly use a new context for every connection, both as server and
1454 client, there is no benefit for these.
1455 GnuTLS appears to not support tickets server-side by default (we don't
1456 call gnutls_session_ticket_enable_server()) but client side is enabled
1457 by default on recent versions (3.1.3 +) unless the PFS priority string
1460 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1461 <https://reproducible-builds.org/specs/source-date-epoch/>.
1463 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1464 the check for any unsuccessful recipients did not notice the limit, and
1465 erroneously found still-pending ones.
1467 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1468 MSG_MORE. Only in-clear (not on TLS connections).
1470 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1471 transport-filter is involved we can buffer the headers in memory for
1472 creating the signature, and read the spool data file once for the
1473 signature and again for transmission.
1475 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1476 4.77 as the kernel support then wasn't solid, having issues in 64bit
1477 mode. Now, it's been long enough. Add support for FreeBSD also.
1479 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1480 case where the routing stage had gathered several addresses to send to
1481 a host before calling the transport for the first, we previously failed
1482 to close down TLS in the old transport process before passing the TCP
1483 connection to the new process. The new one sent a STARTTLS command
1484 which naturally failed, giving a failed delivery and bloating the retry
1485 database. Investigation and fix prototype from Wolfgang Breyha.
1487 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1488 false-negatives in the check that the sender had not preempted a response
1489 or prompt from Exim (running as a server), due to that code's lack of
1490 awareness of the SMTP input buffering.
1492 PP/04 Add commandline_checks_require_admin option.
1493 Exim drops privileges sanely, various checks such as -be aren't a
1494 security problem, as long as you trust local users with access to their
1495 own account. When invoked by services which pass untrusted data to
1496 Exim, this might be an issue. Set this option in main configuration
1497 AND make fixes to the calling application, such as using `--` to stop
1500 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1501 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1502 responses to those, into a single TLS record each way (this usually means
1503 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1506 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1507 affects you only if you're dancing at the edge of the param size limits.
1508 If you are, and this message makes sense to you, then: raise the
1509 configured limit or use OpenSSL 1.1. Nothing we can do for older
1512 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1513 cmdline element to get the documented default one. Previously it was
1516 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1517 are unsuable for non-message-specific reasons - eg. connection timeout,
1518 banner-time rejection.
1520 JH/16 Fix logging of delivery remote port, when specified by router, under
1523 PP/06 Repair manualroute's ability to take options in any order, even if one
1524 is the name of a transport.
1527 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1529 JH/17 Change the list-building routines interface to use the expanding-string
1530 triplet model, for better allocation and copying behaviour.
1532 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1533 Previously it was constructed the first time a possibly-matching string
1534 was met in the configuration file input during startup; now it is done
1537 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1538 compatible one, to avoid the (poorly documented) possibility of a config
1539 file in the working directory redirecting the DB files, possibly correpting
1540 some existing file. CVE-2017-10140 assigned for BDB.
1542 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1543 cache-hot. Previously, although the result was properly cached, the
1544 initial verify call returned a defer.
1546 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1547 the main verify for receipient in uncached-mode.
1549 JH/22 Retire historical build files to an "unsupported" subdir. These are
1550 defined as "ones for which we have no current evidence of testing".
1552 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1553 if present. Previously it was ignored.
1555 JH/24 Start using specified-initialisers in C structure init coding. This is
1556 a C99 feature (it's 2017, so now considered safe).
1558 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1559 if was a fixed-sized field and bitmask ops via macros; it is now more
1562 PP/07 GitHub PR 56: Apply MariaDB build fix.
1563 Patch provided by Jaroslav Škarvada.
1565 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1566 during Coverity cleanups [4.87 JH/47]
1567 Diagnosis and fix provided by Michael Fischer v. Mollard.
1569 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1570 the right size to place the terminating semicolon on its own folded
1571 line, the header hash was calculated to an incorrect value thanks to
1572 the (relaxed) space the fold became.
1574 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1575 and confused the parent.
1577 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1578 which could crash as a result. This could lead to undeliverable messages.
1580 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1583 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1584 matters on fast-turnover and PID-randomising systems, which were getting
1585 out-of-order delivery.
1587 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1588 a possibly-overlapping copy. The symptom was that "Remote host closed
1589 connection in response to HELO" was logged instead of the actual 4xx
1592 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1593 Previously only that bufferd was discarded, resulting in SYMTP command
1596 JH/32 DKIM: when a message has multiple signatures matching an identity given
1597 in dkim_verify_signers, run the dkim acl once for each. Previously only
1598 one run was done. Bug 2189.
1600 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1601 "panic the current process" to "deliberately defer". The panic log is
1602 still written with the problem list name; the mail and reject logs now
1603 get a temp-reject line for the message that was being handled, saying
1604 something like "domains check lookup or other defer". The SMTP 451
1605 message is still "Temporary local problem".
1607 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1608 A crafted sequence of BDAT commands could result in in-use memory beeing
1609 freed. CVE-2017-16943.
1611 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1612 from SMTP input. Previously it was always done; now only done for DATA
1613 and not BDAT commands. CVE-2017-16944.
1615 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1616 to the message (such as an overlong header line). Previously this was
1617 not done and we did not exit BDAT mode. Followon from the previous item
1618 though a different problem.
1624 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1625 than -2003 did; needs libidn2 in addition to libidn.
1627 JH/02 The path option on a pipe transport is now expanded before use.
1629 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1630 Patch provided by "Björn", documentation fix added too.
1632 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1633 missing a wire-to-host endian conversion.
1635 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1636 close after a BDAT command line could be taken as a following command,
1637 giving a synch failure. Fix by only checking for synch immediately
1638 before acknowledging the chunk.
1640 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1641 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1642 macro. Patches provided by Josh Soref.
1644 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1645 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1646 by IANA as a service extension. However, John Klensin suggests that we
1649 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1650 the dkim code may be unix-mode line endings rather than smtp wire-format
1651 CRLF, so prepend a CR to any bare LF.
1653 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1654 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1656 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1657 the first were themselves being wrongly included in the feed into dkim
1658 processing; with most chunk sizes in use this resulted in an incorrect
1659 body hash calculated value.
1661 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1662 DKIM signature block, for verification. Although advised against by
1663 standards it is specifically not ruled illegal.
1665 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1667 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1668 missing a body hash (the bh= tag).
1670 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1671 It seems that HAProxy sends the Proxy Protocol information in clear and
1672 only then does a TLS startup, so do the same.
1674 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1675 TCP connections (such as for Spamd) unless the daemon successfully set
1676 Fast Open mode on its listening sockets. This fixes breakage seen on
1677 too-old kernels or those not configured for Fast Open, at the cost of
1678 requiring both directions being enabled for TFO, and TFO never being used
1679 by non-daemon-related Exim processes.
1681 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1682 endings, at least on the first header line. Try to canonify any that get
1683 past that check, despite the cost.
1685 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1686 now limited to an arbitrary five deep, while parsing addresses with the
1687 strip_excess_angle_brackets option enabled.
1689 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1690 instead leave the unprompted TLS handshake in socket buffer for the
1691 TLS library to consume.
1693 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1695 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1697 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1698 region in one operation, for speed, and this leaves assigned pointers
1699 dangling. Add checks run only under the testsuite which checks all
1700 variables at a store-reset and panics on a dangling pointer; add code
1701 explicitly nulling out all the variables discovered. Fixes one known
1702 bug: a transport crash, where a dangling pointer for $sending_ip_address
1703 originally assigned in a verify callout, is re-used.
1705 PP/06 Drop '.' from @INC in various Perl scripts.
1707 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1709 PP/08 Reduce a number of compilation warnings under clang; building with
1710 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1711 should be warning-free.
1713 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1715 HS/01 Fix portability problems introduced by PP/08 for platforms where
1716 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1718 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1719 chunk. This allows us to accept broken chunked messages. We need a more
1720 general solution here.
1722 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1723 already-broken messages in the queue.
1725 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1727 JH/19 Fix reference counting bug in routing-generated-address tracking.
1733 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1734 supports it and a size is available (ie. the sending peer gave us one).
1736 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1737 years of being deprecated). The replacements are the ACLs
1738 acl_smtp_mime and acl_not_smtp_mime.
1740 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1741 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1742 or even an in-clear connection were permitted. Now, if the host lookup
1743 was dnssec and dane was requested then the host is only used if the
1744 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1745 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1746 if one fails this test.
1747 This means that a poorly-configured remote DNS will make it incommunicado;
1748 but it protects against a DNS-interception attack on it.
1750 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1751 non-noisy when a race steals the message being considered.
1753 JH/05 If main configuration option tls_certificate is unset, generate a
1754 self-signed certificate for inbound TLS connections.
1756 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1757 in rewrites and routers.
1759 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1760 and logged a warning sing 4.83; now they are a configuration file error.
1762 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1763 (lacking @domain). Apply the same qualification processing as RCPT.
1765 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1767 JH/10 Support ${sha256:} applied to a string (as well as the previous
1770 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1771 a cutthrough deliver is pending, as we always want to make a connection.
1772 This also avoids re-routing the message when later placing the cutthrough
1773 connection after a verify cache hit.
1774 Do not update it with the verify result either.
1776 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1777 when routing results in more than one destination address.
1779 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1780 signing (which inhibits the cutthrough capability). Previously only
1781 the presence of an option was tested; now an expansion evaluating as
1782 empty is permissible (obviously it should depend only on data available
1783 when the cutthrough connection is made).
1785 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1786 the relevant preceding SMTP command did not note the pipelining mode.
1788 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1789 Previously they were not counted.
1791 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1792 as one having no matching records. Previously we deferred the message
1793 that needed the lookup.
1795 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1796 distinguished as "(=".
1798 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1799 for missing MX records. Previously it only worked for missing A records.
1801 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1803 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1804 after the data-go-ahead and data-ack. Patch from Jason Betts.
1806 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1807 even for a "none" policy. Patch from Tony Meyer.
1809 JH/22 Fix continued use of a connection for further deliveries. If a port was
1810 specified by a router, it must also match for the delivery to be
1813 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1814 When one of the recipients of a message was unsuitable for the connection
1815 (has no matching addresses), we lost track of needing to mark it
1816 deferred. As a result mail would be lost.
1818 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1820 JH/25 Decoding ACL controls is now done using a binary search; the source code
1821 takes up less space and should be simpler to maintain. Merge the ACL
1822 condition decode tables also, with similar effect.
1824 JH/26 Fix problem with one_time used on a redirect router which returned the
1825 parent address unchanged. A retry would see the parent address marked as
1826 delivered, so not attempt the (identical) child. As a result mail would
1829 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1830 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1831 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1834 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1835 Main config options check_{log,spool}_{inodes,space} are now
1836 100 inodes, 10MB unless set otherwise in the configuration.
1838 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1839 Previously it only applied to the main-section connection policy
1842 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1844 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1845 by me. Added RFC7919 DH primes as an alternative.
1847 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1848 are not in the system include path.
1850 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1851 GnuTLS, when a session startup failed (eg because the client disconnected)
1852 Exim did stdio operations after fclose. This was exposed by a recent
1853 change which nulled out the file handle after the fclose.
1855 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1856 signed directly by the cert-signing cert, rather than an intermediate
1857 OCSP-signing cert. This is the model used by LetsEncrypt.
1859 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1861 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1862 an incoming connection.
1864 HS/02 Bug 1802: Do not half-close the connection after sending a request
1867 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1868 fallback to "prime256v1".
1870 JH/34 SECURITY: Use proper copy of DATA command in error message.
1871 Could leak key material. Remotely exploitable. CVE-2016-9963.
1877 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1878 and 3.4.4 - once the server is enabled to respond to an OCSP request
1879 it does even when not requested, resulting in a stapling non-aware
1880 client dropping the TLS connection.
1882 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1883 support variable-length bit vectors. No functional change.
1885 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1886 The I= interface field on outgoing lines is now after the H= remote
1887 host field, same as incoming lines. There is a separate
1888 outgoing_interface log selector which allows you to disable the
1891 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1892 If not running log_selector +smtp_connection the mainlog would be held
1893 open indefinitely after a "too many connections" event, including to a
1894 deleted file after a log rotate. Leave the per net connection logging
1895 leaving it open for efficiency as that will be quickly detected by the
1896 check on the next write.
1898 HS/01 Bug 1671: Fix post transport crash.
1899 Processing the wait-<transport> messages could crash the delivery
1900 process if the message IDs didn't exist for some reason. When
1901 using 'split_spool_directory=yes' the construction of the spool
1902 file name failed already, exposing the same netto behaviour.
1904 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1905 mime_regex ACL conditions.
1907 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1908 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1909 to HELO, local diagnostic string.
1911 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1912 log line to debug. Even when configured with a tls authenticator many
1913 client connections are expected to not authenticate in this way, so
1914 an authenticate fail is not an error.
1916 HS/02 Add the Exim version string to the process info. This way exiwhat
1917 gives some more detail about the running daemon.
1919 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1920 matter for fast-change records such as DNSBLs.
1922 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1923 retry record, even if constant. There may be multiple transports with
1924 different interface settings and the retry behaviour needs to be kept
1927 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1929 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1931 JH/10 Bug 840: fix log_defer_output option of pipe transport
1933 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1934 hosts_randomize. This matters a lot when combined with mua_wrapper.
1936 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1937 ${quote_pgsql:<string>} operator.
1939 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1941 JH/14 Change header file name for freeradius-client. Relevant if compiling
1942 with Radius support; from the Gentoo tree and checked under Fedora.
1944 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1946 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1947 option transport_filter, meaning no filtering.
1949 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1951 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1952 defaults to "*" (all hosts). The variable is now available when not built
1953 with TLS, default unset, mainly to enable keeping the testsuite sane.
1954 If a server certificate is not supplied (via tls_certificate) an error is
1955 logged, and clients will find TLS connections fail on startup. Presumably
1956 they will retry in-clear.
1957 Packagers of Exim are strongly encouraged to create a server certificate
1958 at installation time.
1960 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1961 with the $config_file variable.
1963 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1964 in transport context, after the attempt, and per-recipient. The latter type
1965 is per host attempted. The event data is the error message, and the errno
1966 information encodes the lookup type (A vs. MX) used for the (first) host,
1967 and the trailing two digits of the smtp 4xx response.
1969 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1970 to write to mainlog (or rejectlog, paniclog) in the window between file
1971 creation and permissions/ownership being changed. Particularly affects
1972 installations where exicyclog is run as root, rather than exim user;
1973 result is that the running daemon panics and dies.
1975 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1977 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1978 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1979 "pri" and "weight". Note that the previous implicit priority given by the
1980 list order is no longer honoured.
1982 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1983 for DKIM processing.
1985 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1986 by defining SUPPORT_SOCKS.
1988 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1989 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1990 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1991 variables are renamed to proxy_{local,external}_{address,port}.
1993 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1994 for a build by defining SUPPORT_I18N
1996 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1997 of the query string, and make ${quote_redis:} do that quoting.
1999 JH/29 Move Events support from Experimental to mainline, enabled by default
2000 and removable for a build by defining DISABLE_EVENT.
2002 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2004 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2005 cached by the daemon.
2007 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2008 by defining LOOKUP_REDIS. The libhiredis library is required.
2010 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2011 keys are given for lookup.
2013 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2014 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2015 only supported when built with TLS support. The PolarSSL SHA routines
2016 are still used when the TLS library is too old for convenient support.
2018 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2019 openssl_options), for security. OpenSSL forces this from version 1.1.0
2020 server-side so match that on older versions.
2022 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2023 allocation for $value could be released as the expansion processing
2024 concluded, but leaving the global pointer active for it.
2026 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2027 and to use the domains and local_parts ACL conditions.
2029 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2030 incorrectly not doubled on cutthrough transmission, hence seen as a
2031 body-termination at the receiving system - resulting in truncated mails.
2032 Commonly the sender saw a TCP-level error, and retransmitted the message
2033 via the normal store-and-forward channel. This could result in duplicates
2034 received - but deduplicating mailstores were liable to retain only the
2035 initial truncated version.
2037 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2039 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2041 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2042 we're in there, support oversigning also; bug 1309.
2044 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2046 HS/04 Add support for keep_environment and add_environment options.
2048 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2049 either intentional arithmetic overflow during PRNG, or testing config-
2052 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2053 delivery resulted in actual delivery. Cancel cutthrough before DATA
2056 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2057 rejects a recipient: pass the reject to the originator.
2059 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2060 Many were false-positives and ignorable, but it's worth fixing the
2063 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2064 for the new environment-manipulation done at startup. Move the routines
2065 from being local to tls.c to being global via the os.c file.
2067 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2068 an extract embedded as result-arg for a map, the first arg for extract
2069 is unavailable so we cannot tell if this is a numbered or keyed
2070 extraction. Accept either.
2076 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2079 JH/02 The smtp transport option "multi_domain" is now expanded.
2081 JH/03 The smtp transport now requests PRDR by default, if the server offers
2084 JH/04 Certificate name checking on server certificates, when exim is a client,
2085 is now done by default. The transport option tls_verify_cert_hostnames
2086 can be used to disable this per-host. The build option
2087 EXPERIMENTAL_CERTNAMES is withdrawn.
2089 JH/05 The value of the tls_verify_certificates smtp transport and main options
2090 default to the word "system" to access the system default CA bundle.
2091 For GnuTLS, only version 3.0.20 or later.
2093 JH/06 Verification of the server certificate for a TLS connection is now tried
2094 (but not required) by default. The verification status is now logged by
2095 default, for both outbound TLS and client-certificate supplying inbound
2098 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2101 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2102 Status Notification (bounce) messages are now MIME format per RFC 3464.
2103 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2104 under the control of the dsn_advertise_hosts option, and routers may
2105 have a dsn_lasthop option.
2107 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2108 default, modifiable by a malware= option. The list separator for
2109 the options can now be changed in the usual way. Bug 68.
2111 JH/10 The smtp_receive_timeout main option is now expanded before use.
2113 JH/11 The incoming_interface log option now also enables logging of the
2114 local interface on delivery outgoing connections.
2116 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2117 if the interface and destination host and port all match.
2119 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2122 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2123 Patch from Andrew Lewis.
2125 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2126 now supports optional time-restrictions, weighting, and priority
2127 modifiers per server. Patch originally by <rommer@active.by>.
2129 JH/16 The spamd_address main option now supports a mixed list of local
2130 and remote servers. Remote servers can be IPv6 addresses, and
2131 specify a port-range.
2133 JH/17 Bug 68: The spamd_address main option now supports an optional
2134 timeout value per server.
2136 JH/18 Bug 1581: Router and transport options headers_add/remove can
2137 now have the list separator specified.
2139 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2142 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2145 JH/21 Support for the A6 type of dns record is withdrawn.
2147 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2148 rather than the verbs used.
2150 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2151 from 255 to 1024 chars.
2153 JH/24 Verification callouts now attempt to use TLS by default.
2155 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2156 are generic router options now. The defaults didn't change.
2158 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2159 Original patch from Alexander Shikoff, worked over by JH.
2161 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2162 files if ZCAT_COMMAND is not executable.
2164 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2166 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2168 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2169 Normally benign, it bites when the pair was led to by a CNAME;
2170 modern usage is to not canonicalize the domain to a CNAME target
2171 (and we were inconsistent anyway for A-only vs AAAA+A).
2173 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2175 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2176 when evaluating $sender_host_dnssec.
2178 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2179 $sender_helo_dnssec variable.
2181 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2183 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2185 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2187 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2188 documented as working, but never had. Support all but $spam_report.
2190 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2191 added for tls authenticator.
2193 HS/03 Add perl_taintmode main config option
2199 TL/01 When running the test suite, the README says that variables such as
2200 no_msglog_check are global and can be placed anywhere in a specific
2201 test's script, however it was observed that placement needed to be near
2202 the beginning for it to behave that way. Changed the runtest perl
2203 script to read through the entire script once to detect and set these
2204 variables, reset to the beginning of the script, and then run through
2205 the script parsing/test process like normal.
2207 TL/02 The BSD's have an arc4random API. One of the functions to induce
2208 adding randomness was arc4random_stir(), but it has been removed in
2209 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2210 function when detected.
2212 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2213 cause callback expansion.
2215 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2216 syntax errors in an expansion can be treated as a string instead of
2217 logging or causing an error, due to the internal use of bool_lax
2218 instead of bool when processing it.
2220 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2221 server certificates when making smtp deliveries.
2223 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2225 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2227 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2229 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2230 Merged patch from Sebastian Wiedenroth.
2232 JH/05 Fix results-pipe from transport process. Several recipients, combined
2233 with certificate use, exposed issues where response data items split
2234 over buffer boundaries were not parsed properly. This eventually
2235 resulted in duplicates being sent. This issue only became common enough
2236 to notice due to the introduction of connection certificate information,
2237 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2239 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2240 size buffer was used, resulting in syntax errors when an expansion
2243 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2244 version 3.3.6 or later.
2246 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2247 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2248 both become "event_action", the variables become $event_name, $event_data
2249 and $event_defer_errno. There is a new variable $verify_mode, usable in
2250 routers, transports and related events. The tls:cert event is now also
2251 raised for inbound connections, if the main configuration event_action
2254 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2255 early OCSP support, but no stapling (appears to be less than 1.0.0).
2257 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2258 server certificate names available under the smtp transport option
2259 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2262 JH/10 Time-related extraction expansions from certificates now use the main
2263 option "timezone" setting for output formatting, and are consistent
2264 between OpenSSL and GnuTLS compilations. Bug 1541.
2266 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2267 encoded parameter in the incoming message. Bug 1558.
2269 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2270 include certificate info, eximon was claiming there were spoolfile
2273 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2275 JH/14 Log delivery-related information more consistently, using the sequence
2276 "H=<name> [<ip>]" wherever possible.
2278 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2279 are problematic for Debian distribution, omit them from the release
2282 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2284 JH/16 Fix string representation of time values on 64bit time_t architectures.
2287 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2288 output list separator was used.
2293 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2294 checkers that were complaining about end of non-void function with no
2297 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2298 This was a regression introduced in 4.83 by another bugfix.
2300 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2302 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2303 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2309 TF/01 Correctly close the server side of TLS when forking for delivery.
2311 When a message was received over SMTP with TLS, Exim failed to clear up
2312 the incoming connection properly after forking off the child process to
2313 deliver the message. In some situations the subsequent outgoing
2314 delivery connection happened to have the same fd number as the incoming
2315 connection previously had. Exim would try to use TLS and fail, logging
2316 a "Bad file descriptor" error.
2318 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2319 utilities have not been installed.
2321 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2322 temporary space as the ACL may create new global variables.
2324 TL/01 LDAP support uses per connection or global context settings, depending
2325 upon the detected version of the libraries at build time.
2327 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2328 to extract and use the src ip:port in logging and expansions as if it
2329 were a direct connection from the outside internet. PPv2 support was
2330 updated based on HAProxy spec change in May 2014.
2332 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2334 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2335 Properly escape header and check for NULL return.
2337 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2340 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2342 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2343 characters in header names, implemented as a verify condition.
2344 Contributed by Michael Fischer v. Mollard.
2346 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2347 results permerror and temperror. Previous values are deprecated but
2348 still accepted. In a future release, err_perm and err_temp will be
2349 completely removed, which will be a backward incompatibility if the
2350 ACL tests for either of these two old results. Patch contributed by
2351 user bes-internal on the mailing list.
2353 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2355 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2356 selectors, in both main and reject logs.
2358 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2361 JH/07 Add malware type "sock" for talking to simple daemon.
2363 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2365 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2366 routers/transports under cutthrough routing.
2368 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2369 numbers. Touch up "bool" conditional to keep the same definition.
2371 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2373 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2375 JH/12 Expand items in router/transport headers_add or headers_remove lists
2376 individually rather than the list as a whole. Bug 1452.
2378 Required for reasonable handling of multiple headers_ options when
2379 they may be empty; requires that headers_remove items with embedded
2380 colons must have them doubled (or the list-separator changed).
2382 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2383 view the policy declared in the DMARC record. Currently, $dmarc_status
2384 is a combined value of both the record presence and the result of the
2387 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2389 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2390 dnslookup router and the smtp transport (applying to the forward
2393 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2394 of ldap servers used for a specific lookup. Patch provided by Heiko
2397 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2398 New variable $lookup_dnssec_authenticated for observability.
2400 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2401 Patch submitted by Lars Timman.
2403 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2405 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2406 Requires trusted mode and valid format message id, aborts otherwise.
2407 Patch contributed by Heiko Schlichting.
2409 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2410 certextract with support for various fields. Bug 1358.
2412 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2413 is requested by default, modifiable by smtp transport option
2416 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2417 operate on certificate variables to give certificate fingerprints
2418 Also new ${sha256:cert_variable}.
2420 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2422 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2423 Christian Aistleitner.
2425 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2427 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2428 file. Patch from Wolfgang Breyha.
2430 JH/25 Expand the coverage of the delivery $host and $host_address to
2431 client authenticators run in verify callout. Bug 1476.
2433 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2434 align with daemon_smtp_ports. Bug 72.
2436 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2437 support and error reporting did not work properly.
2439 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2440 and is readable. Patch from Andrew Colin Kissa.
2442 TL/14 Enhance documentation of ${run expansion and how it parses the
2443 commandline after expansion, particularly in the case when an
2444 unquoted variable expansion results in an empty value.
2446 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2448 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2449 by renaming away from it. Addresses GH issue 15, reported by
2452 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2454 TL/15 SECURITY: prevent double expansion in math comparison functions
2455 (can expand unsanitized data). Not remotely exploitable.
2462 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2464 PP/02 Make -n do something, by making it not do something.
2465 When combined with -bP, the name of an option is not output.
2467 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2470 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2471 $sender_host_name and config options to manage this, and basic check
2474 PP/05 DSCP support for outbound connections and control modifier for inbound.
2476 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2477 (Only plugin which currently uses this is kerberos4, which nobody should
2478 be using, but we should make it available and other future plugins might
2479 conceivably use it, even though it would break NAT; stuff *should* be
2480 using channel bindings instead).
2482 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2483 name; added for Sendmail compatibility; requires admin caller.
2484 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2485 just ignore it); requires trusted caller.
2486 Also parse but ignore: -Ac -Am -X<logfile>
2489 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2491 TL/02 Add +smtp_confirmation as a default logging option.
2493 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2494 Patch by Magnus Holmgren from 2007-02-20.
2496 TL/04 Bugzilla 1281 - Spec typo.
2497 Bugzilla 1283 - Spec typo.
2498 Bugzilla 1290 - Spec grammar fixes.
2500 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2502 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2504 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2505 mailing list by by Dmitry Isaikin.
2507 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2509 JH/02 Support "G" suffix to numbers in ${if comparisons.
2511 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2513 NM/01 Bugzilla 1197 - Spec typo
2514 Bugzilla 1196 - Spec examples corrections
2516 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2518 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2519 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2522 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2523 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2525 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2526 "acl {{name}{arg}...}", and optional args on acl condition
2529 JH/05 Permit multiple router/transport headers_add/remove lines.
2531 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2533 JH/07 Avoid using a waiting database for a single-message-only transport.
2534 Performance patch from Paul Fisher. Bugzilla 1262.
2536 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2539 JH/09 Add $headers_added variable, with content from use of ACL modifier
2540 add_header (but not yet added to the message). Bugzilla 199.
2542 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2543 Pulled from Bugzilla 817 by Wolfgang Breyha.
2545 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2547 (nb: this is the same fix as in Exim 4.80.1)
2549 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2552 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2555 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2557 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2558 advertise SMTP AUTH mechanism to us, instead of a generic
2559 protocol violation error. Also, make Exim more robust to bad
2560 data from the Dovecot auth socket.
2562 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2564 When a queue runner is handling a message, Exim first routes the
2565 recipient addresses, during which it prunes them based on the retry
2566 hints database. After that it attempts to deliver the message to
2567 any remaining recipients. It then updates the hints database using
2570 So if a recipient address works intermittently, it can get repeatedly
2571 deferred at routing time. The retry hints record remains fresh so the
2572 address never reaches the final cutoff time.
2574 This is a fairly common occurrence when a user is bumping up against
2575 their storage quota. Exim had some logic in its local delivery code
2576 to deal with this. However it did not apply to per-recipient defers
2577 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2579 This change adds a proper retry rule check during routing so that the
2580 final cutoff time is checked against the message's age. We only do
2581 this check if there is an address retry record and there is not a
2582 domain retry record; this implies that previous attempts to handle
2583 the address had the retry_use_local_parts option turned on. We use
2584 this as an approximation for the destination being like a local
2585 delivery, as in LMTP.
2587 I suspect this new check makes the old local delivery cutoff check
2588 redundant, but I have not verified this so I left the code in place.
2590 TF/02 Correct gecos expansion when From: is a prefix of the username.
2592 Test 0254 submits a message to Exim with the header
2596 When I ran the test suite under the user fanf2, Exim expanded
2597 the header to contain my full name, whereas it should have added
2598 a Resent-Sender: header. It erroneously treats any prefix of the
2599 username as equal to the username.
2601 This change corrects that bug.
2603 GF/01 DCC debug and logging tidyup
2604 Error conditions log to paniclog rather than rejectlog.
2605 Debug lines prefixed by "DCC: " to remove any ambiguity.
2607 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2609 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2610 Bug spotted by Jeremy Harris; was flawed since initial commit.
2611 Would have resulted in OCSP responses post-SNI triggering an Exim
2612 NULL dereference and crash.
2614 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2616 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2617 Bug detection, analysis and fix by Samuel Thibault.
2618 Bugzilla 1331, Debian bug #698092.
2620 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2622 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2623 Server implementation by Todd Lyons, client by JH.
2624 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2625 config variable "prdr_enable" controls whether the server
2626 advertises the facility. If the client requests PRDR a new
2627 acl_data_smtp_prdr ACL is called once for each recipient, after
2628 the body content is received and before the acl_smtp_data ACL.
2629 The client is controlled by both of: a hosts_try_prdr option
2630 on the smtp transport, and the server advertisement.
2631 Default client logging of deliveries and rejections involving
2632 PRDR are flagged with the string "PRDR".
2634 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2635 fclose(). Diagnosis by Todd Lyons.
2637 PP/17 Update configure.default to handle IPv6 localhost better.
2638 Patch by Alain Williams (plus minor tweaks).
2641 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2642 This is now consistent with GnuTLS, and is now documented: the
2643 previous undocumented portable approach to treating the option as
2644 unset was to force an expansion failure. That still works, and
2645 an empty string is now equivalent.
2647 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2648 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2649 not performing validation itself.
2651 PP/20 Added force_command boolean option to pipe transport.
2652 Patch from Nick Koston, of cPanel Inc.
2654 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2657 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2659 PP/21 Fix eximon continuous updating with timestamped log-files.
2660 Broken in a format-string cleanup in 4.80, missed when I repaired the
2661 other false fix of the same issue.
2662 Report and fix from Heiko Schlichting.
2665 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2666 Report from Prashanth Katuri.
2668 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2669 It's SecureTransport, so affects any MacOS clients which use the
2670 system-integrated TLS libraries, including email clients.
2672 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2673 using a MIME ACL for non-SMTP local injection.
2674 Report and assistance in diagnosis by Warren Baker.
2676 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2678 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2680 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2681 last id that failed so it may be referenced in subsequent ACL's.
2683 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2686 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2687 ldap library initialization, allowing self-signed CA's to be
2688 used. Also properly sets require_cert option later in code by
2689 using NULL (global ldap config) instead of ldap handle (per
2690 session). Bug diagnosis and testing by alxgomz.
2692 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2693 the src/util/ subdirectory.
2695 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2696 renamed to Transport Post Delivery Action by Jeremy Harris, as
2699 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2700 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2701 redis_servers = needs to be configured which will be used by the redis
2702 lookup. Patch from Warren Baker, of The Packet Hub.
2704 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2706 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2707 hostname or reverse DNS when processing a host list. Used suggestions
2708 from multiple comments on this bug.
2710 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2712 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2713 Missed a few lines, added it to make the runtest require no keyboard
2716 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2717 contains upper case chars. Make router use caseful_local_part.
2719 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2720 support when GnuTLS has been built with p11-kit.
2726 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2728 This, or similar/improved, will also be change PP/11 of 4.82.
2734 PP/01 Handle short writes when writing local log-files.
2735 In practice, only affects FreeBSD (8 onwards).
2736 Bugzilla 1053, with thanks to Dmitry Isaikin.
2738 NM/01 Bugzilla 949 - Documentation tweak
2740 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2743 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2745 PP/02 Implemented gsasl authenticator.
2747 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2749 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2750 `pkg-config foo` for cflags/libs.
2752 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2753 with rest of GSASL and with heimdal_gssapi.
2755 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2756 `pkg-config foo` for cflags/libs for the TLS implementation.
2758 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2759 properties get this fed in as external SSF. A number of robustness
2760 and debugging improvements to the cyrus_sasl authenticator.
2762 PP/08 cyrus_sasl server now expands the server_realm option.
2764 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2765 Patch by Jeremy Harris.
2767 PP/10 Added dbmjz lookup type.
2769 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2771 PP/12 MAIL args handles TAB as well as SP, for better interop with
2772 non-compliant senders.
2773 Analysis and variant patch by Todd Lyons.
2775 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2776 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2777 Patch from Dirk Mueller <dmueller@suse.com>
2779 PP/13 tls_peerdn now print-escaped for spool files.
2780 Observed some $tls_peerdn in wild which contained \n, which resulted
2781 in spool file corruption.
2783 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2784 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2785 or write after TLS renegotiation, which otherwise led to messages
2788 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2789 as a tracking header (ie: a signed header comes before the signature).
2790 Patch from Wolfgang Breyha.
2792 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2793 comma-sep list; embedded commas doubled.
2795 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2797 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2799 Report and patch from Dmitry Banschikov.
2801 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2802 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2803 failures. We appear to now support TLS1.1+ with Exim.
2805 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2806 lets Exim select keys and certificates based upon TLS SNI from client.
2807 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2808 before an outbound SMTP session. New log_selector, +tls_sni.
2810 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2811 NULL dereference. Report and patch from Alun Jones.
2813 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2814 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2815 Not seeing resolver debug output on NetBSD, but suspect this is a
2816 resolver implementation change.
2818 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2819 Left warnings. Added "eximon gdb" invocation mode.
2821 PP/21 Defaulting "accept_8bitmime" to true, not false.
2823 PP/22 Added -bw for inetd wait mode support.
2825 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2826 locate the relevant includes and libraries. Made this the default.
2828 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2829 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2831 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2832 This may cause build issues on older platforms.
2834 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2835 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2836 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2837 Added SNI support via GnuTLS too.
2838 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2840 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2842 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2843 Applied second patch from Janne, implementing suggestion to default
2844 multiple-strings-in-record handling to match SPF spec.
2846 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2848 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2849 read-only, out of scope).
2850 Patch from Wolfgang Breyha, report from Stuart Northfield.
2852 PP/29 Fix three issues highlighted by clang analyser static analysis.
2853 Only crash-plausible issue would require the Cambridge-specific
2854 iplookup router and a misconfiguration.
2855 Report from Marcin Mirosław.
2857 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2859 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2860 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2861 As part of this, removing so much warning spew let me fix some minor
2862 real issues in debug logging.
2864 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2865 assignment on my part. Fixed.
2867 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2868 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2869 Janne Snabb (who went above and beyond: thank you).
2871 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2872 string otherwise requires a connection and a bunch more work and it's
2873 relatively easy to get wrong. Should also expose TLS library linkage
2876 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2877 64-bit ${eval} (JH/03).
2879 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2880 GNU libc to support some of the 64-bit stuff, should not lead to
2881 conflicts. Defined before os.h is pulled in, so if a given platform
2882 needs to override this, it can.
2884 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2885 protection layer was required, which is not implemented.
2886 Bugzilla 1254, patch from Wolfgang Breyha.
2888 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2889 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2890 tls_dhparam take prime identifiers. Also unbreak combination of
2891 OpenSSL+DH_params+TLSSNI.
2893 PP/39 Disable SSLv2 by default in OpenSSL support.
2899 PP/01 Solaris build fix for Oracle's LDAP libraries.
2900 Bugzilla 1109, patch from Stephen Usher.
2902 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2904 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2907 TF/02 Fix a couple more cases where we did not log the error message
2908 when unlink() failed. See also change 4.74-TF/03.
2910 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2911 lock up or crash if it happened to be inside a call to libc when it
2912 got a SIGUSR1 from exiwhat.
2914 The SIGUSR1 handler appends the current process status to the process
2915 log which is later printed by exiwhat. It used to use the general
2916 purpose logging code to do this, but several functions it calls are
2917 not safe for signals.
2919 The new output code in the SIGUSR1 handler is specific to the process
2920 log, and simple enough that it's easy to inspect for signal safety.
2921 Removing some special cases also simplifies the general logging code.
2922 Removing the spurious timestamps from the process log simplifies
2925 TF/04 Improved ratelimit ACL condition.
2927 The /noupdate option has been deprecated in favour of /readonly which
2928 has clearer semantics. The /leaky, /strict, and /readonly update modes
2929 are mutually exclusive. The update mode is no longer included in the
2930 database key; it just determines when the database is updated. (This
2931 means that when you upgrade Exim will forget old rate measurements.)
2933 Exim now checks that the per_* options are used with an update mode that
2934 makes sense for the current ACL. For example, when Exim is processing a
2935 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2936 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2937 must specify per_mail/readonly. If you omit the update mode it defaults to
2938 /leaky where that makes sense (as before) or /readonly where required.
2940 The /noupdate option is now undocumented but still supported for
2941 backwards compatibility. It is equivalent to /readonly except that in
2942 ACLs where /readonly is required you may specify /leaky/noupdate or
2943 /strict/noupdate which are treated the same as /readonly.
2945 A useful new feature is the /count= option. This is a generalization
2946 of the per_byte option, so that you can measure the throughput of other
2947 aggregate values. For example, the per_byte option is now equivalent
2948 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2950 The per_rcpt option has been generalized using the /count= mechanism
2951 (though it's more complicated than the per_byte equivalence). When it is
2952 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2953 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2954 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2955 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2956 Note that using per_rcpt with a non-readonly update mode in more than
2957 one ACL will cause the recipients to be double-counted. (The per_mail
2958 and per_byte options don't have this problem.)
2960 The handling of very low rates has changed slightly. If the computed rate
2961 is less than the event's count (usually one) then this event is the first
2962 after a long gap. In this case the rate is set to the same as this event's
2963 count, so that the first message of a spam run is counted properly.
2965 The major new feature is a mechanism for counting the rate of unique
2966 events. The new per_addr option counts the number of different
2967 recipients that someone has sent messages to in the last time period. It
2968 behaves like per_rcpt if all the recipient addresses are different, but
2969 duplicate recipient addresses do not increase the measured rate. Like
2970 the /count= option this is a general mechanism, so the per_addr option
2971 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2972 example, measure the rate that a client uses different sender addresses
2973 with the options per_mail/unique=$sender_address. There are further
2974 details in the main documentation.
2976 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2978 TF/06 Removed a few PCRE remnants.
2980 TF/07 Automatically extract Exim's version number from tags in the git
2981 repository when doing development or release builds.
2983 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2984 Bugzilla 879. Patch from Paul Fisher.
2986 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2987 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2990 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2992 PP/05 Variable $av_failed, true if the AV scanner deferred.
2993 Bugzilla 1078. Patch from John Horne.
2995 PP/06 Stop make process more reliably on build failure.
2996 Bugzilla 1087. Patch from Heiko Schlittermann.
2998 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2999 Bugzilla 1089. Patch from Heiko Schlittermann.
3001 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3002 Bugzilla 1131. Patch from Holger Weiß.
3004 PP/09 Handle IPv6 addresses with SPF.
3005 Bugzilla 860. Patch from Wolfgang Breyha.
3007 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3009 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3012 PP/11 match_* no longer expand right-hand-side by default.
3013 New compile-time build option, EXPAND_LISTMATCH_RHS.
3014 New expansion conditions, "inlist", "inlisti".
3016 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3018 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3020 PP/14 fix log_write() format string regression from TF/03.
3021 Bugzilla 1152. Patch from Dmitry Isaikin.
3027 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3029 PP/02 Harmonised TLS library version reporting; only show if debugging.
3030 Layout now matches that introduced for other libraries in 4.74 PP/03.
3032 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3034 PP/04 New "dns_use_edns0" global option.
3036 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3039 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3040 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3042 TK/01 Updated PolarSSL code to 0.14.2.
3043 Bugzilla 1097. Patch from Andreas Metzler.
3045 PP/07 Catch divide-by-zero in ${eval:...}.
3046 Fixes bugzilla 1102.
3048 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3051 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3052 format-string attack -- SECURITY: remote arbitrary code execution.
3054 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3055 time unintentionally subject to list matching rules, letting the header
3056 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3057 arbitrary string expansion). This allowed for information disclosure.
3059 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3060 INT_MIN/-1 -- value coerced to INT_MAX.
3066 NM/01 Workaround for PCRE version dependency in version reporting
3069 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3070 This fixes portability to compilers other than gcc, notably
3071 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3073 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3074 makefiles for portability to HP-UX and POSIX correctness.
3076 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3077 Also via indented variable definition in the Makefile.
3078 (Debugging by Oliver Heesakkers).
3080 PP/02 Restore caching of spamd results with expanded spamd_address.
3081 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3083 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3084 Improves build reliability. Fix from: Frank Elsner
3086 NM/02 Fix wide character breakage in the rfc2047 coding
3087 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3089 NM/03 Allow underscore in dnslist lookups
3090 Fixes bug 1026. Patch from Graeme Fowler
3092 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3093 Code patches from Adam Ciarcinski of NetBSD.
3095 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3098 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3099 is logged, to avoid truncation. Patch from John Horne.
3101 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3102 Patch from Jakob Hirsch.
3104 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3105 SQL string expansion failure details.
3106 Patch from Andrey Oktyabrski.
3108 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3109 Patch from Simon Arlott.
3111 PP/09 New lookups functionality failed to compile on old gcc which rejects
3112 extern declarations in function scope.
3113 Patch from Oliver Fleischmann
3115 PP/10 Use sig_atomic_t for flags set from signal handlers.
3116 Check getgroups() return and improve debugging.
3117 Fixed developed for diagnosis in bug 927 (which turned out to be
3120 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3121 Patch from Mark Zealey.
3123 PP/12 Bugzilla 1056: Improved spamd server selection.
3124 Patch from Mark Zealey.
3126 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3127 Based on patch from Heiko Schlittermann.
3129 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3130 Patch from Uwe Doering, sign-off by Michael Haardt.
3132 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3133 variable declaration deep within a block. Bug and patch from
3136 PP/15 lookups-Makefile IRIX compatibility coercion.
3138 PP/16 Make DISABLE_DKIM build knob functional.
3140 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3141 Patch by Simon Arlott
3143 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3144 variable argument macros. Our copy now differs from upstream.
3150 TF/01 Failure to get a lock on a hints database can have serious
3151 consequences so log it to the panic log.
3153 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3154 controlled using the smtp_confirmation log selector.
3156 TF/03 Include the error message when we fail to unlink a spool file.
3158 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3159 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3160 for maintaining out-of-tree patches for some time.
3162 PP/01 Bugzilla 139: Documentation and portability issues.
3163 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3164 Handle per-OS dynamic-module compilation flags.
3166 PP/02 Let /dev/null have normal permissions.
3167 The 4.73 fixes were a little too stringent and complained about the
3168 permissions on /dev/null. Exempt it from some checks.
3169 Reported by Andreas M. Kirchwitz.
3171 PP/03 Report version information for many libraries, including
3172 Exim version information for dynamically loaded libraries. Created
3173 version.h, now support a version extension string for distributors
3174 who patch heavily. Dynamic module ABI change.
3176 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3177 privilege escalation vulnerability whereby the Exim run-time user
3178 can cause root to append content of the attacker's choosing to
3181 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3184 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3185 If dropping privileges for untrusted macros, we disabled normal logging
3186 on the basis that it would fail; for the Exim run-time user, this is not
3187 the case, and it resulted in successful deliveries going unlogged.
3188 Fixed. Reported by Andreas Metzler.
3194 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3195 only prepend for the Resent-* case. Fixes regression introduced in
3196 Exim 4.70 by NM/22 for Bugzilla 607.
3198 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3199 increasing numbers of administrators be bitten by this.
3201 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3203 PP/03 Bugzilla 994: added openssl_options main configuration option.
3205 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3207 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3209 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3211 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3212 without a peer certificate, leading to a segfault because of an
3213 assumption that peers always have certificates. Be a little more
3214 paranoid. Problem reported by Martin Tscholak.
3216 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3217 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3218 NB: ClamAV planning to remove STREAM in "middle of 2010".
3219 CL also introduces -bmalware, various -d+acl logging additions and
3220 more caution in buffer sizes.
3222 PP/09 Implemented reverse_ip expansion operator.
3224 PP/10 Bugzilla 937: provide a "debug" ACL control.
3226 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3228 PP/12 Bugzilla 973: Implement --version.
3230 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3232 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3234 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3236 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3237 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3238 ignore trailing whitespace.
3240 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3242 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3245 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3246 Notification from Dr Andrew Aitchison.
3248 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3249 ExtendedDetectionInfo response format.
3250 Notification from John Horne.
3252 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3255 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3256 XSL and documented dependency on system catalogs, with examples of how
3259 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3262 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3263 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3264 configuration file which is writeable by the Exim user or group.
3266 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3267 of configuration files to cover files specified with the -C option if
3268 they are going to be used with root privileges, not just the default
3271 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3272 option (effectively making it always true).
3274 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3275 files to be used while preserving root privileges.
3277 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3278 that rogue child processes cannot use them.
3280 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3281 run-time user, instead of root.
3283 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3284 Exim run-time user without dropping privileges.
3286 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3287 result string, instead of calling string_vformat() twice with the same
3290 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3291 for other users. Others should always drop root privileges if they use
3292 -C on the command line, even for a whitelisted configure file.
3294 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3296 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3302 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3303 $data_path, and $header_path variables; fixed documentation bugs and
3306 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3307 exipick to access non-standard spools, including the "frozen" queue
3310 NM/01 Bugzilla 965: Support mysql stored procedures.
3311 Patch from Alain Williams
3313 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3315 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3316 Patch from Andreas Metzler
3318 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3319 Patch from Kirill Miazine
3321 NM/05 Bugzilla 671: Added umask to procmail example.
3323 JJ/03 installed exipick 20100323.0, fixing doc bug
3325 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3326 directory. Notification and patch from Dan Rosenberg.
3328 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3330 TK/02 Improve log output when DKIM signing operation fails.
3332 MH/01 Treat the transport option dkim_domain as a colon separated
3333 list, not as a single string, and sign the message with each element,
3334 omitting multiple occurences of the same signer.
3336 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3337 Bugzilla 985, 986. Patch by Simon Arlott
3339 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3340 Patch by Simon Arlott
3342 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3343 MBX locking. Notification from Dan Rosenberg.
3349 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3351 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3353 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3355 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3357 NM/04 Bugzilla 915: Flag broken perl installation during build.
3363 TK/01 Added patch by Johannes Berg that expands the main option
3364 "spamd_address" if it starts with a dollar sign.
3366 TK/02 Write list of recipients to X-Envelope-Sender header when building
3367 the mbox-format spool file for content scanning (suggested by Jakob
3370 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3371 (http://www.dcc-servers.net/) support via dccifd. Activated by
3372 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3374 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3375 by Mark Daniel Reidel <mr@df.eu>.
3377 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3378 When building exim an external PCRE library is now needed -
3379 PCRE is a system library on the majority of modern systems.
3380 See entry on PCRE_LIBS in EDITME file.
3382 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3383 conversation. Added nologin parameter to request.
3384 Patch contributed by Kirill Miazine.
3386 TF/01 Do not log submission mode rewrites if they do not change the address.
3388 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3390 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3391 log files in place. Contributed by Roberto Lima.
3393 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3395 TF/03 Bugzilla 615: When checking the local_parts router precondition
3396 after a local_part_suffix or local_part_prefix option, Exim now
3397 does not use the address's named list lookup cache, since this
3398 contains cached lookups for the whole local part.
3400 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3401 Robert Millan. Documentation is in experimental-spec.txt.
3403 TF/04 Bugzilla 668: Fix parallel build (make -j).
3405 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3407 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3408 Patch provided by Jan Srzednicki.
3410 TF/05 Leading white space used to be stripped from $spam_report which
3411 wrecked the formatting. Now it is preserved.
3413 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3414 that they are available at delivery time.
3416 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3418 TF/08 TLS error reporting now respects the incoming_interface and
3419 incoming_port log selectors.
3421 TF/09 Produce a more useful error message if an SMTP transport's hosts
3422 setting expands to an empty string.
3424 NM/06 Bugzilla 744: EXPN did not work under TLS.
3425 Patch provided by Phil Pennock.
3427 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3428 Patch provided by Richard Godbee.
3430 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3431 acl_smtp_notquit, added index entry.
3433 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3434 Patch provided by Eugene Bujak.
3436 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3437 accept(). Patch provided by Maxim Dounin.
3439 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3440 Patch provided by Phil Pennock.
3442 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3444 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3445 Patch provided by Brad "anomie" Jorsch.
3447 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3448 Patch provided by Dean Brooks.
3450 TK/05 Add native DKIM support (does not depend on external libraries).
3452 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3453 Patch provided by Graeme Fowler.
3455 NM/16 Bugzilla 851: Documentation example syntax fix.
3457 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3459 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3462 NM/19 Bugzilla 745: TLS version reporting.
3463 Patch provided by Phil Pennock.
3465 NM/20 Bugzilla 167: bool: condition support.
3466 Patch provided by Phil Pennock.
3468 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3469 clients. Patch provided by Phil Pennock.
3471 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3472 Patch provided by Brad "anomie" Jorsch.
3474 NM/23 Bugzilla 687: Fix misparses in eximstats.
3475 Patch provided by Heiko Schlittermann.
3477 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3478 Patch provided by Heiko Schlittermann.
3480 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3481 plus update to original patch.
3483 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3485 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3486 Patch provided by David Brownlee.
3488 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3490 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3492 NM/30 Bugzilla 888: TLS documentation bugfixes.
3494 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3496 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3497 Unlike the original bugzilla I have changed all shell scripts in src tree.
3499 NM/33 Bugzilla 898: Transport filter timeout fix.
3500 Patch by Todd Rinaldo.
3502 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3503 Patch by Serge Demonchaux.
3505 NM/35 Bugzilla 39: Base64 decode bug fixes.
3506 Patch by Jakob Hirsch.
3508 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3510 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3512 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3514 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3520 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3521 ALT-N's libdkim that I have put here:
3522 http://duncanthrax.net/exim-experimental/
3524 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3525 were called 'true' and it seems that C99 defines that as a reserved
3526 keyword to be used with 'bool' variable types. That means you could
3527 not include C99-style headers which use bools without triggering
3528 build errors in sieve.c.
3530 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3531 as mailq or other aliases. Changed the --help handling significantly
3532 to do whats expected. exim_usage() emits usage/help information.
3534 SC/01 Added the -bylocaldomain option to eximstats.
3536 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3538 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3540 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3546 PH/01 Another patch from the Sieve maintainer.
3548 PH/02 When an IPv6 address is converted to a string for single-key lookup
3549 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3550 dots are used instead of colons so that keys in lsearch files need not
3551 contain colons. This was done some time before quoting was made available
3552 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3553 (notated using the quote facility) so as to distinguish them from IPv4
3554 keys. This meant that lookups for IP addresses in host lists did not work
3555 for iplsearch lookups.
3557 This has been fixed by arranging for IPv6 addresses to be expressed with
3558 colons if the lookup type is iplsearch. This is not incompatible, because
3559 previously such lookups could never work.
3561 The situation is now rather anomalous, since one *can* have colons in
3562 ordinary lsearch keys. However, making the change in all cases is
3563 incompatible and would probably break a number of configurations.
3565 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3568 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3569 conversion specification without a maximum field width, thereby enabling
3570 a rogue spamd server to cause a buffer overflow. While nobody in their
3571 right mind would setup Exim to query an untrusted spamd server, an
3572 attacker that gains access to a server running spamd could potentially
3573 exploit this vulnerability to run arbitrary code as the Exim user.
3575 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3576 $primary_hostname instead of what libspf2 thinks the hosts name is.
3578 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3579 a directory entry by the name of the lookup key. Previously, if a
3580 symlink pointed to a non-existing file or a file in a directory that
3581 Exim lacked permissions to read, a lookup for a key matching that
3582 symlink would fail. Now it is enough that a matching directory entry
3583 exists, symlink or not. (Bugzilla 503.)
3585 PH/03 The body_linecount and body_zerocount variables are now exported in the
3588 PH/04 Added the $dnslist_matched variable.
3590 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3591 This means they are set thereafter only if the connection becomes
3594 PH/06 Added the client_condition to authenticators so that some can be skipped
3595 by clients under certain conditions.
3597 PH/07 The error message for a badly-placed control=no_multiline_responses left
3598 "_responses" off the end of the name.
3600 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3602 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3603 (without spaces) instead of just copying the configuration text.
3605 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3607 PH/11 Added $max_received_linelength.
3609 PH/12 Added +ignore_defer and +include_defer to host lists.
3611 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3612 way in which PCRE > 7.0 is built.
3614 PH/14 Implemented queue_only_load_latch.
3616 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3617 MAIL command. The effect was to mangle the value on 64-bit systems.
3619 PH/16 Another patch from the Sieve maintainer.
3621 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3623 PH/18 If a system quota error occurred while trying to create the file for
3624 a maildir delivery, the message "Mailbox is full" was not appended to the
3625 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3626 only to a quota excession during the actual writing of the file.
3628 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3629 characters?) which causes problems in log lines. The DN values are now
3630 passed through string_printing() before being added to log lines.
3632 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3633 and InterBase are left for another time.)
3635 PH/21 Added message_body_newlines option.
3637 PH/22 Guard against possible overflow in moan_check_errorcopy().
3639 PH/23 POSIX allows open() to be a macro; guard against that.
3641 PH/24 If the recipient of an error message contained an @ in the local part
3642 (suitably quoted, of course), incorrect values were put in $domain and
3643 $local_part during the evaluation of errors_copy.
3649 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3650 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3653 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3654 issue a MAIL command.
3656 PH/02 In an ACL statement such as
3658 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3660 if a client was not listed at all, or was listed with a value other than
3661 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3662 the condition was not true (as it should be), so access was not denied.
3663 The bug was that the ! inversion was incorrectly passed on to the second
3664 item. This has been fixed.
3666 PH/03 Added additional dnslists conditions == and =& which are different from
3667 = and & when the dns lookup returns more than one IP address.
3669 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3670 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3672 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3673 FSYNC, which compiles an option called disable_fsync that allows for
3674 bypassing fsync(). The documentation is heavily laced with warnings.
3676 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3678 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3679 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3680 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3681 including adding "make clean"; (3) Added -fPIC when compiling the test
3682 dynamically loaded module, to get rid of a warning.
3684 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3685 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3686 = 0s. The bug is otherwise harmless.
3688 PH/07 There was a bug in the dovecot authenticator such that the value of
3689 $auth1 could be overwritten, and so not correctly preserved, after a
3690 successful authentication. This usually meant that the value preserved by
3691 the server_setid option was incorrect.
3693 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3695 PH/09 Installed PCRE release 7.0.
3697 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3698 run for batched SMTP input. It is now run at the start of every message
3699 in the batch. While fixing this I discovered that the process information
3700 (output by running exiwhat) was not always getting set for -bs and -bS
3701 input. This is fixed, and it now also says "batched" for BSMTP.
3703 PH/11 Added control=no_pipelining.
3705 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3706 patch, slightly modified), and move the expansion of helo_data till after
3707 the connection is made in the smtp transport (so it can use these
3710 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3712 PH/14 Added log_selector = +pid.
3714 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3716 PH/16 Add ${if forany and ${if forall.
3718 PH/17 Added dsn_from option to vary the From: line in DSNs.
3720 PH/18 Flush SMTP output before performing a callout, unless control =
3721 no_callout_flush is set.
3723 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3724 was true (the default) a successful delivery failed to delete the retry
3725 item, thus causing premature timeout of the address. The bug is now
3728 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3730 PH/21 Long custom messages for fakedefer and fakereject are now split up
3731 into multiline responses in the same way that messages for "deny" and
3732 other ACL rejections are.
3734 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3735 with slight modification.
3737 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3738 draft, changing the syntax and factoring some duplicate code".
3740 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3741 for deliveries of the second and subsequent messages over the same SMTP
3744 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3745 ${reduce, with only minor "tidies".
3747 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3749 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3750 expansion side effects.
3752 PH/27 When a message times out after an over-quota error from an Exim-imposed
3753 quota, the bounce message says "mailbox is full". This message was not
3754 being given when it was a system quota that was exceeded. It now should
3757 MH/03 Made $recipients available in local_scan(). local_scan() already has
3758 better access to the recipient list through recipients_list[], but
3759 $recipients can be useful in postmaster-provided expansion strings.
3761 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3762 in the case of a MAIL command with additional options following the
3763 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3764 were accidentally chopped off.
3766 PH/29 SMTP synchronization checks are implemented when a command is read -
3767 there is a check that no more input is waiting when there shouldn't be
3768 any. However, for some commands, a delay in an ACL can mean that it is
3769 some time before the response is written. In this time, more input might
3770 arrive, invalidly. So now there are extra checks after an ACL has run for
3771 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3772 pipelining has not been advertised.
3774 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3776 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3777 returned in the SMTP response when smtp_return_error_details was false.
3778 This has been fixed.
3780 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3781 instead of the C I/O that was originally supplied, because problems were
3782 reported on Solaris.
3784 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3785 Exim which did not show up earlier: it was assuming that a call to
3786 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3787 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3788 was a macro that became an assignment, so it seemed to work. This has
3789 changed to a proper function call with a void return, hence the compile
3790 error. Exim's code has been fixed.
3792 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3795 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3797 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3799 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3800 because it turns out to be too big or there is a timeout) there is no
3801 "Completed" line in the log. When some messages of this type were
3802 selected by exigrep, they were listed as "not completed". Others were
3803 picked up by some special patterns. I have improved the selection
3804 criteria to be more general.
3806 PH/38 The host_find_failed option in the manualroute router can now be set
3807 to "ignore", to completely ignore a host whose IP address cannot be
3808 found. If all hosts are ignored, the behaviour is controlled by the new
3809 host_all_ignored option.
3811 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3812 homing or because of multiple MX records with /mx) generated more than
3813 one IP address, and the following item turned out to be the local host,
3814 all the secondary addresses of the first item were incorrectly removed
3815 from the list, along with the local host and any following hosts (which
3816 is what is supposed to happen).
3818 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3819 whoever called Exim into the -H file. In the case of the daemon it was
3820 behaving confusingly. When first started, it used values for whoever
3821 started the daemon, but after a SIGHUP it used the Exim user (because it
3822 calls itself on a restart). I have changed the code so that it now always
3825 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3826 message are rejected with the same error (e.g. no authentication or bad
3827 sender address), and a DATA command is nevertheless sent (as can happen
3828 with PIPELINING or a stupid MUA), the error message that was given to the
3829 RCPT commands is included in the rejection of the DATA command. This is
3830 intended to be helpful for MUAs that show only the final error to their
3833 PH/42 Another patch from the Sieve maintainer.
3835 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3836 Eximstats - Fixed some broken HTML links and added missing column headers
3838 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3839 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3841 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3843 PH/43 Yet another patch from the Sieve maintainer.
3845 PH/44 I found a way to check for a TCP/IP connection going away before sending
3846 the response to the final '.' that terminates a message, but only in the
3847 case where the client has not sent further data following the '.'
3848 (unfortunately, this is allowed). However, in many cases there won't be
3849 any further data because there won't be any more messages to send. A call
3850 to select() can be used: if it shows that the input is "ready", there is
3851 either input waiting, or the socket has been closed. An attempt to read
3852 the next input character can distinguish the two cases. Previously, Exim
3853 would have sent an OK response which the client would never have see.
3854 This could lead to message repetition. This fix should cure that, at
3855 least in a lot of common cases.
3857 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3858 advertised in response to EHLO.
3864 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3865 fixed by 4.65/MH/01 (is this a record?) are fixed:
3867 (i) An empty string was always treated as zero by the numeric comparison
3868 operators. This behaviour has been restored.
3870 (ii) It is documented that the numeric comparison operators always treat
3871 their arguments as decimal numbers. This was broken in that numbers
3872 starting with 0 were being interpreted as octal.
3874 While fixing these problems I realized that there was another issue that
3875 hadn't been noticed. Values of message_size_limit (both the global option
3876 and the transport option) were treated as octal if they started with 0.
3877 The documentation was vague. These values are now always treated as
3878 decimal, and I will make that clear in the documentation.
3884 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3885 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3888 MH/01 Don't check that the operands of numeric comparison operators are
3889 integers when their expansion is in "skipping" mode (fixes bug
3890 introduced by 4.64-PH/07).
3892 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3893 child addresses, Exim now panics and dies. Previously, because the count
3894 is held in a short int, deliveries were likely to be lost. As such a
3895 large number of recipients for a single message is ridiculous
3896 (performance will be very, very poor), I have chosen to impose a limit
3897 rather than extend the field.
3903 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3904 leftover -K file (the existence of which was triggered by #402).
3905 While we were at it, introduced process PID as part of the -K
3906 filename. This should rule out race conditions when creating
3909 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3910 processing considerably. Previous code took too long for large mails,
3911 triggering a timeout which in turn triggers #401.
3913 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3914 in the DK code in transports.c. sendfile() is not really portable,
3915 hence the _LINUX specificness.
3917 TF/01 In the add_headers option to the mail command in an Exim filter,
3918 there was a bug that Exim would claim a syntax error in any
3919 header after the first one which had an odd number of characters
3922 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3923 callout verification, Exim cached a "reject" for the entire domain. This
3924 is correct for most verifications, but it is not correct for a recipient
3925 verification with use_sender or use_postmaster set, because in that case
3926 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3927 case of MAIL FROM:<> rejection from other early rejections (e.g.
3928 rejection of HELO). When verifying a recipient using a non-null MAIL
3929 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3930 Whatever the result of the callout, the value of the domain cache is
3931 left unchanged (for any other kind of callout, getting as far as trying
3932 RCPT means that the domain itself is ok).
3934 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3937 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3938 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3939 session, the daemon ignores ECONNECT errors and logs others; it now
3940 ignores EPIPE as well.
3942 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3943 (quoted-printable decoding).
3945 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3946 later the small subsequent patch to fix an introduced bug.
3948 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3950 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3952 PH/08 An error is now given if message_size_limit is specified negative.
3954 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3955 to be given (somewhat) arbitrary names.
3957 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3960 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3961 miscellaneous code fixes
3963 PH/10 Added the log_reject_target ACL modifier to specify where to log
3966 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3967 hostname. This is wrong, because it relates to the incoming message (and
3968 probably the interface on which it is arriving) and not to the outgoing
3969 callout (which could be using a different interface). This has been
3970 changed to use the value of the helo_data option from the smtp transport
3971 instead - this is what is used when a message is actually being sent. If
3972 there is no remote transport (possible with a router that sets up host
3973 addresses), $smtp_active_hostname is used.
3975 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3976 tweaks were necessary in order to get it to work (see also 21 below):
3977 (a) The code assumed that strncpy() returns a negative number on buffer
3978 overflow, which isn't the case. Replaced with Exim's string_format()
3980 (b) There were several signed/unsigned issues. I just did the minimum
3981 hacking in of casts. There is scope for a larger refactoring.
3982 (c) The code used strcasecmp() which is not a standard C function.
3983 Replaced with Exim's strcmpic() function.
3984 (d) The code set only $1; it now sets $auth1 as well.
3985 (e) A simple test gave the error "authentication client didn't specify
3986 service in request". It would seem that Dovecot has changed its
3987 interface. Fortunately there's a specification; I followed it and
3988 changed what the client sends and it appears to be working now.
3990 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3993 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3994 address A is aliased to B and C, where B exists and C does not. Without
3995 -v the output is "A verified" because verification stops after a
3996 successful redirection if more than one address is generated. However,
3997 with -v the child addresses are also verified. Exim was outputting "A
3998 failed to verify" and then showing the successful verification for C,
3999 with its parentage. It now outputs "B failed to verify", showing B's
4000 parentage before showing the successful verification of C.
4002 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4003 look up a TXT record in a specific list after matching in a combined
4006 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4007 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4008 they consult the DNS. I had assumed they would set it the way they
4009 wanted; and indeed my experiments on Linux seem to show that in some
4010 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4011 To be on the safe side, however, I have now made the interface to
4012 host_find_byname() similar to host_find_bydns(), with an argument
4013 containing the DNS resolver options. The host_find_byname() function now
4014 sets these options at its start, just as host_find_bydns() does. The smtp
4015 transport options dns_qualify_single and dns_search_parents are passed to
4016 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4017 of host_find_byname() use the default settings of RES_DEFNAMES
4018 (qualify_single) but not RES_DNSRCH (search_parents).
4020 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4021 spool_read_header() do less string testing, by means of a preliminary
4022 switch on the second character of optional "-foo" lines. (This is
4023 overdue, caused by the large number of possibilities that now exist.
4024 Originally there were few.) While I was there, I also converted the
4025 str(n)cmp tests so they don't re-test the leading "-" and the first
4026 character, in the hope this might squeeze out yet more improvement.
4028 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4029 flag allowing group syntax was set by the header_syntax check but not
4030 turned off, possible causing trouble later; (2) The flag was not being
4031 set at all for the header_verify test, causing "group"-style headers to
4032 be rejected. I have now set it in this case, and also caused header_
4033 verify to ignore an empty address taken from a group. While doing this, I
4034 came across some other cases where the code for allowing group syntax
4035 while scanning a header line wasn't quite right (mostly, not resetting
4036 the flag correctly in the right place). These bugs could have caused
4037 trouble for malformed header lines. I hope it is now all correct.
4039 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4040 with the "reply" argument non-NULL. The code, however (which originally
4041 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4042 but it didn't always do it. This confused somebody who was copying the
4043 code for some other use. I have removed all the tests.
4045 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4046 feature that was used to support insecure browsers during the U.S. crypto
4047 embargo. It requires special client support, and Exim is probably the
4048 only MTA that supported it -- and would never use it because real RSA is
4049 always available. This code has been removed, because it had the bad
4050 effect of slowing Exim down by computing (never used) parameters for the
4051 RSA_EXPORT functionality.
4053 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4054 authenticator to fail if there's a tab character in the incoming data
4055 (there should never be unless someone is messing about, as it's supposed
4056 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4057 if the connection is using TLS or if the remote IP is the same as the
4058 local IP, and the "valid-client-cert option" if a client certificate has
4061 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4062 authenticators. This can be used for authorization after authentication
4063 succeeds. (In the case of plaintext, it servers for both authentication
4066 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4067 if any retry times were supplied.
4069 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4070 connection, where there is no client IP address to check. In this
4071 situation, the verify now always succeeds.
4073 PH/25 Applied John Jetmore's -Mset patch.
4075 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4077 PH/27 In a string expansion for a processed (not raw) header when multiple
4078 headers of the same name were present, leading whitespace was being
4079 removed from all of them, but trailing whitespace was being removed only
4080 from the last one. Now trailing whitespace is removed from each header
4081 before concatenation. Completely empty headers in a concatenation (as
4082 before) are ignored.
4084 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4085 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4087 PH/29 [Removed. This was a change that I later backed out, and forgot to
4088 correct the ChangeLog entry (that I had efficiently created) before
4089 committing the later change.]
4091 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4092 address errors (4xx response to RCPT) over the same connection as other
4093 messages routed to the same hosts. Such deliveries are always "forced",
4094 so retry times are not inspected. This resulted in far too many retries
4095 for the affected addresses. The effect occurred only when there were more
4096 hosts than the hosts_max_try setting in the smtp transport when it had
4097 the 4xx errors. Those hosts that it had tried were not added to the list
4098 of hosts for which the message was waiting, so if all were tried, there
4099 was no problem. Two fixes have been applied:
4101 (i) If there are any address or message errors in an SMTP delivery, none
4102 of the hosts (tried or untried) are now added to the list of hosts
4103 for which the message is waiting, so the message should not be a
4104 candidate for sending over the same connection that was used for a
4105 successful delivery of some other message. This seems entirely
4106 reasonable: after all the message is NOT "waiting for some host".
4107 This is so "obvious" that I'm not sure why it wasn't done
4108 previously. Hope I haven't missed anything, but it can't do any
4109 harm, as the worst effect is to miss an optimization.
4111 (ii) If, despite (i), such a delivery is accidentally attempted, the
4112 routing retry time is respected, so at least it doesn't keep
4113 hammering the server.
4115 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4116 in ${readsocket because some servers need this prod.
4118 PH/32 Added some extra debug output when updating a wait-xxx database.
4120 PH/33 The hint "could be header name not terminated by colon", which has been
4121 given for certain expansion errors for a long time, was not being given
4122 for the ${if def:h_colon_omitted{... case.
4124 PH/34 The spec says: "With one important exception, whenever a domain list is
4125 being scanned, $domain contains the subject domain." There was at least
4126 one case where this was not true.
4128 PH/35 The error "getsockname() failed: connection reset by peer" was being
4129 written to the panic log as well as the main log, but it isn't really
4130 panic-worthy as it just means the connection died rather early on. I have
4131 removed the panic log writing for the ECONNRESET error when getsockname()
4134 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4135 runs only) independently of the message's sender address. This meant
4136 that, if the 4xx error was in fact related to the sender, a different
4137 message to the same recipient with a different sender could confuse
4138 things. In particular, this can happen when sending to a greylisting
4139 server, but other circumstances could also provoke similar problems.
4140 I have changed the default so that the retry time for these errors is now
4141 based a combination of the sender and recipient addresses. This change
4142 can be overridden by setting address_retry_include_sender=false in the
4145 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4146 remote server are returned as part of bounce messages. This was not
4147 happening for LMTP over a pipe (the lmtp transport), but now it is the
4148 same for both kinds of LMTP.
4150 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4151 in header lines that were in fact CNAMEs. This is no longer the case.
4153 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4154 and queue runs started by the daemon processed all messages. This has
4155 been fixed so that -R and -S can now usefully be given with -q<time>.
4157 PH/40 Import PCRE release 6.7 (fixes some bugs).
4159 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4161 PH/42 Give an error if -q is specified more than once.
4163 PH/43 Renamed the variables $interface_address and $interface_port as
4164 $received_ip_address and $received_port, to make it clear that these
4165 values apply to message reception, and not to the outgoing interface when
4166 a message is delivered. (The old names remain recognized, of course.)
4168 PH/44 There was no timeout on the connect() call when using a Unix domain
4169 socket in the ${readsocket expansion. There now is.
4171 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4172 be meaningful with "accept".
4174 SC/01 Eximstats V1.43
4175 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4177 SC/02 Eximstats V1.44
4178 Use a glob alias rather than an array ref in the generated
4179 parser. This improves both readability and performance.
4181 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4182 Collect SpamAssassin and rejection statistics.
4183 Don't display local sender or destination tables unless
4184 there is data to show.
4185 Added average volumes into the top table text output.
4187 SC/04 Eximstats V1.46
4188 Collect data on the number of addresses (recipients)
4189 as well as the number of messages.
4191 SC/05 Eximstats V1.47
4192 Added 'Message too big' to the list of mail rejection
4193 reasons (thanks to Marco Gaiarin).
4195 SC/06 Eximstats V1.48
4196 Mainlog lines which have GMT offsets and are too short to
4197 have a flag are now skipped.
4199 SC/07 Eximstats V1.49 (Alain Williams)
4200 Added the -emptyok flag.
4202 SC/08 Eximstats V1.50
4203 Fixes for obtaining the IP address from reject messages.
4205 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4206 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4207 whitespace changes from 4.64-PH/27
4209 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4212 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4213 are found, allow negative numbers in numeric criteria)
4215 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4217 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4218 to match changes made in 4.64-PH/43
4220 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4222 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4223 "domain" is actually an IP address, and force a failure. This locks out
4224 those revolvers/nameservers that support "A-for-A" lookups, in
4225 contravention of the specifications.
4227 PH/48 When a host name was looked up from an IP address, and the subsequent
4228 forward lookup of the name timed out, the host name was left in
4229 $sender_host_name, contrary to the specification.
4231 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4232 restricted to single-key lookups, Exim was not diagnosing an error if
4233 * or *@ was used with a query-style lookup.
4235 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4237 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4238 long ago, but noone interested enough thought of it. Let's just say that
4239 the "1.1" means that there are some new functions that weren't there at
4240 some point in the past.
4242 PH/51 Error processing for expansion failure of helo_data from an smtp
4243 transport during callout processing was broken.
4245 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4246 tested/used via the -bh/-bhc/-bs options.
4248 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4249 bug, fixed in subsequent PCRE releases).
4251 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4252 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4254 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4260 SC/01 Use a glob alias rather than an array ref in eximstats generated
4261 parser. This improves both readability and performance.
4263 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4264 Don't display local sender or destination tables in eximstats unless
4265 there is data to show.
4266 Added average volumes into the eximstats top table text output.
4268 SC/03 Collect data on the number of addresses (recipients) as well
4269 as the number of messages in eximstats.
4271 TF/01 Correct an error in the documentation for the redirect router. Exim
4272 does (usually) call initgroups() when daemonizing.
4274 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4275 with consistent privilege compared to when running as a daemon.
4277 TF/03 Note in the spec that $authenticated_id is not set for local
4278 submissions from trusted users.
4280 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4281 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4283 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4284 by adding some example configuration directives to the default
4285 configuration file. A little bit of work is required to uncomment the
4286 directives and define how usernames and passwords are checked, but
4287 there is now a framework to start from.
4289 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4290 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4291 without this. I don't know how relevant this is to other LDAP libraries.
4293 PH/02 Add the verb name to the "unknown ACL verb" error.
4295 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4297 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4299 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4300 directory not expanded when it should be if an expanded home directory
4301 was set for the address (which is overridden by the transport).
4303 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4306 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4307 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4308 because it is too late at that time, and has no effect.
4310 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4311 security issue with \' (bugzilla #107). I could not use the
4312 PQescapeStringConn() function, because it needs a PGconn value as one of
4315 PH/08 When testing addresses using -bt, indicate those final addresses that
4316 are duplicates that would not cause an additional delivery. At least one
4317 person was confused, thinking that -bt output corresponded to deliveries.
4318 (Suppressing duplicates isn't a good idea as you lose the information
4319 about possibly different redirections that led to the duplicates.)
4321 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4322 systems where poll() doesn't work, in particular OS X.
4324 PH/10 Added more information to debugging output for retry time not reached.
4326 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4327 operations in malware.c.
4329 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4332 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4333 syslog_duplication set false, log lines that would normally be written
4334 both the the main log and to the reject log were not written to syslog at
4337 PH/14 In the default configuration, change the use of "message" in ACL warn
4338 statements to "add_header".
4340 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4341 not followed by a command (e.g. "seen endif").
4343 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4344 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4347 PH/17 Added extra conditions to the default value of delay_warning_condition
4351 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4352 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4353 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4356 The Auto-Submitted: and various List- headers are standardised, whereas I
4357 don't think Precedence: ever was.
4359 PH/18 Refactored debugging code in route_finduser() to show more information,
4360 in particular, the error code if getpwnam() issues one.
4362 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4363 This is apparently needed in addition to the PH/07 change above to avoid
4364 any possible encoding problems.
4366 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4367 but not after initializing Perl.
4369 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4370 output them only if debugging. By default they are written stderr,
4371 apparently, which is not desirable.
4373 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4376 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4379 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4381 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4382 authenticated or an ident call has been made. Suppress the default
4383 values for $authenticated_id and $authenticated_sender (but permit -oMai
4384 and -oMas) when testing with -bh.
4386 PH/24 Re-jigged the order of the tests in the default configuration so that the
4387 tests for valid domains and recipients precede the DNS black list and CSA
4388 tests, on the grounds that those ones are more expensive.
4390 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4391 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4392 This bug exists in every version of Exim that I still have, right back to
4395 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4396 However, an attempt to turn on case-sensitivity in a regex key by
4397 including (?-i) didn't work because the subject string was already
4398 lowercased, and the effects were non-intuitive. It turns out that a
4399 one-line patch can be used to allow (?-i) to work as expected.
4405 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4406 other effects) broke the use of negated acl sub-conditions.
4408 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4411 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4412 "Deny" causes Exim to reject the incoming connection with a 554 error.
4413 Unfortunately, if there is a major crisis, such as a disk failure,
4414 tcp-wrappers gives "deny", whereas what one would like would be some
4415 kind of temporary error. A kludge has been added to help with this.
4416 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4417 554 error is used if errno is still zero or contains ENOENT (which occurs
4418 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4421 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4423 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4424 errors. Otherwise a message that provokes a temporary error (when other
4425 messages do not) can cause a whole host to time out.
4427 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4428 addresses were routed directly to files or pipes from a redirect router.
4429 File deliveries just didn't batch; pipe deliveries might have suffered
4432 PH/06 A failure to get a lock for a hints database would erroneously always say
4433 "Failed to get write lock", even when it was really a read lock.
4435 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4436 of 0600. This has been changed to use the value of the lockfile_mode
4437 option (which defaults to 0600).
4439 PH/08 Applied small patch from the Sieve maintainer.
4441 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4442 folder from quota calculations, a direct delivery into this folder messed
4443 up the contents of the maildirsize file. This was because the regex was
4444 used only to exclude .Trash (or whatever) when the size of the mailbox
4445 was calculated. There was no check that a delivery was happening into an
4446 excluded directory. This bug has been fixed by ignoring all quota
4447 processing for deliveries into excluded directories.
4449 PH/10 Added the maildirfolder_create_regex option to appendfile.
4455 PH/01 The code for finding all the local interface addresses on a FreeBSD
4456 system running IPv6 was broken. This may well have applied to all BSD
4457 systems, as well as to others that have similar system calls. The broken
4458 code found IPv4 interfaces correctly, but gave incorrect values for the
4459 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4460 that it would not match correctly against @[] and not recognize the IPv6
4463 PH/02 The ipliteral router was not recognizing addresses of the form user@
4464 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4466 PH/03 Added disable_ipv6.
4468 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4469 decoded form, because it is most often used to construct To: headers
4470 lines in autoreplies, and the decoded form may well be syntactically
4471 invalid. However, $reply_address has leading white space removed, and all
4472 newlines turned into spaces so that the autoreply transport does not
4475 PH/05 If group was specified without a user on a router, and no group or user
4476 was specified on a transport, the group from the router was ignored.
4478 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4479 for visible compile-time settings that can be used to change these
4480 numbers, for those that want even more. Backwards compatibility with old
4481 spool files has been maintained. However, going back to a previous Exim
4482 release will lost any variables that are in spool files.
4484 PH/07 Two small changes when running in the test harness: increase delay when
4485 passing a TCP/IP connection to a new process, in case the original
4486 process has to generate a bounce, and remove special handling of
4487 127.0.0.2 (sic), which is no longer necessary.
4489 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4490 be the same on different OS.
4492 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4495 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4496 whether --show-vars was specified or not
4498 JJ/02 exipick: Added support for new ACL variable spool format introduced
4501 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4502 syntactically invalid From: or Reply-to: line, and a filter used this to
4503 generate an autoreply, and therefore failed to obtain an address for the
4504 autoreply, Exim could try to deliver to a non-existent relative file
4505 name, causing unrelated and misleading errors. What now happens is that
4506 it logs this as a hard delivery error, but does not attempt to create a
4509 PH/11 The exinext utility has a -C option for testing purposes, but although
4510 the given file was scanned by exinext itself; it wasn't being passed on
4511 when Exim was called.
4513 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4514 an end-of-file indication when reading a command response.
4516 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4517 compiled. In many other places in Exim, IPv6 addresses are always
4518 recognized, so I have changed this. It also means that IPv4 domain
4519 literals of the form [IPV4:n.n.n.n] are now always recognized.
4521 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4522 used if the router is not running as root, for example, when verifying at
4523 ACL time, or when using -bh. The debugging output from this situation was
4524 non-existent - all you got was a failure to exec. I have made two
4527 (a) Failures to set uid/gid, the current directory, or a process leader
4528 in a subprocess such as that created by queryprogram now generate
4529 suitable debugging output when -d is set.
4531 (b) The queryprogram router detects when it is not running as root,
4532 outputs suitable debugging information if -d is set, and then runs
4533 the subprocess without attempting to change uid/gid.
4535 PH/15 Minor change to Makefile for building test_host (undocumented testing
4538 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4539 additional section of a DNS packet that returns MX or SRV records.
4540 Instead, it always explicitly searches for A/AAAA records. This avoids
4541 major problems that occur when a DNS server includes only records of one
4542 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4543 fixed another bug: if SRV records were looked up and the corresponding
4544 address records were *not* found in the additional section, the port
4545 values from the SRV records were lost.
4547 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4548 using the correct key (the original address) when searching the retry
4549 rules in order to find which one to use for generating the retry hint.
4551 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4552 adding the default one. Similarly, if it contains a Reply-To: header, the
4553 errors_reply_to option, if set, is not used.
4555 PH/19 When calculating a retry time, Exim used to measure the "time since
4556 failure" by looking at the "first failed" field in the retry record. Now
4557 it does not use this if it is later than than the arrival time of the
4558 message. Instead it uses the arrival time. This makes for better
4559 behaviour in cases where some deliveries succeed, thus re-setting the
4560 "first failed" field. An example is a quota failure for a huge message
4561 when small messages continue to be delivered. Without this change, the
4562 "time since failure" will always be short, possible causing more frequent
4563 delivery attempts for the huge message than are intended.
4564 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4566 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4567 $1, $2, $3) because the numerical variables can be reset during some
4568 expansion items (e.g. "match"), thereby losing the authentication data.
4570 PH/21 Make -bV show the size of off_t variables so that the test suite can
4571 decide whether to run tests for quotas > 2G.
4573 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4574 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4575 filecount value is greater than 2G or if a quota value is greater than 2G
4576 on a system where the size of off_t is not greater than 4, a panic error
4579 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4580 never match. The debug and -bh output now contains an explicit error
4581 message indicating a malformed IPv4 address or mask.
4583 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4584 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4585 PH/23 above applies.
4587 PH/25 Do not write to syslog when running in the test harness. The only
4588 occasion when this arises is a failure to open the main or panic logs
4589 (for which there is an explicit test).
4591 PH/26 Added the /no_tell option to "control=freeze".
4593 PH/27 If a host name lookup failed very early in a connection, for example, if
4594 the IP address matched host_lookup and the reverse lookup yielded a name
4595 that did not have a forward lookup, an error message of the form "no IP
4596 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4597 could be logged. Now it outputs the IP address instead of "NULL".
4599 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4600 allows the sender and the authenticated sender to be set when
4601 submitting a message from within Exim. Since child_open_exim() is
4602 documented for local_scan(), the new function should be too.
4604 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4605 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4606 results in an empty string is now treated as unset.
4608 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4610 PH/31 Added sender_verify_fail logging option.
4612 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4613 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4614 tidied the source and removed it altogether.
4616 PH/33 When a queue run was abandoned because the load average was too high, a
4617 log line was always written; now it is written only if the queue_run log
4618 selector is set. In addition, the log line for abandonment now contains
4619 information about the queue run such as the pid. This is always present
4620 in "start" and "stop" lines but was omitted from the "abandon" line.
4622 PH/34 Omit spaces between a header name and the colon in the error message that
4623 is given when verify = headers_syntax fails (if there are lots of them,
4624 the message gets confusing).
4626 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4627 names, as there are now some PTR records that contain slashes. This check
4628 is only to protect against broken name servers that fall over on strange
4629 characters, so the fact that it applies to all lookups doesn't matter.
4631 PH/36 Now that the new test suite is complete, we can remove some of the
4632 special code in Exim that was needed for the old test suite. For example,
4633 sorting DNS records because real resolvers return them in an arbitrary
4634 order. The new test suite's fake resolver always returns records in the
4637 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4638 bounces) except when queue_only is set, to avoid logging races between
4639 the different processes.
4641 PH/38 Panic-die if .include specifies a non-absolute path.
4643 PH/39 A tweak to the "H" retry rule from its user.
4645 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4646 a label. They prevented compilation on older perls.
4648 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4649 a warning to be raised on newish perls.
4651 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4652 on queue. Changes to match documented behaviour of showing count of
4653 messages matching specified criteria.
4655 PH/40 Changed the default ident timeout from 30s to 5s.
4657 PH/41 Added support for the use of login_cap features, on those BSD systems
4658 that have them, for controlling the resources used by pipe deliveries.
4660 PH/42 The content-scanning code uses fopen() to create files in which to put
4661 message data. Previously it was not paying any attention to the mode of
4662 the files. Exim runs with umask(0) because the rest of the code creates
4663 files with open(), and sets the required mode explicitly. Thus, these
4664 files were ending up world-writeable. This was not a big issue, because,
4665 being within the spool directory, they were not world-accessible. I have
4666 created a function called modefopen, which takes an additional mode
4667 argument. It sets umask(777), creates the file, chmods it to the required
4668 mode, then resets the umask. All the relevant calls to fopen() in the
4669 content scanning code have been changed to use this function.
4671 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4672 to 24 hours. This avoids potential overflow problems when processing G
4673 and H retry rules. I suspect nobody ever tinkers with this value.
4675 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4677 PH/45 When the plaintext authenticator is running as a client, the server's
4678 challenges are checked to ensure they are valid base64 strings. By
4679 default, the authentication attempt is cancelled if an invalid string is
4680 received. Setting client_ignore_invalid_base64 true ignores these errors.
4681 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4682 they are received. Thus, the responses can be made to depend on the
4683 challenges. If an invalid string is ignored, an empty string is placed in
4686 PH/46 Messages that are created by the autoreply transport now contains a
4687 References: header, in accordance with RFCs 2822 and 3834.
4689 PH/47 Added authenticated_sender_force to the smtp transport.
4691 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4693 PH/49 Installed latest patch from the Sieve maintainer.
4695 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4696 was also set, the appendfile transport was unnecessarily scanning a
4697 directory of message files (e.g. for maildir delivery) to find the count
4698 of files (along with the size), even though it did not need this
4699 information. It now does the scan only if it needs to find either the
4700 size of the count of files.
4702 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4704 PH/52 Two bugs concerned with error handling when the smtp transport is
4707 (i) Exim was not creating retry information for temporary errors given
4708 for individual recipients after the DATA command when the smtp transport
4709 was used in LMTP mode. This meant that they could be retried too
4710 frequently, and not timed out correctly.
4712 (ii) Exim was setting the flag that allows error details to be returned
4713 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4714 recipients that were returned after the DATA command.
4716 PH/53 This is related to PH/52, but is more general: for any failing address,
4717 when detailed error information was permitted to be returned to the
4718 sender, but the error was temporary, then after the final timeout, only
4719 "retry timeout exceeded" was returned. Now it returns the full error as
4720 well as "retry timeout exceeded".
4722 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4723 do this, and (what is worse) MTAs that accept it.
4725 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4726 will now be deprecated.
4728 PH/56 New os.c-cygwin from the Cygwin maintainer.
4730 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4731 formats (previously only available in exim formats via -bpr, -bpru,
4732 and -bpra. Now also available in native and exiqgrep formats)
4734 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4735 with very large, slow to parse queues
4737 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4739 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4741 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4742 responses to authentication challenges, though it was showing the
4743 challenges; (ii) I've removed the CR characters from the debug output for
4746 PH/58 Allow for the insertion of a newline as well as a space when a string
4747 is turned into more than one encoded-word during RFC 2047 encoding. The
4748 Sieve code now uses this.
4750 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4751 data_4xx, lost_connection, tls_required.
4753 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4754 message was being sent as an SMTP response.
4756 PH/61 Add -l and -k options to exicyclog.
4758 PH/62 When verifying, if an address was redirected to one new address, so that
4759 verification continued, and the new address failed or deferred after
4760 having set something in $address_data, the value of $address_data was not
4761 passed back to the ACL. This was different to the case when no
4762 redirection occurred. The value is now passed back in both cases.
4764 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4765 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4766 use login_cap.h, so on its own it isn't the distinguishing feature. The
4767 new name refers directly to the setclassresources() function.
4769 PH/65 Added configuration files for NetBSD3.
4771 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4773 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4774 is preferred over IPv4.
4776 PH/68 The bounce_return_message and bounce_return_body options were not being
4777 honoured for bounces generated during the reception of non-SMTP messages.
4778 In particular, this applied to messages rejected by the ACL. This bug has
4779 been fixed. However, if bounce_return_message is true and bounce_return_
4780 body is false, the headers that are returned for a non-SMTP message
4781 include only those that have been read before the error was detected.
4782 (In the case of an ACL rejection, they have all been read.)
4784 PH/69 The HTML version of the specification is now built in a directory called
4785 spec_html instead of spec.html, because the latter looks like a path with
4786 a MIME-type, and this confuses some software.
4788 PH/70 Catch two compiler warnings in sieve.c.
4790 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4791 function verify_get_ident() calls ip_connect() to connect a socket, but
4792 if the "connect()" function timed out, ip_connect() used to close the
4793 socket. However, verify_get_ident() also closes the socket later, and in
4794 between Exim writes to the log, which may get opened at this point. When
4795 the socket was closed in ip_connect(), the log could get the same file
4796 descriptor number as the socket. This naturally causes chaos. The fix is
4797 not to close the socket in ip_connect(); the socket should be closed by
4798 the function that creates it. There was only one place in the code where
4799 this was missing, in the iplookup router, which I don't think anybody now
4800 uses, but I've fixed it anyway.
4802 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4803 well as to direct DNS lookups. Otherwise the handling of names in host
4804 lists is inconsistent and therefore confusing.
4810 PH/01 Two changes to the default runtime configuration:
4812 (1) Move the checks for relay_from_hosts and authenticated clients from
4813 after to before the (commented out) DNS black list checks.
4815 (2) Add control=submission to the relay_from_hosts and authenticated
4816 clients checks, on the grounds that messages accepted by these
4817 statements are most likely to be submissions.
4819 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4821 (1) Generate an error if the third argument for the ${prvs expansion is
4824 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4827 (3) Reset the variables that are obtained from the first argument of
4828 ${prvscheck and used in the second argument before leaving the code,
4829 because their memory is reclaimed, so using them afterwards may do
4832 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4833 one (it's much easier than Tom thought :-).
4835 (5) Because of (4), we can now allow for the use of $prvscheck_result
4836 inside the third argument.
4838 PH/03 For some reason, the default setting of PATH when running a command from
4839 a pipe transport was just "/usr/bin". I have changed it to
4842 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4843 anything to be listed in the output from -bV.
4845 PH/05 When a filter generated an autoreply, the entire To: header line was
4846 quoted in the delivery log line, like this:
4848 => >A.N.Other <ano@some.domain> <original@ddress> ...
4850 This has been changed so that it extracts the operative address. There
4851 may be more than one such address. If so, they are comma-separated, like
4854 => >ano@some.domain,ona@other.domain <original@ddress> ...
4856 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4857 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4858 not being looked up in the rDNS to get a host name, Exim was showing the
4859 IP address twice in Received: lines, even though the IP addresses were
4860 identical. For example:
4862 Received: from [1.2.3.4] (helo=[1.2.3.4])
4864 However, if the real host name was known, it was omitting the HELO data
4865 if it matched the actual IP address. This has been tidied up so that it
4866 doesn't show the same IP address twice.
4868 PH/07 When both +timestamp and +memory debugging was on, the value given by
4869 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4870 called by the debug printing, thereby overwriting the timestamp buffer.
4871 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4873 PH/08 When the original message was included in an autoreply transport, it
4874 always said "this is a copy of the message, including all the headers",
4875 even if body_only or headers_only was set. It now gives an appropriate
4878 PH/09 Applied a patch from the Sieve maintainer which:
4880 o fixes some comments
4881 o adds the (disabled) notify extension core
4882 o adds some debug output for the result of if/elsif tests
4883 o points to the current vacation draft in the documentation
4884 and documents the missing references header update
4888 o fixes a bug in processing the envelope test (when testing
4889 multiple envelope elements, the last element determined the
4892 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4893 Electronic Mail") by including:
4895 Auto-submitted: auto-generated
4897 in the messages that it generates (bounce messages and others, such as
4898 warnings). In the case of bounce messages for non-SMTP messages, there was
4899 also a typo: it was using "Auto_submitted" (underscore instead of
4900 hyphen). Since every message generated by Exim is necessarily in response
4901 to another message, thes have all been changed to:
4903 Auto-Submitted: auto-replied
4905 in accordance with these statements in the RFC:
4907 The auto-replied keyword:
4909 - SHOULD be used on messages sent in direct response to another
4910 message by an automatic process,
4912 - MUST NOT be used on manually-generated messages,
4914 - MAY be used on Delivery Status Notifications (DSNs) and Message
4915 Disposition Notifications (MDNs),
4917 - MUST NOT be used on messages generated by automatic or periodic
4918 processes, except for messages which are automatic responses to
4921 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4922 to the default Received: header definition.
4924 PH/12 Added log selector acl_warn_skipped (default on).
4926 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4927 variables because (a) they are in the wrong storage pool and (b) even if
4928 they were copied, it wouldn't work properly because of the caching.
4930 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4931 checking when decoding. Apparently there are clients that generate
4932 overlong encoded strings. Why am I not surprised?
4934 PH/15 If the first argument of "${if match_address" was not empty, but did not
4935 contain an "@" character, Exim crashed. Now it writes a panic log message
4936 and treats the condition as false.
4938 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4940 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4941 extension "envelope-auth". The code is finished and in agreement with
4942 other implementations, but there is no documentation so far and in fact,
4943 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4944 not changing the active code.
4946 Print executed "if" and "elsif" statements when debugging is used. This
4947 helps a great deal to understand what a filter does.
4949 Document more things not specified clearly in RFC3028. I had all this
4950 sorted out, when out of a sudden new issues came to my mind. Oops."
4952 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4955 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4956 canonical form (as documented). However, after a host name lookup from
4957 the IP address, check_host() was doing a simple string comparison with
4958 addresses acquired from the DNS when checking that the found name did
4959 have the original IP as one of its addresses. Since any found IPv6
4960 addresses are likely to be in abbreviated form, the comparison could
4961 fail. Luckily, there already exists a function for doing the comparison
4962 by converting both addresses to binary, so now that is used instead of
4963 the text comparison.
4965 PH/20 There was another similar case to PH/19, when a complete host name was
4966 given in a host list; looking up its IP address could give an abbreviated
4967 form, whereas the current host's name might or might not be abbreviated.
4968 The same fix has been applied.
4974 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4975 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4978 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4979 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4981 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4983 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4984 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4985 Also, for "Auto-Submitted", treat anything other than "no" as
4986 non-personal, in accordance with RFC 3834. (Previously it treated
4987 anything starting "auto-" as non-personal.)
4989 TF/01 The control=submission/name=... option had a problem with syntax
4990 errors if the name included a slash character. The /name= option
4991 now slurps the rest of the string, so it can include any characters
4992 but it must come last in the list of options (after /sender_retain
4995 PH/05 Some modifications to the interface to the fake nameserver for the new
5003 TK/01 Added the "success_on_redirect" address verification option. See
5004 NewStuff for rationale and an example.
5006 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5008 PH/02 Patch to exigrep to allow it to work on syslog lines.
5010 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5011 fread() to skip over the body file's header line, because in Cygwin the
5012 header line is locked and is inaccessible.
5014 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5015 co-exist for some time) to make it clear that it is the Exim ID that is
5016 referenced, not the Message-ID: header line.
5018 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5019 string_format() function, because snprintf() does not exist on all
5022 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5023 ${stat: expansion item.
5025 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5026 protocol synchronization error", to keep the pedants happy.
5028 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5029 well as for IRIX systems, when gcc is being used. See the host.c source
5032 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5034 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5037 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5038 transport and to the smtp transport in LMTP mode.
5040 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5042 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5043 run a filter in a subprocess. This could lead to confusion in subsequent
5044 lookups in the parent process. There should also be a search_tidyup() at
5045 the end of the subprocess.
5047 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5048 only if the host matched helo_try_verify_hosts, which caused the
5049 verification to occur when the EHLO/HELO command was issued. The ACL just
5050 tested the remembered result. Now, if a previous verification attempt has
5051 not happened, "verify = helo" does it there and then.
5053 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5055 TK/03 Fix log output including CR from clamd.
5057 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5058 exist provoked a memory error which could cause a segfault.
5060 PH/15 Installed PCRE 6.2
5062 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5064 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5065 of the problem. Specifically, suggested +O2 rather than +O1 for the
5068 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5070 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5071 record, the port was not correctly logged when the outgoing_port log
5072 selector was set (it logged the transort's default port).
5074 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5075 fallback_hosts, and "hosts" in the smtp transport.
5077 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5078 host errors such as "Connection refused".
5080 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5081 authentication with radiusclient 0.4.9:
5083 - Error returned from rc_read_config was caught wrongly
5084 - Username/password not passed on to radius server due to wrong length.
5086 The presumption is that some radiusclient API changes for 4.51/PH/17
5087 were not taken care of correctly. The code is still untested by me (my
5088 Linux distribution still has 0.3.2 of radiusclient), but it was
5089 contributed by a Radius user.
5091 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5092 expanding the "port" option of the smtp transport.
5094 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5095 while reading a MIME header. Thanks to Tom Hughes for a patch.
5097 PH/24 Include config.h inside local_scan.h so that configuration settings are
5100 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5101 that in an ACL for RCPT (for example), you can examine exactly what was
5104 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5105 commands, but it was not correctly comparing the address with the actual
5106 client host address. Thus, it would show the EHLO address in Received:
5107 header lines when this was not necessary.
5109 PH/27 Added the % operator to ${eval:}.
5111 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5112 it should be ignoring failures (because with -C, for example, it has lost
5113 privilege). It wasn't ignoring creation failures other than "already
5116 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5119 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5120 because an input line was too long, either on its own, or by virtue of
5121 too many continuations, the temporary file was not being removed, and the
5122 return code was incorrect.
5124 PH/31 Missing "BOOL" in function definition in filtertest.c.
5126 PH/32 Applied Sieve patches from the maintainer.
5128 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5130 PH/33 Added "verify = not_blind".
5132 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5133 Local/Makefile (with some defaults set). These are used in built scripts
5134 such as exicyclog, but they have never been used in the exim_install
5135 script (though there are many overriding facilities there). I have
5136 arranged that the exim_install script now takes note of these two
5139 PH/35 Installed configuration files for Dragonfly.
5141 PH/36 When a locally submitted message by a trusted user did not contain a
5142 From: header, and the sender address was obtained from -f or from an SMTP
5143 MAIL command, and the trusted user did not use -F to supply a sender
5144 name, $originator_name was incorrectly used when constructing a From:
5145 header. Furthermore, $originator_name was used for submission mode
5146 messages from external hosts without From: headers in a similar way,
5147 which is clearly wrong.
5149 PH/37 Added control=suppress_local_fixups.
5151 PH/38 When log_selector = +received_sender was set, and the addition of the
5152 sender made the log line's construction buffer exactly full, or one byte
5153 less than full, an overflow happened when the terminating "\n" was
5156 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5157 when the result of a list match is failure because a DNS lookup failed.
5159 PH/40 RM_COMMAND is now used in the building process.
5161 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5162 the "build-* directories that it finds.
5164 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5165 address in a domain literal was a prefix of an interface address.
5167 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5168 when verifying a sender address, unless rewrite_headers is false.
5170 PH/44 Wrote a long comment about why errors_to addresses are verified as
5171 recipients, not senders.
5173 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5174 the ratelimit ACL was added.
5176 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5178 PH/46 Added extra information about PostgreSQL errors to the error string.
5180 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5181 avoiding the need to install special zones in a real server. This is
5182 backwards compatible; if it can't find the fake resolver, it drops back.
5183 Thus, both old and new test suites can be run.
5185 TF/02 Added util/ratelimit.pl
5187 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5188 clock is set back in time.
5190 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5191 Candler <B.Candler@pobox.com>.
5193 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5194 OK for addresses that are the result of redirections.
5196 PH/48 A number of further additions for the benefit of the new test suite,
5197 including a fake gethostbyname() that interfaces to the fake DNS resolver
5200 TF/06 The fix for widen_domains has also been applied to qualify_single and
5201 search_parents which are the other dnslookup options that can cause
5204 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5207 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5209 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5210 service" if it encountered an MX record with an empty target hostname.
5211 The message is now "an MX or SRV record indicated no SMTP service".
5213 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5214 if the DNS of the sending site is misconfigured. This is quite a
5215 common situation. This change restores the behaviour of treating a
5216 helo verification defer as a failure.
5218 PH/51 If self=fail was set on a router, the bounce message did not include the
5219 actual error message.
5225 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5227 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5228 command itself ended in error, the underlying message about the transport
5229 filter timeout was being overwritten with the pipe command error. Now the
5230 underlying error message should be appended to the second error message.
5232 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5234 PH/02 Reduce the amount of output that "make" produces by default. Full output
5235 can still be requested.
5237 PH/03 The warning log line about a condition test deferring for a "warn" verb
5238 was being output only once per connection, rather than after each
5239 occurrence (because it was using the same function as for successful
5240 "warn" verbs). This seems wrong, so I have changed it.
5242 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5243 it should not have, which might have caused a crash in the right
5244 circumstances, but probably never did.
5246 PH/04 Installed a modified version of Tony Finch's patch to make submission
5247 mode fix the return path as well as the Sender: header line, and to
5248 add a /name= option so that you can make the user's friendly name appear
5251 TF/03 Added the control = fakedefer ACL modifier.
5253 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5254 Mark Lowes for thorough testing.
5256 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5258 TK/03 Merged latest SRS patch from Miles Wilton.
5260 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5261 with the definition in sysexits.h (which is #included earlier).
5262 Fortunately, Exim does not actually use EX_OK. The code used to try to
5263 preserve the sysexits.h value, by assuming that macro definitions were
5264 scanned for macro replacements. I have been disabused of this notion,
5265 so now the code just undefines EX_OK before #including unistd.h.
5267 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5268 in the smtp transport. When a block could not be written in a single
5269 write() function, the timeout was being re-applied to each part-write.
5270 This seems wrong - if the receiver was accepting one byte at a time it
5271 would take for ever. The timeout is now adjusted when this happens. It
5272 doesn't have to be particularly precise.
5274 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5275 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5277 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5278 without a domain if the check to <postmaster@domain> fails.
5280 SC/01 Eximstats: added -xls and the ability to specify output files
5281 (patch written by Frank Heydlauf).
5283 SC/02 Eximstats: use FileHandles for outputting results.
5285 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5287 SC/04 Eximstats: fixed display of large numbers with -nvr option
5289 SC/05 Eximstats: fixed merging of reports with empty tables.
5291 SC/06 Eximstats: added the -include_original_destination flag
5293 SC/07 Eximstats: removed tabs and trailing whitespace.
5295 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5297 TK/06 MBOX spool code: Add real "From " MBOX separator line
5298 so the .eml file is really in mbox format (even though
5299 most programs do not really care). Patch from Alex Miller.
5301 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5302 The latter is generated from $received_to and is only set if the
5303 message has one envelope recipient. SA can use these headers,
5304 obviously out-of-the-box. Patch from Alex Miller.
5306 PH/08 The ${def test on a variable was returning false if the variable's
5307 value was "0", contrary to what the specification has always said!
5308 The result should be true unless the variable is empty.
5310 PH/09 The syntax error of a character other than { following "${if
5311 def:variable_name" (after optional whitespace) was not being diagnosed.
5312 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5313 accidental colon was present, for example, could give incorrect results.
5315 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5316 result is used (not including appendfile, where other changes are about
5319 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5320 This involved changing a lot of size variables from int to off_t. It
5321 should work with maildirs and everything.
5323 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5324 spamd dying while we are connected to it.
5326 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5329 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5330 function for BDB 4.3.
5332 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5334 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5335 the sources for PCRE have been re-arranged and more files are now
5338 PH/15 The code I had for printing potentially long long variables in PH/11
5339 above was not the best (it lost precision). The length of off_t variables
5340 is now inspected at build time, and an appropriate printing format (%ld
5341 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5342 to be "long long int" or "long int". This is needed for the internal
5343 formatting function string_vformat().
5345 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5346 the configuration file to be ":syslog", then the script "guesses" where
5347 the logs files are, rather than using the compiled in default. In our
5348 case the guess is not the same as the compiled default, so the script
5349 suddenly stopped working when I started to use syslog. The patch checks
5350 to see if log_file_path is "". If so, it attempts to read it from exim
5351 with no configuration file to get the compiled in version, before it
5352 falls back to the previous guessing code."
5354 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5355 implementing BATV in an Exim configuration. See NewStuff for the gory
5358 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5359 Makefile that are specific to HP-UX.
5361 PH/18 If the "use_postmaster" option was set for a recipient callout together
5362 with the "random" option, the postmaster address was used as the MAIL
5363 FROM address for the random test, but not for the subsequent recipient
5364 test. It is now used for both.
5366 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5367 patch removes a few documentation additions to RFC 3028, because the
5368 latest draft now contains them. It adds the new en;ascii-case comparator
5369 and a new error check for 8bit text in MIME parts. Comparator and
5370 require names are now matched exactly. I enabled the subaddress
5371 extension, but it is not well tested yet (read: it works for me)."
5373 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5374 rework some of the code of TK/09 above to avoid the hardwired use of
5375 "%lld" and "long long". Replaced the call to snprintf() with a call to
5378 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5379 records point to non-existent hosts", "retry timeout exceeded", and
5380 "retry time not reached for any host after a long failure period".
5382 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5383 experimental DomainKeys support:
5385 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5386 (2) On an error such as an illegally used "control", the wrong name for
5387 the control was given.
5389 These problems did NOT occur unless DomainKeys support was compiled.
5391 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5393 PH/24 Added ${if match_ip condition.
5395 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5396 databases so that it will be absolutely obvious if a crash occurs in the
5397 DB library. This is a regular occurrence (often caused by mis-matched
5400 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5401 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5402 picked up on a user's system that detects such things. There doesn't seem
5403 to be a gcc warning option for this - only an attribute that has to be
5404 put on the function's prototype. It seems that in Fedora Core 4 they have
5405 set this on a number of new functions. No doubt there will be more in due
5408 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5409 specify a transport. However, if an address that was verified by such a
5410 router was the subject of a callout, Exim crashed because it tried to
5411 read the rcpt_include_affixes from the non-existent transport. Now it
5412 just assumes that the setting of that option is false. This bug was
5413 introduced by 4.51/PH/31.
5415 PH/28 Changed -d+all to exclude +memory, because that information is very
5416 rarely of interest, but it makes the output a lot bigger. People tend to
5417 do -d+all out of habit.
5419 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5420 code in os-type was giving problems when libc.so lives in lib64, like on
5423 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5424 aren't the modern standard, and it seems that some systems' include files
5425 don't always have them. Exim was already checking for some of the newer
5426 ones like T_AAAA, and defining it itself. I've added checks for all the
5427 record types that Exim uses.
5429 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5430 not automatically generating a new one, as it is supposed to. This
5431 prevented TLS from working. If the file did exist, but contained invalid
5432 data, a new version was generated, as expected. It was only the case of a
5433 non-existent file that was broken.
5435 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5436 with a change in libdomainkeys > 0.64.
5438 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5439 from DNS. If the selector record carries the flag, it now has
5440 precedence over the domain-wide flag.
5442 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5444 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5445 the use of a port name that isn't defined in /etc/services) occurred, the
5446 message was deferred as in a normal delivery, and thus remained on the
5447 spool, instead of being failed because of the mua_wrapper setting. This
5448 is now fixed, and I tidied up some of the mua_wrapper messages at the
5451 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5452 the messages in a hash of arrays rather than using individual hashes.
5453 This is a bit cleaner and results in dramatic memory savings, albeit
5454 at a slight CPU cost.
5456 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5457 as requested by Marc Sherman.
5459 SC/10 Eximstats: added histograms for user specified patterns as requested
5462 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5464 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5465 fopen() in the content-scanning modules that did not already have it.
5471 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5472 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5474 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5476 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5478 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5479 to test Sieve filters that use "vacation".
5481 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5482 that changes the way the GnuTLS parameters are stored in the cache file.
5483 The new format can be generated externally. For backward compatibility,
5484 if the data in the cache doesn't make sense, Exim assumes it has read an
5485 old-format file, and it generates new data and writes a new file. This
5486 means that you can't go back to an older release without removing the
5489 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5490 work if there are any delivery delays because "one_time" forces the
5491 parent to be marked "delivered", so its unseen clone is never tried
5492 again. For this reason, Exim now forbids the simultaneous setting of
5495 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5496 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5497 entry if you want to know the details. The fix, however, neglected to
5498 consider the case when local delivery batching is involved. The test for
5499 "previously delivered" was not happening when checking to see if an
5500 address could be batched with a previous (undelivered) one; under
5501 certain circumstances this could lead to multiple deliveries to the same
5504 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5505 in its include files, and this causes problems building Exim.
5507 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5508 header_syntax) but Exim was just ignoring anything given after a slash.
5509 In particular, this caused confusion with an attempt to use "verify =
5510 reverse_host_lookup/defer_ok". An error is now given when options are
5511 supplied for verify items that do not have them. (Maybe reverse_host_
5512 lookup should have a defer_ok option, but that's a different point.)
5514 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5515 defined by RFC 821) to 2048, because there were problems with some AUTH
5516 commands, and RFC 1869 says the size should be increased for extended
5517 SMTP commands that take arguments.
5519 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5522 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5523 "unknown" error; now it says that the functionality isn't in the binary.
5525 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5526 an address' error message when a string expansion fails (syntax or
5527 whatever). Otherwise the password may appear in the log. Following change
5528 PH/42 below, there is no longer a chance of it appearing in a bounce
5531 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5533 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5534 of its addresses was ever tried. (Bugzilla bug #2.)
5536 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5537 the result incorrectly in the debug output. (It correctly added a newline
5538 to what was transported.)
5540 TF/01 Added $received_time.
5542 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5543 commented out examples of how to interface to a virus scanner and to
5544 SpamAssassin. Also added commented examples of av_scanner and
5545 spamd_address settings.
5547 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5548 and controls are allowed in which ACLs. There were a couple of minor
5549 errors. Some of the entries in the conditions table (which is a table of
5550 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5551 negation of where the condition IS allowed.
5553 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5555 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5556 header file does not have a version number, so I've had to invent a new
5557 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5558 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5559 radiusclient), but it was contributed by a Radius user.
5561 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5562 files or directories, for OpenSSL.
5564 PH/19 When an Exim process that is running as root has to create an Exim log
5565 file, it does so in a subprocess that runs as exim:exim so as to get the
5566 ownership right at creation (otherwise, other Exim processes might see
5567 the file with the wrong ownership). There was no test for failure of this
5568 fork() call, which would lead to the process getting stuck as it waited
5569 for a non-existent subprocess. Forks do occasionally fail when resources
5570 run out. I reviewed all the other calls to fork(); they all seem to check
5573 PH/20 When checking for unexpected SMTP input at connect time (before writing
5574 the banner), Exim was not dealing correctly with a non-positive return
5575 from the read() function. If the client had disconnected by this time,
5576 the result was a log entry for a synchronization error with an empty
5577 string after "input=" when read() returned zero. If read() returned -1
5578 (an event I could not check), uninitialized data bytes were printed.
5579 There were reports of junk text (parts of files, etc) appearing after
5582 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5584 PH/22 Added support for macro redefinition, and (re)definition in between
5585 driver and ACL definitions.
5587 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5588 forgetting to use the resulting value; it was using the unexpanded value.
5590 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5591 hadn't been configured. The fix is from Juergen Kreileder, who
5592 understands it better than I do:
5594 "Here's what I see happening with three configured cyrus_sasl
5595 authenticators configured (plain, login, cram-md5):
5597 On startup auth_cyrus_sasl_init() gets called for each of these.
5598 This means three calls to sasl_listmech() without a specified mech_list.
5599 => SASL tests which mechs of all available mechs actually work
5600 => three warnings about OTP not working
5601 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5603 With the patch, sasl_listmech() also gets called three times. But now
5604 SASL's mech_list option is set to the server_mech specified in the the
5605 authenticator. Or in other words, the answer from sasl_listmech()
5606 gets limited to just the mech you're testing for (which is different
5608 => the return list contains just 'plain' or 'login', 'cram-md5' or
5609 nothing depending on the value of ob->server_mech.
5611 I've just tested the patch: Authentication still works fine,
5612 unavailable mechs specified in the exim configuration are still
5613 caught, and the auth.log warnings about OTP are gone."
5615 PH/25 When debugging is enabled, the contents of the command line are added
5616 to the debugging output, even when log_selector=+arguments is not
5619 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5620 answer is "GNU", and only if the return is "GNU/something" is the answer
5623 PH/27 $acl_verify_message is now set immediately after the failure of a
5624 verification in an ACL, and so is available in subsequent modifiers. In
5625 particular, the message can be preserved by coding like this:
5627 warn !verify = sender
5628 set acl_m0 = $acl_verify_message
5630 Previously, $acl_verify_message was set only while expanding "message"
5631 and "log_message" when a very denied access.
5633 PH/28 Modified OS/os.c-Linux with
5635 -#ifndef OS_LOAD_AVERAGE
5636 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5638 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5639 nomenclature these days.)
5641 PH/29 Installed patch from the Sieve maintainer that adds the options
5642 sieve_useraddress and sieve_subaddress to the redirect router.
5644 PH/30 In these circumstances:
5645 . Two addresses routed to the same list of hosts;
5646 . First host does not offer TLS;
5647 . First host accepts first address;
5648 . First host gives temporary error to second address;
5649 . Second host offers TLS and a TLS session is established;
5650 . Second host accepts second address.
5651 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5652 and peerdn, if requested) that were in fact used only for the second
5655 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5656 attention to any local part prefix or suffix that was matched by the
5657 router that accepted the address. It now behaves in the same way as it
5658 does for delivery: the affixes are removed from the local part unless
5659 rcpt_include_affixes is set on the transport.
5661 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5662 timeout during the DATA phase of an incoming message.
5664 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5665 applied a patch sanctioned by the Sieve maintainer.
5667 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5668 the uid or gid is negative. A case of a negative gid caused this to be
5669 noticed. The fix allows for either to be negative.
5671 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5672 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5675 PH/36 The change PH/12 above was broken. Fixed it.
5677 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5678 the grounds that routing the same address twice would always produce the
5679 same answer. This might have been true once, but it is certainly no
5680 longer true now. Routing a child address may depend on the previous
5681 routing that produced that child. Some complicated redirection strategies
5682 went wrong when messages had multiple recipients, and made Exim's
5683 behaviour dependent on the order in which the addresses were given.
5685 I have moved the duplicate checking until after the routing is complete.
5686 Exim scans the addresses that are assigned to local and remote
5687 transports, and removes any duplicates. This means that more work will be
5688 done, as duplicates will always all be routed, but duplicates are
5689 presumably rare, so I don't expect this is of any significance.
5691 For deliveries to pipes, files, and autoreplies, the duplicate checking
5692 still happens during the routing process, since they are not going to be
5695 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5696 It corrects a timeout issue with spamd. This is Ian's comment: "The
5697 background is that sometimes spamd either never reads data from a
5698 connection it has accepted, or it never writes response data. The exiscan
5699 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5700 blindly assumes that writes won't block so it may never time out."
5702 PH/39 Allow G after quota size as well as K and M.
5704 PH/40 The value set for $authenticated_id in an authenticator may not contain
5705 binary zeroes or newlines because the value is written to log lines and
5706 to spool files. There was no check on this. Now the value is run through
5707 the string_printing() function so that such characters are converted to
5708 printable escape sequences.
5710 PH/41 $message_linecount is a new variable that contains the total number of
5711 lines in the message. Compare $body_linecount, which is the count for the
5714 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5715 bounce and delay warning messages, except in certain special cases, which
5718 (a) An SMTP error message from a remote host;
5719 (b) A message specified in a :fail: redirection;
5720 (c) A message specified in a "fail" command in a system filter;
5721 (d) A message specified in a FAIL return from the queryprogram router;
5722 (e) A message specified by the cannot_route_message router option.
5724 In these cases only, Exim does include the error details in bounce and
5725 warning messages. There are also a few cases where bland messages such
5726 as "unrouteable address" or "local delivery error" are given.
5728 PH/43 $value is now also set for the "else" part of a ${run expansion.
5730 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5731 being worked on, but at least Exim now implements the latest version to
5734 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5735 process to complete was treated as a delivery failure, a timeout while
5736 writing the message to the pipe was logged, but erroneously treated as a
5737 successful delivery. Such timeouts include transport filter timeouts. For
5738 consistency with the overall process timeout, these timeouts are now
5739 treated as errors, giving rise to delivery failures by default. However,
5740 there is now a new Boolean option for the pipe transport called
5741 timeout_defer, which, if set TRUE, converts the failures into defers for
5742 both kinds of timeout. A transport filter timeout is now identified in
5745 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5746 systems where "make" and "gmake" are different, calling "gmake" at top
5747 level broke things. I've arranged for the value of $(MAKE) to be passed
5748 from the Makefile to this script so that it can call the same version of
5752 A note about Exim versions 4.44 and 4.50
5753 ----------------------------------------
5755 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5756 changes of various kinds. As a consequence, a big documentation update was
5757 needed. This delayed the release for rather longer than seemed good, especially
5758 in the light of a couple of (minor) security issues. Therefore, the changes
5759 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5760 release. So 4.44 and 4.50 are in effect two different branches that both start
5763 I have left the 4.50 change log unchanged; it contains all the changes since
5764 4.43. The change log for 4.44 is below; many of its items are identical to
5765 those for 4.50. This seems to be the most sensible way to preserve the
5766 historical information.
5772 1. Minor wording change to the doc/README.SIEVE file.
5774 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5775 computation of the current number of files was incorrect.
5777 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5778 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5781 4. Give more explanation in the error message when the command for a transport
5782 filter fails to execute.
5784 5. There are several places where Exim runs a non-Exim command in a
5785 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5786 was being done only for the command run by the queryprogram router. It is
5787 now done for all such subprocesses. The other cases are: ${run, transport
5788 filters, and the commands run by the lmtp and pipe transports.
5790 6. Added CONFIGURE_GROUP build-time option.
5792 7. Some older OS have a limit of 256 on the maximum number of file
5793 descriptors. Exim was using setrlimit() to set 1000 as a large value
5794 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5795 systems. I've change it so that if it can't get 1000, it tries for 256.
5797 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5798 was an oversight, and furthermore, ever since the addition of extra
5799 controls (e.g. 4.43/32), the checks on when to allow different forms of
5800 "control" were broken. There should now be diagnostics for all cases when a
5801 control that does not make sense is encountered.
5803 9. Added the /retain_sender option to "control=submission".
5805 10. $recipients is now available in the predata ACL (oversight).
5807 11. Tidy the search cache before the fork to do a delivery from a message
5808 received from the command line. Otherwise the child will trigger a lookup
5809 failure and thereby defer the delivery if it tries to use (for example) a
5810 cached ldap connection that the parent has called unbind on.
5812 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5813 of $address_data from the recipient verification was clobbered by the
5814 sender verification.
5816 13. The value of address_data from a sender verification is now available in
5817 $sender_address_data in subsequent conditions in the ACL statement.
5819 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5821 15. Added a new option "connect=<time>" to callout options, to set a different
5824 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5825 was its contents. (It was OK if the option was not defined at all.)
5827 17. A "Completed" log line is now written for messages that are removed from
5828 the spool by the -Mrm option.
5830 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5831 information about exactly what failed.
5833 19. Added -dd to debug only the daemon process.
5835 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5836 handles timeouts, both on the server side and network timeouts. Renamed the
5837 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5839 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5840 It is now set to "smtps".
5842 22. $host_address is now set to the target address during the checking of
5843 ignore_target_hosts.
5845 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5846 being passed; this would have caused $sender_host_name to have been used if
5847 matching the list had actually called for a host name (not very likely,
5848 since this list is usually IP addresses). A host name is now passed as
5851 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5852 code that specifies a non-restarting handler (typically sigaction() in
5853 modern systems) in an attempt to fix a rare and obscure crash bug.
5855 25. Narrowed the window for a race in the daemon that could cause it to ignore
5856 SIGCHLD signals. This is not a major problem, because they are used only to
5857 wake it up if nothing else does.
5859 26. A malformed maildirsize file could cause Exim to calculate negative values
5860 for the mailbox size or file count. Odd effects could occur as a result.
5861 The maildirsize information is now recalculated if the size or filecount
5864 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5865 support for a long time. Removed HAVE_SYS_VFS_H.
5867 28. Installed the latest version of exipick from John Jetmore.
5869 29. In an address list, if the pattern was not a regular expression, an empty
5870 subject address (from a bounce message) matched only if the pattern was an
5871 empty string. Non-empty patterns were not even tested. This was the wrong
5872 because it is perfectly reasonable to use an empty address as part of a
5873 database query. An empty address is now tested by patterns that are
5874 lookups. However, all the other forms of pattern expect the subject to
5875 contain a local part and a domain, and therefore, for them, an empty
5876 address still always fails if the pattern is not itself empty.
5878 30. Exim went into a mad DNS loop when attempting to do a callout where the
5879 host was specified on an smtp transport, and looking it up yielded more
5880 than one IP address.
5882 31. Re-factored the code for checking spool and log partition space into a
5883 function that finds that data and another that does the check. The former
5884 is then used to implement four new variables: $spool_space, $log_space,
5885 $spool_inodes, and $log_inodes.
5887 32. The RFC2047 encoding function was originally intended for short strings
5888 such as real names; it was not keeping to the 75-character limit for
5889 encoded words that the RFC imposes. It now respects the limit, and
5890 generates multiple encoded words if necessary. To be on the safe side, I
5891 have increased the buffer size for the ${rfc2047: expansion operator from
5894 33. It is now permitted to omit both strings after an "if" condition; if the
5895 condition is true, the result is "true". As before, when the second string
5896 is omitted, a false condition yields an empty string. This makes it less
5897 cumbersome to write custom ACL and router conditions.
5899 34. Failure to deliver a bounce message always caused it to be frozen, even if
5900 there was an errors_to setting on the router. The errors_to setting is now
5903 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5904 canonical form (fully expanded) before being placed in
5905 $sender_host_address.
5907 36. The table in the code that translates DNS record types into text (T_A to
5908 "A" for instance) was missing entries for NS and CNAME. It is just possible
5909 that this could have caused confusion if both these types were looked up
5910 for the same domain, because the text type is used as part of Exim's
5911 per-process caching. But the chance of anyone hitting this buglet seems
5914 37. The dnsdb lookup has been extended in a number of ways.
5916 (1) There is a new type, "zns", which walks up the domain tree until it
5917 finds some nameserver records. It should be used with care.
5919 (2) There is a new type, "mxh", which is like "mx" except that it returns
5920 just the host names, not the priorities.
5922 (3) It is now possible to give a list of domains (or IP addresses) to be
5923 looked up. The behaviour when one of the lookups defers can be
5924 controlled by a keyword.
5926 (4) It is now possible to specify the separator character for use when
5927 multiple records are returned.
5929 38. The dnslists ACL condition has been extended: it is now possible to supply
5930 a list of IP addresses and/or domains to be looked up in a particular DNS
5933 39. Added log_selector=+queue_time_overall.
5935 40. When running the queue in the test harness, wait just a tad after forking a
5936 delivery process, to get repeatability of debugging output.
5938 41. Include certificate and key file names in error message when GnuTLS fails
5939 to set them up, because the GnuTLS error message doesn't include the name
5940 of the failing file when there is a problem reading it.
5942 42. Allow both -bf and -bF in the same test run.
5944 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5946 44. The "Exiscan patch" is now merged into the mainline Exim source.
5948 45. Sometimes the final signoff response after QUIT could fail to get
5949 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5950 < 0 before doing a fflush(). This bug looks as though it goes back to the
5951 introduction of TLS in release 3.20, but "sometimes" must have been rare
5952 because the tests only now provoked it.
5954 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5955 (this can affect the format of dates).
5957 47. exim_tidydb, when checking for the continued existence of a message for
5958 which it has found a message-specific retry record, was not finding
5959 messages that were in split spool directories. Consequently, it was
5960 deleting retry records that should have stayed in existence.
5962 48. Steve fixed some bugs in eximstats.
5964 49. The SPA authentication driver was not abandoning authentication and moving
5965 on to the next authenticator when an expansion was forced to fail,
5966 contradicting the general specification for all authenticators. Instead it
5967 was generating a temporary error. It now behaves as specified.
5969 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5970 (the order specifies the preference for clients). The order is now AES256,
5971 AES128, 3DES, ARCFOUR128.
5973 51. Small patch to Sieve code - explicitly set From: when generating an
5976 52. Exim crashed if a remote delivery caused a very long error message to be
5977 recorded - for instance if somebody sent an entire SpamAssassin report back
5978 as a large number of 550 error lines. This bug was coincidentally fixed by
5979 increasing the size of one of Exim's internal buffers (big_buffer) that
5980 happened as part of the Exiscan merge. However, to be on the safe side, I
5981 have made the code more robust (and fixed the comments that describe what
5984 53. Now that there can be additional text after "Completed" in log lines (if
5985 the queue_time_overall log selector is set), a one-byte patch to exigrep
5986 was needed to allow it to recognize "Completed" as not the last thing in
5989 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5990 patch that reportedly fixes this has been added. I am not expert enough to
5991 create a test for it. This is what the patch creator wrote:
5993 "I found a little strange behaviour of ldap code when working with
5994 Windows 2003 AD Domain, where users was placed in more than one
5995 Organization Units. When I tried to give exim partial DN, the exit code
5996 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5997 But simultaneously result of request was absolutely normal ldap result,
5998 so I produce this patch..."
6000 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6001 REFERENCE, so I have modified the code to exclude the patch when that macro
6004 55. Some experimental protocols are using DNS PTR records for new purposes. The
6005 keys for these records are domain names, not reversed IP addresses. The
6006 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6007 leaves it alone. Component reversal etc. now happens only for IP addresses.
6010 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6012 57. Double the size of the debug message buffer (to 2048) so that more of very
6013 long debug lines gets shown.
6015 58. The exicyclog utility now does better if the number of log files to keep
6016 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6018 59. Two changes related to the smtp_active_hostname option:
6020 (1) $smtp_active_hostname is now available as a variable.
6021 (2) The default for smtp_banner uses $smtp_active_hostname instead
6022 of $primary_hostname.
6024 60. The host_aton() function is supposed to be passed a string that is known
6025 to be a valid IP address. However, in the case of IPv6 addresses, it was
6026 not checking this. This is a hostage to fortune. Exim now panics and dies
6027 if the condition is not met. A case was found where this could be provoked
6028 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6029 components; fortuitously, this particular loophole had already been fixed
6030 by change 4.50/55 above.
6032 If there are any other similar loopholes, the new check in host_aton()
6033 itself should stop them being exploited. The report I received stated that
6034 data on the command line could provoke the exploit when Exim was running as
6035 exim, but did not say which command line option was involved. All I could
6036 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6037 running as the user.
6040 61. There was a buffer overflow vulnerability in the SPA authentication code
6041 (which came originally from the Samba project). I have added a test to the
6042 spa_base64_to_bits() function which I hope fixes it.
6045 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6046 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6048 63. The daemon start-up calls getloadavg() while still root for those OS that
6049 need the first call to be done as root, but it missed one case: when
6050 deliver_queue_load_max is set with deliver_drop_privilege. This is
6051 necessary for the benefit of the queue runner, because there is no re-exec
6052 when deliver_drop_privilege is set.
6054 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6055 This has been fixed.
6057 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6058 was in use, was not putting the data itself into the right store pool;
6059 consequently, it could be overwritten for a subsequent message in the same
6060 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6063 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6065 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6066 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6067 the function were treating the return as a boolean value, which happened to
6068 work because 0=false and not-0=true, but is not correct code.
6070 68. The host_aton() function was not handling scoped IPv6 addresses (those
6071 with, for example, "%eth0" on the end) correctly.
6073 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6074 negated items (that is, ~something) in unsigned ints. Some compilers
6075 apparently mutter when there is no cast.
6077 70. If an address verification called from an ACL failed, and did not produce a
6078 user-specific message (i.e. there was only a "system" message), nothing was
6079 put in $acl_verify_message. In this situation, it now puts the system
6082 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6083 session; change 4.31/43 added the unwanted input to the log line - except
6084 that it did not do this in the start of session case. It now does.
6086 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6087 This is wrong and can cause the other end to generate a synchronization
6088 error if it is another Exim or anything else that does the synchronization
6089 check. A QUIT command is no longer sent after a timeout.
6091 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6092 during host lookups.
6094 74. The defer_ok option of callout verification was not working if it was used
6095 when verifying addresses in header lines, that is, for this case:
6097 verify = header_sender/callout=defer_ok
6099 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6100 those file descriptors could be used for SMTP connections. If anything
6101 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6102 could be sent to the SMTP client, causing chaos. The daemon now opens
6103 stdin, stdout, and stderr to /dev/null when it puts itself into the
6106 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6107 log by default. The user can override this with suitable Perl magic.
6109 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6110 the log message when discard triggers, was not working for the DATA ACL or
6111 for the non-SMTP ACL.
6113 78. Error message wording change in sieve.c.
6115 79. If smtp_accept_max_per_host was set, the number of connections could be
6116 restricted to fewer than expected, because the daemon was trying to set up
6117 a new connection before checking whether the processes handling previous
6118 connections had finished. The check for completed processes is now done
6119 earlier. On busy systems, this bug wouldn't be noticed because something
6120 else would have woken the daemon, and it would have reaped the completed
6123 80. If a message was submitted locally by a user whose login name contained one
6124 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6125 It caused a spool format error. I have fixed the spool reading code. A
6126 related problem was that the "from" clause in the Received: line became
6127 illegal because of the space(s). It is now covered by ${quote_local_part.
6129 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6132 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6134 83. Patch from Sieve maintainer to fix unterminated string problem in
6135 "vacation" handling.
6137 84. Some minor changes to the Linux configuration files to help with other
6138 OS variants using glibc.
6140 85. One more patch for Sieve to update vacation handling to latest spec.
6143 ----------------------------------------------------
6144 See the note above about the 4.44 and 4.50 releases.
6145 ----------------------------------------------------
6151 1. Change 4.43/35 introduced a bug that caused file counts to be
6152 incorrectly computed when quota_filecount was set in an appendfile
6155 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6156 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6159 3. Give more explanation in the error message when the command for a transport
6160 filter fails to execute.
6162 4. There are several places where Exim runs a non-Exim command in a
6163 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6164 was being done only for the command run by the queryprogram router. It is
6165 now done for all such subprocesses. The other cases are: ${run, transport
6166 filters, and the commands run by the lmtp and pipe transports.
6168 5. Some older OS have a limit of 256 on the maximum number of file
6169 descriptors. Exim was using setrlimit() to set 1000 as a large value
6170 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6171 systems. I've change it so that if it can't get 1000, it tries for 256.
6173 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6174 was an oversight, and furthermore, ever since the addition of extra
6175 controls (e.g. 4.43/32), the checks on when to allow different forms of
6176 "control" were broken. There should now be diagnostics for all cases when a
6177 control that does not make sense is encountered.
6179 7. $recipients is now available in the predata ACL (oversight).
6181 8. Tidy the search cache before the fork to do a delivery from a message
6182 received from the command line. Otherwise the child will trigger a lookup
6183 failure and thereby defer the delivery if it tries to use (for example) a
6184 cached ldap connection that the parent has called unbind on.
6186 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6187 of $address_data from the recipient verification was clobbered by the
6188 sender verification.
6190 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6191 was its contents. (It was OK if the option was not defined at all.)
6193 11. A "Completed" log line is now written for messages that are removed from
6194 the spool by the -Mrm option.
6196 12. $host_address is now set to the target address during the checking of
6197 ignore_target_hosts.
6199 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6200 being passed; this would have caused $sender_host_name to have been used if
6201 matching the list had actually called for a host name (not very likely,
6202 since this list is usually IP addresses). A host name is now passed as
6205 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6206 code that specifies a non-restarting handler (typically sigaction() in
6207 modern systems) in an attempt to fix a rare and obscure crash bug.
6209 15. Narrowed the window for a race in the daemon that could cause it to ignore
6210 SIGCHLD signals. This is not a major problem, because they are used only to
6211 wake it up if nothing else does.
6213 16. A malformed maildirsize file could cause Exim to calculate negative values
6214 for the mailbox size or file count. Odd effects could occur as a result.
6215 The maildirsize information is now recalculated if the size or filecount
6218 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6219 support for a long time. Removed HAVE_SYS_VFS_H.
6221 18. Updated exipick to current release from John Jetmore.
6223 19. Allow an empty sender to be matched against a lookup in an address list.
6224 Previously the only cases considered were a regular expression, or an
6227 20. Exim went into a mad DNS lookup loop when doing a callout where the
6228 host was specified on the transport, if the DNS lookup yielded more than
6231 21. The RFC2047 encoding function was originally intended for short strings
6232 such as real names; it was not keeping to the 75-character limit for
6233 encoded words that the RFC imposes. It now respects the limit, and
6234 generates multiple encoded words if necessary. To be on the safe side, I
6235 have increased the buffer size for the ${rfc2047: expansion operator from
6238 22. Failure to deliver a bounce message always caused it to be frozen, even if
6239 there was an errors_to setting on the router. The errors_to setting is now
6242 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6243 canonical form (fully expanded) before being placed in
6244 $sender_host_address.
6246 24. Updated eximstats to version 1.33
6248 25. Include certificate and key file names in error message when GnuTLS fails
6249 to set them up, because the GnuTLS error message doesn't include the name
6250 of the failing file when there is a problem reading it.
6252 26. Expand error message when OpenSSL has problems setting up cert/key files.
6255 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6256 (this can affect the format of dates).
6258 28. exim_tidydb, when checking for the continued existence of a message for
6259 which it has found a message-specific retry record, was not finding
6260 messages that were in split spool directories. Consequently, it was
6261 deleting retry records that should have stayed in existence.
6263 29. eximstats updated to version 1.35
6264 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6265 1.35 - bugfix such that pie charts by volume are generated correctly
6267 30. The SPA authentication driver was not abandoning authentication and moving
6268 on to the next authenticator when an expansion was forced to fail,
6269 contradicting the general specification for all authenticators. Instead it
6270 was generating a temporary error. It now behaves as specified.
6272 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6273 (the order specifies the preference for clients). The order is now AES256,
6274 AES128, 3DES, ARCFOUR128.
6276 31. Small patch to Sieve code - explicitly set From: when generating an
6279 32. Exim crashed if a remote delivery caused a very long error message to be
6280 recorded - for instance if somebody sent an entire SpamAssassin report back
6281 as a large number of 550 error lines. This bug was coincidentally fixed by
6282 increasing the size of one of Exim's internal buffers (big_buffer) that
6283 happened as part of the Exiscan merge. However, to be on the safe side, I
6284 have made the code more robust (and fixed the comments that describe what
6287 33. Some experimental protocols are using DNS PTR records for new purposes. The
6288 keys for these records are domain names, not reversed IP addresses. The
6289 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6290 leaves it alone. Component reversal etc. now happens only for IP addresses.
6293 34. The host_aton() function is supposed to be passed a string that is known
6294 to be a valid IP address. However, in the case of IPv6 addresses, it was
6295 not checking this. This is a hostage to fortune. Exim now panics and dies
6296 if the condition is not met. A case was found where this could be provoked
6297 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6298 components; fortuitously, this particular loophole had already been fixed
6299 by change 4.50/55 or 4.44/33 above.
6301 If there are any other similar loopholes, the new check in host_aton()
6302 itself should stop them being exploited. The report I received stated that
6303 data on the command line could provoke the exploit when Exim was running as
6304 exim, but did not say which command line option was involved. All I could
6305 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6306 running as the user.
6309 35. There was a buffer overflow vulnerability in the SPA authentication code
6310 (which came originally from the Samba project). I have added a test to the
6311 spa_base64_to_bits() function which I hope fixes it.
6314 36. The daemon start-up calls getloadavg() while still root for those OS that
6315 need the first call to be done as root, but it missed one case: when
6316 deliver_queue_load_max is set with deliver_drop_privilege. This is
6317 necessary for the benefit of the queue runner, because there is no re-exec
6318 when deliver_drop_privilege is set.
6320 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6321 was in use, was not putting the data itself into the right store pool;
6322 consequently, it could be overwritten for a subsequent message in the same
6323 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6326 38. Sometimes the final signoff response after QUIT could fail to get
6327 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6328 < 0 before doing a fflush(). This bug looks as though it goes back to the
6329 introduction of TLS in release 3.20, but "sometimes" must have been rare
6330 because the tests only now provoked it.
6336 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6337 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6338 Now it returns an integer. A number of calls were still expecting a T/F
6339 return. Fortuitously, in all cases, the tests worked in OK situations,
6340 which is the norm. However, things would have gone wrong on any write
6341 failures on the smtp file descriptor. This function is used when sending
6342 messages over SMTP and also when doing verify callouts.
6344 2. When Exim is called to do synchronous delivery of a locally submitted
6345 message (the -odf or -odi options), it no longer closes stderr before doing
6348 3. Implemented the mua_wrapper option.
6350 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6352 5. Implemented the functions header_remove(), header_testname(),
6353 header_add_at_position(), and receive_remove_recipient(), and exported them
6356 6. If an ACL "warn" statement specified the addition of headers, Exim already
6357 inserted X-ACL-Warn: at the start if there was no header name. However, it
6358 was not making this test for the second and subsequent header lines if
6359 there were newlines in the string. This meant that an invalid header could
6360 be inserted if Exim was badly configured.
6362 7. Allow an ACL "warn" statement to add header lines at the start or after all
6363 the Received: headers, as well as at the end.
6365 8. Added the rcpt_4xx retry error code.
6367 9. Added postmaster_mailfrom=xxx to callout verification option.
6369 10. Added mailfrom=xxxx to the callout verification option, for verify=
6372 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6373 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6375 12. Inserted some casts to stop certain compilers warning when using pointer
6376 differences as field lengths or precisions in printf-type calls (mostly
6377 affecting debugging statements).
6379 13. Added optional readline() support for -be (dynamically loaded).
6381 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6382 same clock tick as a message's arrival, so that its received time was the
6383 same as the "first fail" time on the retry record, and that message
6384 remained on the queue past the ultimate address timeout, every queue runner
6385 would try a delivery (because it was past the ultimate address timeout) but
6386 after another failure, the ultimate address timeout, which should have then
6387 bounced the address, did not kick in. This was a "< instead of <=" error;
6388 in most cases the first failure would have been in the next clock tick
6389 after the received time, and all would be well.
6391 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6392 being recognized when the domain list was tested by the match_domain
6393 condition in an expansion string.
6395 16. Added the ${str2b64: operator.
6397 17. Exim was always calling setrlimit() to set a large limit for the number of
6398 processes, without checking whether the existing limit was already
6399 adequate. (It did check for the limit on file descriptors.) Furthermore,
6400 errors from getrlimit() and setrlimit() were being ignored. Now they are
6401 logged to the main and panic logs, but Exim does carry on, to try to do its
6402 job under whatever limits there are.
6404 18. Imported PCRE 5.0.
6406 19. Trivial typo in log message " temporarily refused connection" (the leading
6409 20. If the log selector return_path_on_delivery was set and an address was
6410 redirected to /dev/null, the delivery process crashed because it assumed
6411 that a return path would always be set for a "successful" delivery. In this
6412 case, the whole delivery is bypassed as an optimization, and therefore no
6415 21. Internal re-arrangement: the function for sending a challenge and reading
6416 a response while authentication was assuming a zero-terminated challenge
6417 string. It's now changed to take a pointer and a length, to allow for
6418 binary data in such strings.
6420 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6422 23. Exim was not respecting finduser_retries when seeking the login of the
6423 uid under which it was called; it was always trying 10 times. (The default
6424 setting of finduser_retries is zero.) Also, it was sleeping after the final
6425 failure, which is pointless.
6427 24. Implemented tls_on_connect_ports.
6429 25. Implemented acl_smtp_predata.
6431 26. If the domain in control=submission is set empty, Exim assumes that the
6432 authenticated id is a complete email address when it generates From: or
6433 Sender: header lines.
6435 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6436 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6437 chown and chgrp in /bin and hostname in /usr/bin.
6439 28. Exim was keeping the "process log" file open after each use, just as it
6440 does for the main log. This opens the possibility of it remaining open for
6441 long periods when the USR1 signal hits a daemon. Occasional processlog
6442 errors were reported, that could have been caused by this. Anyway, it seems
6443 much more sensible not to leave this file open at all, so that is what now
6446 29. The long-running daemon process does not normally write to the log once it
6447 has entered its main loop, and it closes the log before doing so. This is
6448 so that log files can straightforwardly be renamed and moved. However,
6449 there are a couple of unusual error situations where the daemon does write
6450 log entries, and I had neglected to close the log afterwards.
6452 30. The text of an SMTP error response that was received during a remote
6453 delivery was being truncated at 512 bytes. This is too short for some of
6454 the long messages that one sometimes sees. I've increased the limit to
6457 31. It is now possible to make retry rules that apply only when a message has a
6458 specific sender, in particular, an empty sender.
6460 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6461 it possible to be selective about when SMTP synchronization is enforced.
6463 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6465 32. Implemented hosts_connection_nolog.
6467 33. Added an ACL for QUIT.
6469 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6472 35. Added mailbox_size and mailbox_filecount to appendfile.
6474 36. Added control = no_multiline_responses to ACLs.
6476 37. There was a bug in the logic of the code that waits for the clock to tick
6477 in the case where the clock went backwards by a substantial amount such
6478 that the microsecond fraction of "now" was more than the microsecond
6479 fraction of "then" (but the whole seconds number was less).
6481 38. Added support for the libradius Radius client library this is found on
6482 FreeBSD (previously only the radiusclient library was supported).
6488 1. When certain lookups returned multiple values in the form name=value, the
6489 quoting of the values was not always being done properly. Specifically:
6490 (a) If the value started with a double quote, but contained no whitespace,
6492 (b) If the value contained whitespace other than a space character (i.e.
6493 tabs or newlines or carriage returns) it was not quoted.
6494 This fix has been applied to the mysql and pgsql lookups by writing a
6495 separate quoting function and calling it from the lookup code. The fix
6496 should probably also be applied to nisplus, ibase and oracle lookups, but
6497 since I cannot test any of those, I have not disturbed their existing code.
6499 2. A hit in the callout cache for a specific address caused a log line with no
6500 reason for rejecting RCPT. Now it says "Previous (cached) callout
6501 verification failure".
6503 3. There was an off-by-one bug in the queryprogram router. An over-long
6504 return line was truncated at 256 instead of 255 characters, thereby
6505 overflowing its buffer with the terminating zero. As well as fixing this, I
6506 have increased the buffer size to 1024 (and made a note to document this).
6508 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6509 when Exim is waiting for an SMTP response from a remote server, Exim
6510 restarts its select() call on the socket, thereby resetting its timeout.
6511 This is not a problem when such interrupts are rare. Somebody set up a cron
6512 job to run exiwhat every 2 minutes, which is less than the normal select()
6513 timeout (5 or 10 minutes). This meant that the select() timeout never
6514 kicked in because it was always reset. I have fixed this by comparing the
6515 time when an interrupt arrives with the time at the start of the first call
6516 to select(). If more time than the timeout has elapsed, the interrupt is
6517 treated as a timeout.
6519 5. Some internal re-factoring in preparation for the addition of Sieve
6520 extensions (by MH). In particular, the "personal" test is moved to a
6521 separate function, and given an option for scanning Cc: and Bcc: (which is
6522 not set for Exim filters).
6524 6. When Exim created an email address using the login of the caller as the
6525 local part (e.g. when creating a From: or Sender: header line), it was not
6526 quoting the local part when it contained special characters such as @.
6528 7. Installed new OpenBSD configuration files.
6530 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6531 try to make them clearer.
6533 9. Callout options, other than the timeout value, were being ignored when
6534 verifying sender addresses in header lines. For example, when using
6536 verify = header_sender/callout=no_cache
6538 the cache was (incorrectly) being used.
6540 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6541 only the Cygwin environment.
6543 11. When return_path_on_delivery was set as a log selector, if different remote
6544 addresses in the same message used different return paths and parallel
6545 remote delivery occurred, the wrong values would sometimes be logged.
6546 (Whenever a remote delivery process finished, the return path value from
6547 the most recently started remote delivery process was logged.)
6549 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6550 lines when AUTH and/or TLS are in use. This is the "received protocol"
6551 field. Exim used to use "asmtp" for authenticated SMTP, without any
6552 indication (in the protocol name) for TLS use. Now it follows the RFC and
6553 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6554 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6555 names appear in log lines as well as in Received: header lines.
6557 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6558 extensions, and comparison tests, and to fix some bugs.
6560 14. Changes to the "personal" filter test:
6562 (1) The test was buggy in that it was just doing the equivalent of
6563 "contains" tests on header lines. For example, if a user's address was
6564 anne@some.where, the "personal" test would incorrectly be true for
6566 To: susanne@some.where
6568 This test is now done by extracting each address from the header in turn,
6569 and checking the entire address. Other tests that are part of "personal"
6570 are now done using regular expressions (for example, to check local parts
6571 of addresses in From: header lines).
6573 (2) The list of non-personal local parts in From: addresses has been
6574 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6575 taken from the Sieve specification recommendations.
6577 (3) If the message contains any header line starting with "List-" it is
6578 treated as non-personal.
6580 (4) The test for "circular" in the Subject: header line has been removed
6581 because it now seems ill-conceived.
6583 15. Minor typos in src/EDITME comments corrected.
6585 16. Installed latest exipick from John Jetmore.
6587 17. If headers_add on a router specified a text string that was too long for
6588 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6589 of string_sprintf() is now avoided.
6591 18. $message_body_size was not set (it was always zero) when running the DATA
6592 ACL and the local_scan() function.
6594 19. For the "mail" command in an Exim filter, no default was being set for
6595 the once_repeat time, causing a random time value to be used if "once" was
6596 specified. (If the value happened to be <= 0, no repeat happened.) The
6597 default is now 0s, meaning "never repeat". The "vacation" command was OK
6598 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6599 (I found it when inspecting the code).
6601 20. There is now an overall timeout for performing a callout verification. It
6602 defaults to 4 times the callout timeout, which applies to individual SMTP
6603 commands during the callout. The overall timeout applies when there is more
6604 than one host that can be tried. The timeout is checked before trying the
6605 next host. This prevents very long delays if there are a large number of
6606 hosts and all are timing out (e.g. when the network connections are timing
6607 out). The value of the overall timeout can be changed by specifying an
6608 additional sub-option for "callout", called "maxwait". For example:
6610 verify = sender/callout=5s,maxwait=20s
6612 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6613 to the end before writing, but this should make it even safer).
6615 22. Exim was forgetting that it had advertised PIPELINING for the second and
6616 subsequent messages on an SMTP connection. It was also not resetting its
6617 memory on STARTTLS and an internal HELO.
6619 23. When Exim logs an SMTP synchronization error within a session, it now
6620 records whether PIPELINING has been advertised or not.
6622 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6623 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6624 rather than long int.
6626 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6628 26. Added the never_mail option to autoreply.
6634 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6635 crash if the getsockname() call failed; this can happen if a connection is
6636 closed very soon after it is established. The problem was simply in the
6637 order in which certain operations were done, causing Exim to try to write
6638 to the SMTP stream before it had set up the file descriptor. The bug has
6639 been fixed by making things happen in the correct order.
6645 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6646 before the connection was closed, thus losing the rejection response.
6648 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6649 some early Solaris releases, but causes trouble in current releases where
6650 socklen_t is defined.
6652 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6655 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6658 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6659 that releasing the top of it at the end releases what was used for sub-
6660 expansions (unless the block got too big). However, discard this block if
6661 the first thing is a variable or header, so that we can use its block when
6662 it is dynamic (useful for very large $message_headers, for example).
6664 6. Lookups now cache *every* query, not just the most recent. A new, separate
6665 store pool is used for this. It can be recovered when all lookup caches are
6666 flushed. Lookups now release memory at the end of their result strings.
6667 This has involved some general refactoring of the lookup sources.
6669 7. Some code has been added to the store_xxx() functions to reduce the amount
6670 of flapping under certain conditions.
6672 8. log_incoming_interface used to affect only the <= reception log lines. Now
6673 it causes the local interface and port to be added to several more SMTP log
6674 lines, for example "SMTP connection from", and rejection lines.
6676 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6678 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6680 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6681 could be overwritten at the end of the current message (or the start of a
6682 new message if it was set in a HELO ACL). The value is now preserved for
6683 the duration of the SMTP connection.
6685 12. If a transport had a headers_rewrite setting, and a matching header line
6686 contained an unqualified address, that address was qualified, even if it
6687 did not match any rewriting rules. The underlying bug was that the values
6688 of the flags that permit the existence of unqualified sender and recipient
6689 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6690 non-local messages, and by -bnq for local messages) were not being
6691 preserved with the message after it was received.
6693 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6694 "next input=" as part of the text comprising the host identity instead of
6695 the correct text. The code was using the same buffer for two different
6696 strings. However, depending on which order the printing function evaluated
6697 its arguments, the bug did not always show up. Under Linux, for example, my
6698 test suite worked just fine.
6700 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6701 doesn't work with some older versions of Perl. It has been changed to "my",
6702 which in any case is probably the better facility to use.
6704 15. A really picky compiler found some instances of statements for creating
6705 error messages that either had too many or two few arguments for the format
6708 16. The size of the buffer for calls to the DNS resolver has been increased
6709 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6710 for addresses that have a lot of PTR records. This alleviates a problem; it
6711 does not fully solve it.
6713 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6714 buffer now truncates the list and logs the incident, which is the same
6715 action as happens when Exim is looking up a host name and its aliases.
6716 Previously in this situation something unpredictable would happen;
6717 sometimes it was "internal error: store_reset failed".
6719 18. If a server dropped the connection unexpectedly when an Exim client was
6720 using GnuTLS and trying to read a response, the client delivery process
6721 crashed while trying to generate an error log message.
6723 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6724 string, for example:
6726 warn message = H1: something\nH2: something
6728 the text was added as a single header line from Exim's point of view
6729 though it ended up OK in the delivered message. However, searching for the
6730 second and subsequent header lines using $h_h2: did not work. This has been
6731 fixed. Similarly, if a system filter added multiple headers in this way,
6732 the routers could not see them.
6734 20. Expanded the error message when iplsearch is called with an invalid key to
6735 suggest using net-iplsearch in a host list.
6737 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6738 ACLs are no longer actually imposed (and a message to that effect is
6741 22. If a "gecos" field in a passwd entry contained escaped characters, in
6742 particular, if it contained a \" sequence, Exim got it wrong when building
6743 a From: or a Sender: header from that name. A second bug also caused
6744 incorrect handling when an unquoted " was present following a character
6745 that needed quoting.
6747 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6748 was not being matched caselessly.
6750 24. Arranged for all hyphens in the exim.8 source to be escaped with
6753 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6754 properly. Recipient callout cache records were still being keyed to include
6755 the sender, even when use_sender was set false. This led to far more
6756 callouts that were necessary. The sender is no longer included in the key
6757 when use_sender is false.
6759 26. Added "control = submission" modifier to ACLs.
6761 27. Added the ${base62d: operator to decode base 62 numbers.
6763 28. dnsdb lookups can now access SRV records.
6765 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6766 the configuration file.
6768 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6769 (-v) mode. This makes the output for a verbose queue run more intelligible.
6771 31. Added a use_postmaster feature to recipient callouts.
6773 32. Added the $body_zerocount variable, containing the number of binary zero
6774 bytes in the message body.
6776 33. The time of last modification of the "new" subdirectory is now used as the
6777 "mailbox time last read" when there is a quota error for a maildir
6780 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6782 35. Added +ignore_unknown as a special item in host lists.
6784 36. Code for decoding IPv6 addresses in host lists is now included, even if
6785 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6786 address was recognized as an IP address, but was then not correctly decoded
6787 into binary, causing unexpected and incorrect effects when compared with
6794 1. Very minor rewording of debugging text in manualroute to say "list of
6795 hosts" instead of "hostlist".
6797 2. If verify=header_syntax was set, and a header line with an unqualified
6798 address (no domain) and a large number of spaces between the end of the
6799 name and the colon was received, the reception process suffered a buffer
6800 overflow, and (when I tested it) crashed. This was caused by some obsolete
6801 code that should have been removed. The fix is to remove it!
6803 3. When running in the test harness, delay a bit after writing a bounce
6804 message to get a bit more predictability in the log output.
6806 4. Added a call to search_tidyup() just before forking a reception process. In
6807 theory, someone could use a lookup in the expansion of smtp_accept_max_
6808 per_host which, without the tidyup, could leave open a database connection.
6810 5. Added the variables $recipient_data and $sender_data which get set from a
6811 lookup success in an ACL "recipients" or "senders" condition, or a router
6812 "senders" option, similar to $domain_data and $local_part_data.
6814 6. Moved the writing of debug_print from before to after the "senders" test
6817 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6818 problems for message scanning, either using a data ACL, or using
6819 local_scan() because the Received: header was not generated till after they
6820 were called (in order to set the time as the time of reception completion).
6821 I have revised the way this works. The header is now generated after the
6822 body is received, but before the ACL or local_scan() are called. After they
6823 are run, the timestamp in the header is updated.
6829 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6830 before starting a queue runner without re-exec. This happened only when
6831 deliver_drop_privilege was set or when the Exim user was set to root. The
6832 effect of the bug was that timeouts during subsequent deliveries caused
6833 crashes instead of being properly handled. The handler is now left at its
6834 default (and expected) setting.
6836 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6837 message, again when deliver_drop_privilege is set or Exim is run as root.
6838 The bug described in (1) was not present in this case, but the tidying up
6839 of the other signals was missing. I have made the two cases consistent.
6841 3. The ignore_target_hosts setting on a manualroute router was being ignored
6842 for hosts that were looked up using the /MX notation.
6844 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6847 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6848 operated on the sender address. After changing the $sender_address to <>
6849 for the sender address verify, Exim was re-instated it as the original
6850 (before rewriting) address, but remembering that it had rewritten it, so it
6851 wasn't rewriting it again. This bug also had the effect of breaking the
6852 sender address verification caching when the sender address was rewritten.
6854 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6855 This has been changed so that if the ip literal address matches
6856 ignore_target_hosts, the router declines.
6858 7. Added expansion conditions match_domain, match_address, and match_local_
6859 part (NOT match_host).
6861 8. The placeholder for the Received: header didn't have a length field set.
6863 9. Added code to Exim itself and to exim_lock to test for a specific race
6864 condition that could lead to file corruption when using MBX delivery. The
6865 issue is with the lockfile that is created in /tmp. If this file is removed
6866 after a process has opened it but before that process has acquired a lock,
6867 there is the potential for a second process to recreate the file and also
6868 acquire a lock. This could lead to two Exim processes writing to the file
6869 at the same time. The added code performs the same test as UW imapd; it
6870 checks after acquiring the lock that its file descriptor still refers to
6871 the same named file.
6873 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6874 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6877 11. Added the smtp_active_hostname option. If used, this will typically be made
6878 to depend on the incoming interface address. Because $interface_address is
6879 not set up until the daemon has forked a reception process, error responses
6880 that can happen earlier (such as "too many connections") no longer contain
6883 12. If an expansion in a condition on a "warn" statement fails because a lookup
6884 defers, the "warn" statement is abandoned, and the next ACL statement is
6885 processed. Previously this caused the whole ACL to be aborted.
6887 13. Added the iplsearch lookup type.
6889 14. Added ident_timeout as a log selector.
6891 15. Added tls_certificate_verified as a log selector.
6893 16. Added a global option tls_require_ciphers (compare the smtp transport
6894 option of the same name). This controls incoming TLS connections.
6896 17. I finally figured out how to make tls_require_ciphers do a similar thing
6897 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6898 before starting the TLS session.
6900 18. Tabs are now shown as \t in -bP output.
6902 19. If the log selector return_path_on_delivery was set, Exim crashed when
6903 bouncing a message because it had too many Received: header lines.
6905 20. If two routers both had headers_remove settings, and the first one included
6906 a superfluous trailing colon, the final name in the first list and the
6907 first name in the second list were incorrectly joined into one item (with a
6908 colon in the middle).
6914 1. Added -C and -D options to the exinext utility, mainly to make it easier
6915 to include in the automated testing, but these could be helpful when
6916 multiple configurations are in use.
6918 2. The exinext utility was not formatting the output nicely when there was
6919 an alternate port involved in the retry record key, nor when there was a
6920 message id as well (for retries that were specific to a specific message
6921 and a specific host). It was also confused by IPv6 addresses, because of
6922 the additional colons they contain. I have fixed the IPv4 problem, and
6923 patched it up to do a reasonable job for IPv6.
6925 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6926 delivery, the log line now contains "pipelined" if PIPELINING was used.
6928 4. An SMTP transport process used to panic and die if the bind() call to set
6929 an explicit outgoing interface failed. This has been changed; it is now
6930 treated in the same way as a connect() failure.
6932 5. A reference to $sender_host_name in the part of a conditional expansion
6933 that was being skipped was still causing a DNS lookup. This no longer
6936 6. The def: expansion condition was not recognizing references to header lines
6937 that used bh_ and bheader_.
6939 7. Added the _cache feature to named lists.
6941 8. The code for checking quota_filecount in the appendfile transport was
6942 allowing one more file than it should have been.
6944 9. For compatibility with Sendmail, the command line option
6952 and sets the incoming protocol and host name (for trusted callers). The
6953 host name and its colon can be omitted when only the protocol is to be set.
6954 Note the Exim already has two private options, -pd and -ps, that refer to
6955 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6956 "s", but I don't think that's a major issue.
6958 10. A number of refactoring changes to the code, none of which should affect
6961 (a) The number of logging options was getting close to filling up the
6962 32-bit word that was used as a bit map. I have split them into two classes:
6963 those that are passed in the argument to log_write(), and those that are
6964 only ever tested independently outside of that function. These are now in
6965 separate 32-bit words, so there is plenty of room for expansion again.
6966 There is no change in the user interface or the logging behaviour.
6968 (b) When building, for example, log lines, the code previously used a
6969 macro that called string_cat() twice, in order to add two strings. This is
6970 not really sufficiently general. Furthermore, there was one instance where
6971 it was actually wrong because one of the argument was used twice, and in
6972 one call a function was used. (As it happened, calling the function twice
6973 did not affect the overall behaviour.) The macro has been replaced by a
6974 function that can join an arbitrary number of extra strings onto a growing
6977 (c) The code for expansion conditions now uses a table and a binary chop
6978 instead of a serial search (which was left over from when there were very
6979 few conditions). Also, it now recognizes conditions like "pam" even when
6980 the relevant support is not compiled in: a suitably worded error message is
6981 given if an attempt is made to use such a condition.
6983 11. Added ${time_interval:xxxxx}.
6985 12. A bug was causing one of the ddress fields not to be passed back correctly
6986 from remote delivery subprocesses. The field in question was not being
6987 subsequently used, so this caused to problems in practice.
6989 13. Added new log selectors queue_time and deliver_time.
6991 14. Might have fixed a bug in maildirsizefile handling that threw up
6992 "unexpected character" debug warnings, and recalculated the data
6993 unnecessarily. In any case, I expanded the warning message to give more
6996 15. Added the message "Restricted characters in address" to the statements in
6997 the default ACL that block characters like @ and % in local parts.
6999 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7000 Three changes have been made:
7002 (a) There was a serious bug; a negative response to MAIL caused the whole
7003 recipient domain to be cached as invalid, thereby blocking all messages
7004 to all local parts at the same domain, from all senders. This bug has
7005 been fixed. The domain is no longer cached after a negative response to
7006 MAIL if the sender used is not empty.
7008 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7011 (c) A new callout option, "use_sender" has been added for people who want
7012 the modified behaviour.
7018 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7021 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7022 indeed breaks things for older releases.
7024 3. Added additional logging to the case where there is a problem reading data
7025 from a filter that is running in a subprocess using a pipe, in order to
7026 try to track down a specific problem.
7028 4. Testing facility fudge: when running in the test harness and attempting
7029 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7030 getting "No route to host". Convert this to a timeout.
7032 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7035 6. Some OS don't have socklen_t but use size_t instead. This affects the
7036 fifth argument of getsockopt() amongst other things. This is now
7037 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7038 can be set for individual OS. I have set it for SunOS5, OSF1, and
7039 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7040 some earlier ones do not.
7042 7. Change 4.30/15 was not doing the test caselessly.
7044 8. The standard form for an IPv6 address literal was being rejected by address
7045 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7046 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7047 this, as well as the form without the "IPv6" on the front (but only when
7048 address literals are enabled, of course).
7050 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7052 10. Exim crashed if a message with an empty sender address specified by -f
7053 encountered a router with an errors_to setting. This could be provoked only
7054 by a command such as
7058 where an empty string was supplied; "<>" did not hit this bug.
7060 11. Installed PCRE release 4.5.
7062 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7063 remained set. It is now erased.
7065 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7066 times from message ids (which are base 36 rather than the normal 62).
7068 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7069 were being counted as actual protocol errors, and logged if the log
7070 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7071 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7072 and DATA following a set of rejected RCPTs do not count as protocol errors.
7073 In other words, Exim assumes they were pipelined, though this may not
7074 actually be the case. Of course, in all cases the client gets an
7075 appropriate error code.
7077 15. If a lookup fails in an ACL condition, a message about the failure may
7078 be available; it is used if testing the ACL cannot continue, because most
7079 such messages specify what the cause of the deferral is. However, some
7080 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7081 that caused an old message to be retained and used if a later statement
7082 caused a defer, replacing the real cause of the deferral.
7084 16. If an IP address had so many PTR records that the DNS lookup buffer
7085 was not large enough to hold them, Exim could crash while trying to process
7086 the truncated data. It now detects and logs this case.
7088 17. Further to 4.21/58, another change has been made: if (and only if) the
7089 first line of a message (the first header line) ends with CRLF, a bare LF
7090 in a subsequent header line has a space inserted after it, so as not to
7091 terminate the header.
7093 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7094 unnecessarily, used atoi() instead of strtol(), and didn't check the
7095 termination when getting file sizes from file names by regex.
7097 19. Completely re-implemented the support for maildirsize files, in the light
7098 of a number of problems with the previous contributed implementation
7099 (4.30/29). In particular:
7101 . If the quota is zero, the maildirsize file is maintained, but no quota is
7104 . If the maildir directory does not exist, it is created before any attempt
7105 to write a maildirsize file.
7107 . The quota value in the file is just a cache; if the quota is changed in
7108 the transport, the new value overrides.
7110 . A regular expression is available for excluding directories from the
7113 20. The autoreply transport checks the characters in options that define the
7114 message's headers; it allows continued headers, but it was checking with
7115 isspace() after an embedded newline instead of explicitly looking for a
7118 21. If all the "regular" hosts to which an address was routed had passed their
7119 expiry times, and had not reached their retry times, the address was
7120 bounced, even if fallback hosts were defined. Now Exim should go on to try
7123 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7124 equivalent code in the SMTP transport. Some hosts send humungous responses
7125 to HELO/EHLO, more than 1024 it seems.
7127 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7128 gives compiler warnings in some environments. I've now done it "properly",
7131 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7132 (because of problems with the built-in one) was declared to return uschar *
7133 instead of char *, causing compiler failure.
7135 25. Fixed a file descriptor leak when processing alias/forward files.
7137 26. Fixed a minor format string issue in dbfn.c.
7139 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7141 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7142 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7145 29. When a sender address is verified, it is cached, to save repeating the test
7146 when there is more than one recipient in a message. However, when the
7147 verification involves a callout, it is possible for different callout
7148 options to be set for different recipients. It is too complicated to keep
7149 track of this in the cache, so now Exim always runs a verification when a
7150 callout is required, relying on the callout cache for the optimization.
7151 The overhead is duplication of the address routing, but this should not be
7154 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7155 to be verified with callout=postmaster, and the main callout worked but the
7156 postmaster check failed, the verification correctly failed. However, if a
7157 subsequent RCPT command asked for sender verification *without* the
7158 postmaster check, incorrect caching caused this verification also to fail,
7161 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7162 it was not caching the DNS options (qualify_single, search_parents) that
7163 were used when the lookup failed. A subsequent lookup with different
7164 options therefore always gave the same answer, though there were cases
7165 where it should not have. (Example: a "domains = !$mx_any" option on a
7166 dnslookup router: the "domains" option is always processed without any
7167 widening, but the router might have qualify_single set.) Now Exim uses the
7168 cached value only when the same options are set.
7170 32. Added John Jetmore's "exipick" utility to the distribution.
7172 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7173 than a timeout (e.g. a certificate is required, and is not provided), an
7174 Exim server now closes the connection immediately. Previously it waited for
7175 the client to close - but if the client is SSL, it seems that they each
7176 wait for each other, leading to a delay before one of them times out.
7178 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7179 maintained 0.8.x compatibility because I don't think many are using it, and
7180 it is clearly obsolete.
7182 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7185 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7186 client certificate was expired. A simple patch fixes this, though I don't
7187 understand the full logic of why the verify callback is called multiple
7190 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7191 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7192 which causes problems with some clients (such as the Certicom SSL Plus
7193 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7194 disables the coutermeasure allowing Eudora to connect."
7196 38. Exim was not checking that a write() to a log file succeeded. This could
7197 lead to Bad Things if a log got too big, in particular if it hit a file
7198 size limit. Exim now panics and dies if it cannot write to a log file, just
7199 as it does if it cannot open a log file.
7201 39. Modified OS/Makefile-Linux so that it now contains
7203 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7205 The two -D definitions ensure that Exim is compiled with large file
7206 support, which makes it possible to handle log files that are bigger than
7209 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7210 instance) a domain was checked against a named list that involved a lookup,
7211 causing $domain_data to be set, then another domain was checked against the
7212 same list, then the first domain was re-checked, the value of $domain_data
7213 after the final check could be wrong. In particular, if the second check
7214 failed, it could be set empty. This bug probably also applied to
7217 41. The strip_trailing_dot option was not being applied to the address given
7218 with the -f command-line option.
7220 42. The code for reading a message's header from the spool was incrementing
7221 $received_count, but never initializing it. This meant that the value was
7222 incorrect (doubled) while delivering a message in the same process in which
7223 it was received. In the most common configuration of Exim, this never
7224 happens - a fresh exec is done - but it can happen when
7225 deliver_drop_privilege is set.
7227 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7228 it now includes up to 150 characters of the unexpected data in the log
7231 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7232 and building data strings. The size of both of these buffers was 10 000
7233 bytes - far larger than anybody would *ever* want, thought I. Needless to
7234 say, somebody hit the limit. I have increased the maximum line length to
7235 20 000 and the maximum data length of concatenated lines to 100 000. I have
7236 also fixed two bugs, because there was no checking on these buffers. Tsk,
7237 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7238 buffer is too small.
7240 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7241 lsearch lookups. Now it does.
7243 46. When parsing a route_list item in a manualroute router, a fixed-length
7244 buffer was used for the list of hosts. I made this 1024 bytes long,
7245 thinking that nobody would ever have a list of hosts that long. Wrong.
7246 Somebody had a whole pile of complicated expansion conditions, and the
7247 string was silently truncated, leading to an expansion error. It turns out
7248 that it is easier to change to an unlimited length (owing to other changes
7249 that have happened since this code was originally written) than to build
7250 structure for giving a limitation error. The length of the item that
7251 expands into the list of hosts is now unlimited.
7253 47. The lsearch lookup could not handle data where the length of text line was
7254 more than 4095 characters. Such lines were truncated, leading to shortened
7255 data being returned. It should now handle lines of any length.
7257 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7258 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7261 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7262 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7263 no longer happens in comment lines. A list of replacements is now placed
7264 at the head of all of the source files, except those whose only change is
7265 to replace PERL_COMMAND in the very first #! line.
7267 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7268 messages on the queue, with a bottom-up merge sort, using code contributed
7269 by Michael Haardt. This should make operations like -bp somewhat faster on
7270 large queues. It won't affect queue runners, except when queue_run_in_order
7273 51. Installed eximstats 1.31 in the distribution.
7275 52. Added support for SRV lookups to the dnslookup router.
7277 53. If an ACL referred to $message_body or $message_body_end, the value was not
7278 reset for any messages that followed in the same SMTP session.
7280 54. The store-handling optimization for building very long strings was not
7281 differentiating between the different store pools. I don't think this
7282 actually made any difference in practice, but I've tidied it.
7284 55. While running the routers to verify a sender address, $sender_address
7285 was still set to the sender address. This is wrong, because when routing to
7286 send a bounce to the sender, it would be empty. Therefore, I have changed
7287 it so that, while verifying a sender address, $sender_address is set to <>.
7288 (There is no change to what happens when verifying a recipient address.)
7290 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7291 target A or AAAA records (if not already returned) without resetting the
7292 qualify_single or search_parents options of the DNS resolver. These are
7293 inappropriate in this case because the targets of MX and SRV records must
7294 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7295 target that, when qualified, matched something in the local domain. These
7296 two options are now turned off when doing these lookups.
7298 57. It seems that at least some releases of Reiserfs (which does not have the
7299 concept of a fixed number of inodes) returns zero and not -1 for the
7300 number of available inodes. This interacted badly with check_spool_inodes,
7301 which assumed that -1 was the "no such thing" setting. What I have done is
7302 to check that the total number of inodes is greater than zero before doing
7303 the test of how many are available.
7305 58. When a "warn" ACL statement has a log_message modifier, the message is
7306 remembered, and not repeated. This is to avoid a lot of repetition when a
7307 message has many recipients that cause the same warning to be written.
7308 However, Exim was preserving the list of already written lines for an
7309 entire SMTP session, which doesn't seem right. The memory is now reset if a
7310 new message is started.
7312 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7313 debugging output unless log_rewrite was also set.
7315 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7316 of (int)(handle) when we know that handle contains (void *)(-1).
7318 61. The Exim daemon panic-logs an error return when it closes the incoming
7319 connection. However "connection reset by peer" seems to be common, and
7320 isn't really an error worthy of noting specially, so that particular error
7323 62. When Exim is trying to find all the local interfaces, it used to panic and
7324 die if the ioctl to get the interface flags failed. However, it seems that
7325 on at least one OS (Solaris 9) it is possible to have an interface that is
7326 included in the list of interfaces, but for which you get a failure error
7327 for this call. This happens when the interface is not "plumbed" into a
7328 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7329 failure of the "get flags" call assumes that the interface is down.
7331 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7332 makes life easier for people who are doing arithmetic on fields extracted
7333 from dates, where you often get leading zeros that should not be
7334 interpreted as octal.
7336 64. Added qualify_domain to the redirect router, to override the global
7339 65. If a pathologically long header line contained very many addresses (the
7340 report of this problem mentioned 10 000) and each of them was rewritten,
7341 Exim could use up a very large amount of memory. (It kept on making new
7342 copies of the header line as it rewrote, and never released the old ones.)
7343 At the expense of a bit more processing, the header rewriting function has
7344 been changed so that it no longer eats memory in this way.
7346 66. The generation of the Received: header has been moved from the time that a
7347 message starts to be received, to the time that it finishes. The timestamp
7348 in the Received: header should now be very close to that of the <= log
7349 line. There are two side-effects of this change:
7351 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7352 logged header lines no longer include the local Received: line, because
7353 it has not yet been created. The same applies to a copy of the message
7354 that is returned to a non-SMTP sender when a message is rejected.
7356 (b) When a filter file is tested using -bf, no additional Received: header
7357 is added to the test message. After some thought, I decided that this
7360 This change does not affect the value of $received_for. It is still set
7361 after address rewriting, but before local_scan() is called.
7363 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7365 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7366 gave an unhelpful panic error message, and a defer error. I have managed to
7367 change this behaviour so that it now rejects any supplied certificate,
7368 which seems right, as the list of acceptable certificates is empty.
7370 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7371 gave an unhelpful defer error. I have not managed to make this reject any
7372 supplied certificates, but the error message it gives is "no certificate
7373 supplied", which is not helpful.
7375 70. exigrep's output now also includes lines that are not associated with any
7376 message, but which match the given pattern. Implemented by a patch from
7377 Martin Sluka, which also tidied up the Perl a bit.
7379 71. Recipient callout verification, like sender verification, was using <> in
7380 the MAIL FROM command. This isn't really the right thing, since the actual
7381 sender may affect whether the remote host accepts the recipient or not. I
7382 have changed it to use the actual sender in the callout; this means that
7383 the cache record is now keyed on a recipient/sender pair, not just the
7384 recipient address. There doesn't seem to be a real danger of callout loops,
7385 since a callout by the remote host to check the sender would use <>.
7386 [SEE ABOVE: changed after hitting problems.]
7388 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7389 temporary errors. However, in the case of such a code being given after
7390 the end of a data transmission (i.e. after ".") Exim was failing to write
7391 a retry record for the message. (Yes, there was some broken host that was
7392 actually sending 8xx at this point.)
7394 73. An unknown lookup type in a host list could cause Exim to panic-die when
7395 the list was checked. (An example that provoked this was putting <; in the
7396 middle of a list instead of at the start.) If this happened during a DATA
7397 ACL check, a -D file could be left lying around. This kind of configuration
7398 error no longer causes Exim to die; instead it causes a defer error. The
7399 incident is still logged to the main and panic logs.
7401 74. Buglet left over from Exim 3 conversion. The message "too many messages
7402 in one connection" was written to the rejectlog but not the mainlog, except
7403 when address rewriting (yes!) was being logged.
7405 75. Added write_rejectlog option.
7407 76. When a system filter was run not as root (that is, when system_filter_user
7408 was set), the values of the $n variables were not being returned to the
7409 main process; thus, they were not subsequently available in the $sn
7412 77. Added +return_path_on_delivery log selector.
7414 78. A connection timeout was being treated differently from recipients deferred
7415 when testing hosts_max_try with a message that was older than the host's
7416 retry timeout. (The host should not be counted, thus allowing all hosts to
7417 be tried at least once before bouncing.) This may have been the cause of an
7418 occasionally reported bug whereby a message would remain on the queue
7419 longer than the retry timeout, but would be bounced if a delivery was
7420 forced. I say "may" because I never totally pinned down the problem;
7421 setting up timeout/retry tests is difficult. See also the next item.
7423 79. The ultimate address timeout was not being applied to errors that involved
7424 a combination of host plus message (for example, a timeout on a MAIL
7425 command). When an address resolved to a number of possible hosts, and they
7426 were not all tried for each delivery (e.g. because of hosts_max_try), a
7427 message could remain on the queue longer than the retry timeout.
7429 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7432 81. Fixed an obscure SMTP outgoing bug which required at least the following
7433 conditions: (a) there was another message waiting for the same server;
7434 (b) the server returned 5xx to all RCPT commands in the first message so
7435 that the message was not completed; (c) the server dropped the connection
7436 or gave a negative response to the RSET that Exim sends to abort the
7437 transaction. The observed case was a dropped connection after DATA that had
7438 been sent in pipelining mode. That is, the server had advertised PIPELINING
7439 but was not implementing it correctly. The effect of the bug was incorrect
7440 behaviour, such as trying another host, and this could lead to a crash.
7446 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7447 and daemon.c were passed as pointers to ints; they should have been
7448 pointers to socklen_t variables (which are typically unsigned ints).
7450 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7453 3. Fixed a really odd bug that affected only the testing scheme; patching a
7454 certain fixed string in the binary changed the value of another string that
7455 happened to be identical to the end of the original first string.
7457 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7458 name", it returns that address as the IP address. On some operating
7459 systems (e.g. Solaris), it also passes back the IP address string as the
7460 "host name". However, on others (e.g. Linux), it passes back an empty
7461 string. Exim wasn't checking for this, and was changing the host name to an
7462 empty string, assuming it had been canonicalized.
7464 5. Although rare, it is permitted to have more than one PTR record for a given
7465 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7466 all the names associated with an address, because they do in Solaris.
7467 However, it seems that they do not in Linux for data that comes from the
7468 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7469 I found this out when I moved to a new Linux workstation and tried to run
7470 the Exim test suite.
7472 To get round this problem I have changed the code so that it now does its
7473 own call to the DNS to look up PTR records when searching for a host name.
7474 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7475 addresses that are only in /etc/hosts are still found.
7477 This behaviour is, however, controlled by an option called host_lookup_
7478 order, which defaults to "bydns:byaddr". If people want to use the other
7479 order, or indeed, just use one or the other means of lookup, they can
7480 specify it in this variable.
7482 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7483 some operating systems, this comes back from gethostbyaddr() as an empty
7484 string, and this is what Exim used to test for. However, it seems that in
7485 other systems, "." is yielded. Exim now tests for this case too.
7487 7. The values of check_spool_space and check_log_space are now held internally
7488 as a number of kilobytes instead of an absolute number of bytes. If a
7489 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7490 kilobyte. This means that much larger values can be stored.
7492 8. Exim monitor: an attempt to get the action menu when not actually pointing
7493 at a message produces an empty menu entitled "No message selected". This
7494 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7495 no entries in it ("Shell widget menu has zero width and/or height"). So I
7496 have added a single, blank menu entry in this case.
7498 9. Added ${quote_local_part.
7500 10. MIME decoding is now applied to the contents of Subject: header lines when
7503 11. Now that a reference to $sender_host_address automatically causes a reverse
7504 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7505 host lookup before query-style lookups in lists that might use this
7506 variable. This has therefore been abolished, and the "net-" prefix is no
7507 longer necessary for query-style lookups.
7509 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7510 have been a typo for LFLAGS, so it has been changed.
7512 13. The install script calls Exim with "-C /dev/null" in order to find the
7513 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7514 to be output. However, since Exim outputs its version number before the
7515 error, it didn't break the script. It just looked ugly. I fixed this by
7516 always allowing "-C /dev/null" if the caller is root.
7518 14. Ignore overlarge ACL variable number when reading spool file - insurance
7519 against a later release with more variables having written the file.
7521 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7522 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7523 this, as well as the form without the "IPv6" on the front.
7525 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7526 OS/Makefile-Darwin file.
7528 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7529 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7531 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7532 variable. However, some OS use ints for inodes. I've added cast to long int
7533 to get rid of the compiler warning.
7535 19. I had forgotten to lock out "/../" in configuration file names when
7536 ALT_CONFIG_PREFIX was set.
7538 20. Routers used for verification do not need to specify transports. However,
7539 if such a router generated a host list, and callout was configured, Exim
7540 crashed, because it could not find a port number from the (non-existent)
7541 transport. It now assumes port 25 in this circumstance.
7543 21. Added the -t option to exigrep.
7545 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7546 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7547 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7548 them is set, LOOKUP_LSEARCH is forced.
7550 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7552 24. Added sender and host information to the "rejected by local_scan()" log
7553 line; previously there was no indication of these.
7555 25. Added .include_if_exists.
7557 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7558 Linux. It turns out that not all file systems support this. Apparently some
7559 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7560 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7561 not supported on the file descriptor, is now ignored when Exim is trying to
7562 sync a directory. This applies only to Linux.
7564 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7566 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7567 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7570 29. Added support for maildirsize files from supplied patch (modified a bit).
7572 30. The use of :fail: followed by an empty string could lead Exim to respond to
7573 sender verification failures with (e.g.):
7575 550 Verification failed for <xxx>
7576 550 Sender verify failed
7578 where the first response line was missing the '-' that indicates it is not
7579 the final line of the response.
7581 31. The loop for finding the name of the user that called Exim had a hardwired
7582 limit of 10; it now uses the value of finduser_retries, which is used for
7583 all other user lookups.
7585 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7588 33. Exim was neglecting to zero errno before one call of strtol() when
7589 expanding a string and expecting an integer value. On some systems this
7590 resulted in spurious "integer overflow" errors. Also, it was casting the
7591 result into an int without checking.
7593 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7594 did not work. The code looks as if it has *never* worked, though it appears
7595 to have been documented since at least release 1.62. I have made it work.
7597 35. The "timeout_DNS" error in retry rules, also documented since at least
7598 1.62, also never worked. As it isn't clear exactly what this means, and
7599 clearly it isn't a major issue, I have abolished the feature by treating it
7600 as "timeout", and writing a warning to the main and panic logs.
7602 36. The display of retry rules for -brt wasn't always showing the error code
7605 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7606 timeout_connect_A, timeout_connect_MX.
7608 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7609 to the empty sender.
7611 39. The daemon was not analysing the content of -oX till after it had closed
7612 stderr and disconnected from the controlling terminal. This meant that any
7613 syntax errors were only noted on the panic log, and the return code from
7614 the command was 0. By re-arranging the code a little, I've made the
7615 decoding happen first, so such errors now appear on stderr, and the return
7616 code is 1. However, the actual setting up of the sockets still happens in
7617 the disconnected process, so errors there are still only recorded on the
7620 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7621 connections (as happens on some IP stacks) was logged at start up time as
7622 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7623 it from "IPv6 and IPv4", which means that two separate sockets are being
7626 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7627 says whether AF_INET or AF_INET6 was passed as an argument.
7629 42. Exiwhat output was messed up when time zones were included in log
7632 43. Exiwhat now gives more information about the daemon's listening ports,
7633 and whether -tls-on-connect was used.
7635 44. The "port" option of the smtp transport is now expanded.
7637 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7638 silently ignored. Now an error message is written to the main and panic
7641 46. There's a new ACL modifier called "logwrite" which writes to a log file
7642 as soon as it is encountered.
7644 47. Added $local_user_uid and $local_user_gid at routing time.
7646 48. Exim crashed when trying to verify a sender address that was being
7649 49. Exim was recognizing only a space character after ".include". It now also
7650 recognizes a tab character.
7652 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7653 extracting the relevant information from the specification. The man page no
7654 longer contains scrambled data for the -d option, and I've added a section
7655 at the front about calling Exim under different names.
7657 51. Added "extra_headers" argument to the "mail" command in filter files.
7659 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7662 53. Installed eximstats 1.29.
7664 54. Added transport_filter_timeout as a generic transport option.
7666 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7667 Cc: header lines. This was required by RFC 822, but it not required by RFC
7670 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7671 incoming messages that did not have them. Now it does so only if the
7672 message originates locally, that is, if there is no associated remote host
7673 address. When Resent- header lines are present, this applies to the Resent-
7674 lines rather than the non-Resent- lines.
7676 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7677 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7679 58. Messages for configuration errors now include the name of the main
7680 configuration file - useful now that there may be more than one file in a
7681 list (.included file names were always shown).
7683 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7684 for those rare installations that do not start the daemon as root or run it
7685 setuid root. I've cut out the call to initgroups() if the daemon is not
7688 60. The Exim user and group can now be bound into the binary as text strings
7689 that are looked up at the start of Exim's processing.
7691 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7693 62. Added $mailstore_basename variable.
7695 63. Installed patch to sieve.c from Michael Haardt.
7697 64. When Exim failed to open the panic log after failing to open the main log,
7698 the original message it was trying to log was written to stderr and debug
7699 output, but if they were not available (the usual case in production), it
7700 was lost. Now it is written to syslog before the two lines that record the
7701 failures to open the logs.
7703 65. Users' Exim filters run in subprocesses under the user's uid. It is
7704 possible for a "deliver" command or an alias in a "personal" command to
7705 provoke an address rewrite. If logging of address rewriting is configured,
7706 this fails because the process is not running as root or exim. There may be
7707 a better way of dealing with this, but for the moment (because 4.30 needs
7708 to be released), I have disabled address rewrite logging when running a
7709 filter in a non-root, non-exim process.
7715 1. The buildconfig auxiliary program wasn't quoting the value set for
7716 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7717 not defined. This bug was masked in 4.22 by the effect that was fixed in
7720 2. Some messages that were rejected after a message id was allocated were
7721 shown as "incomplete" by exigrep. It no longer does this for messages that
7722 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7724 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7725 have allow_domain_literals set, the ID did not get logged in the <= line.
7726 Domain literals are now always recognized in Message-ID: header lines.
7728 4. The first argument for a ${extract expansion item is the key name or field
7729 number. Leading and trailing spaces in this item were not being ignored,
7730 causing some misleading effects.
7732 5. When deliver_drop_privilege was set, single queue runner processes started
7733 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7734 same command in the process it spins off) were not dropping privilege.
7736 6. When the daemon running as "exim" started a queue runner, it always
7737 re-executed Exim in the spun-off process. This is a waste of effort when
7738 deliver_drop_privilege is set. The new process now just calls the
7739 queue-runner function directly.
7745 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7748 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7749 ignored. Though the use of -oP was forcing the writing of a pid file, it
7750 was always written to the default place.
7752 3. If the message "no IP address found for host xxxx" is generated during
7753 incoming verification, it is now followed by identification of the incoming
7754 connection (so you can more easily find what provoked it).
7756 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7758 5. Added some features to "harden" Exim a bit more against certain attacks:
7760 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7761 be put in Local/Makefile. This is like the never_users runtime option,
7762 but it cannot be overridden. The default setting is "root".
7764 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7765 prefix string with which any file named in a -C command line option
7768 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7769 is retained for -C and -D only if the caller of Exim is root. Without
7770 it, the exim user may also use -C and -D and retain privilege.
7772 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7773 command line option is disabled.
7775 6. Macro names set by the -D option must start with an upper case letter, just
7776 like macro names defined in the configuration file.
7778 7. Added "dereference=" facility to LDAP.
7780 8. Two instances of the typo "uknown" in the source files are fixed.
7782 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7783 the Configure-Makefile script screwed up while processing it.
7785 10. Incorporated PCRE 4.4.
7787 11. The SMTP synchronization check was not operating right at the start of an
7788 SMTP session. For example, it could not catch a HELO sent before the client
7789 waited for the greeting. There is now a check for outstanding input at the
7790 point when the greeting is written. Because of the duplex, asynchronous
7791 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7792 way, but not yet received, when the check is performed.
7794 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7795 on TCP/IP sockets, because this apparently causes some broken clients to
7798 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7799 unchanged) from the Cygwin maintainer.
7801 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7802 was supported instead of testing for mbx. Effectively a typo.
7804 15. The spa authenticator server code was not checking that the input it
7805 received was valid base64.
7807 16. The debug output line for the "set" modifier in ACLs was not showing the
7808 name of the variable that was being set.
7810 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7812 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7813 lookup on its own. Something else was needed to trigger the lookup. For
7814 example, a match in host_lookup or the need for a host name in a host list.
7815 Now, if $sender_host_name is referenced and the host name has not yet been
7816 looked up, a lookup is performed. If the lookup fails, the variable remains
7817 empty, and $host_lookup_failed is set to "1".
7819 19. Added "eqi" as a case-independent comparison operator.
7821 20. The saslauthd authentication condition could segfault if neither service
7822 nor realm was specified.
7824 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7825 error message that was logged was misleading, and incoming SMTP
7826 connections were dropped. The message is now more accurate, and temporary
7827 errors are given to SMTP connections.
7829 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7830 (except RSET) with a 421 error, until QUIT is received. However, it was
7831 failing to send a response to QUIT.
7833 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7834 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7835 ACL is now run after the helo_try_verify_hosts code.
7837 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7838 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7839 case-independent, so other case variants are also recognized). Apparently
7840 some systems use these upper case variants.
7842 25. If more than two messages were waiting for the same host, and a transport
7843 filter was specified for the transport, Exim sent two messages over the
7844 same TCP/IP connection, and then failed with "socket operation on non-
7845 socket" when it tried to send the third.
7847 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7849 27. The extern definition of crypt16() in expand.c was not being excluded when
7850 the OS had its own crypt16() function.
7852 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7853 as a preferred synonym for return_size_limit, both as an option and as an
7856 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7858 30. Changed the default configuration ACL to relax the local part checking rule
7859 for addresses that are not in any local domains. For these addresses,
7860 slashes and pipe symbols are allowed within local parts, but the sequence
7861 /../ is explicitly forbidden.
7863 31. SPA server authentication was not clearing the challenge buffer before
7866 32. log_message in a "warn" ACL statement was writing to the reject log as
7867 well as to the main log, which contradicts the documentation and doesn't
7868 seem right (because no rejection is happening). So I have stopped it.
7870 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7871 However, I am unable to do any testing of this.
7873 34. Fixed an infelicity in the appendfile transport. When checking directories
7874 for a mailbox, to see if any needed to be created, it was accidentally
7875 using path names with one or more superfluous leading slashes; tracing
7876 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7878 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7879 contents are added to the log line that is written for every discarded
7880 recipient. (Previously a log_message setting was ignored.)
7882 36. The ${quote: operator now quotes the string if it is empty.
7884 37. The install script runs exim in order to find its version number. If for
7885 some reason other than non-existence or emptiness, which it checks, it
7886 could not run './exim', it was installing it with an empty version number,
7887 i.e. as "exim-". This error state is now caught, and the installation is
7890 38. An argument was missing from the function that creates an error message
7891 when Exim fails to connect to the socket for saslauthd authentication.
7892 This could cause Exim to crash, or give a corrupted message.
7894 39. Added isip, isip4, and isip6 to ${if conditions.
7896 40. The ACL variables $acl_xx are now saved with the message, and can be
7897 accessed later in routers, transports, and filters.
7899 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7900 strings in the file are not string-expanded.
7902 42. If a MAIL command specified a SIZE value that was too large to fit into an
7903 int variable, the check against message_size_limit failed. Such values are
7904 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7905 day this will have to be increased, but I don't think I want to be around
7906 when emails are that large.
7913 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7914 iconv() is not standard in FreeBSD.
7916 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7917 IPv6 enabled. The observed symptom was a segmentation fault on return from
7918 the function os_common_find_running_interfaces() in src/os.c.
7920 3. In the check_special_case() function in daemon.c I had used "errno" as an
7921 argument name, which causes warnings on some systems. This was basically a
7922 typo, since it was named "eno" in the comments!
7924 4. The code that waits for the clock to tick (at a resolution of some fraction
7925 of a second) so as to ensure message-id uniqueness was always waiting for
7926 at least one whole tick, when it could have waited for less. [This is
7927 almost certainly not relevant at current processor speeds, where it is
7928 unlikely to ever wait at all. But we try to future-proof.]
7930 5. The function that sleeps for a time interval that includes fractions of a
7931 second contained a race. It did not block SIGALRM between setting the
7932 timer, and suspending (a couple of lines later). If the interval was short
7933 and the sigsuspend() was delayed until after it had expired, the suspension
7934 never ended. On busy systems this could lead to processes getting stuck for
7937 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7938 process, before it forks any delivery processes. The open lookup caching
7939 mechanism meant that the open file or database connection was passed into
7940 the delivery process. The problem was that delivery processes always tidy
7941 up cached lookup data. This could cause a problem for the next delivery
7942 process started by the queue runner, because the external queue runner
7943 process does not know about the closure. So the next delivery process
7944 still has data in the lookup cache. In the case of a file lookup, there was
7945 no problem because closing a file descriptor in a subprocess doesn't affect
7946 the parent. However, if the lookup was caching a connection to a database,
7947 the connection was closed, and the second delivery process was likely to
7948 see errors such as "PGSQL: query failed: server closed the connection
7949 unexpectedly". The problem has been fixed by closing all cached lookups
7950 in a queue runner before running a delivery process.
7952 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7953 seem to have the "const" qualifier which it has on other OS. I've
7956 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7957 *of the same type* (client or server) with the same public name that an
7958 error should be diagnosed.
7960 9. When Exim looked up a host name for an IP address, but failed to find the
7961 original IP address when looking up the host name (a safety check), it
7962 output the message "<ip address> does not match any IP for NULL", which was
7963 confusing, to say the least. The bug was that the host name should have
7964 appeared instead of "NULL".
7966 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7967 user that is built into the binary, and the -C or -D options is used, root
7968 privilege is dropped before the configuration file is read. In addition,
7969 logging is switched to stderr instead of the normal log files. If the
7970 configuration then re-defines the Exim user, the unprivileged environment
7971 is probably not what is expected, so Exim logs a panic warning message (but
7974 However, if deliver_drop_privilege is set, the unprivileged state may well
7975 be exactly what is intended, so the warning has been cut out in that case,
7976 and Exim is allowed to try to write to its normal log files.
7982 1. smtp_return_error_details was not giving details for temporary sender
7983 or receiver verification errors.
7985 2. Diagnose a configuration error if two authenticators have the same public
7988 3. Exim used not to create the message log file for a message until the first
7989 delivery attempt. This could be confusing when incoming messages were held
7990 for policy or load reasons. The message log file is now created at the time
7991 the message is received, and an initial "Received" line is written to it.
7993 4. The automatically generated man page for command line options had a minor
7994 bug that caused no ill effects; however, a more serious problem was that
7995 the procedure for building the man page automatically didn't always
7996 operate. Consequently, release 4.20 contains an out-of-date version. This
7997 shouldn't happen again.
7999 5. When building Exim with embedded Perl support, the script that builds the
8000 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8001 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8003 6. The freeze_tell option was not being used for messages that were frozen on
8004 arrival, either by an ACL or by local_scan().
8006 7. Added the smtp_incomplete_transaction log selector.
8008 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8009 was accepting AUTH without a new EHLO.
8011 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8012 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8015 10. Logging of TCP/IP connections (when configured) now happens in the main
8016 daemon process instead of the child process, so that the TCP/IP connection
8017 count is more accurate (but it can never be perfect).
8019 11. The use of "drop" in a nested ACL was not being handled correctly in the
8020 outer ACL. Now, if condition failure induced by the nested "drop" causes
8021 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8022 or "require"), the connection is dropped.
8024 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8025 that yield "discard" can now be used with an "accept" or a "discard" verb,
8026 but an error is generated for any others (because I can't see a useful way
8027 to define what should happen).
8029 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8030 are now processed in the same way as lines in the Exim configuration file.
8031 In particular, continuation lines are supported.
8033 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8035 15. Added -ti meaning -t -i.
8037 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8038 domains, and warn by logging if others are found.
8040 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8041 returned from GIFCONF when Exim is trying to find the list of interfaces on
8042 a host. The code in os.c has been modified to copy each ifreq to an aligned
8043 structure in all cases.
8045 Also, in some cases, the returned ifreq's were being copied to a 'struct
8046 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8047 means the last couple of bytes of an IPv6 address could be chopped if the
8048 ifreq contained only a normal sockaddr (14 bytes storage).
8050 18. Named domain lists were not supported in the hosts_treat_as_local option.
8051 An entry such as +xxxx was not recognized, and was treated as a literal
8054 19. Ensure that header lines added by a DATA ACL are included in the reject log
8055 if the ACL subsequently rejects the message.
8057 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8058 MD5 (which is deprecated).
8060 21. When testing a filter file using -bf, Exim was writing a message when it
8061 took the sender from a "From " line in the message, but it was not doing so
8062 when it took $return_path from a Return-Path: header line. It now does.
8064 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8065 with a valid header line field name (a series of printing characters
8066 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8068 23. Changed "disc" in the source to "disk" to conform to the documentation and
8069 the book and for uniformity.
8071 24. Ignore Sendmail's -Ooption=value command line item.
8073 25. When execve() failed while trying to run a command in a pipe transport,
8074 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8075 could be confused with a return value of 69 from the command itself. This
8076 has been changed to 127, the value the shell returns if it is asked to run
8077 a non-existent command. The wording for the related log line suggests a
8078 non-existent command as the problem.
8080 26. If received_header_text expands to an empty string, do not add a Received:
8081 header line to the message. (Well, it adds a token one on the spool, but
8082 marks it "old" so that it doesn't get used or transmitted.)
8084 27. Installed eximstats 1.28 (addition of -nt option).
8086 28. There was no check for failure on the call to getsockname() in the daemon
8087 code. This can fail if there is a shortage of resources on the system, with
8088 ENOMEM, for example. A temporary error is now given on failure.
8090 29. Contrary to the C standard, it seems that in some environments, the
8091 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8092 program. Exim now does this explicitly; it affects the formatting of
8093 timestamps using strftime().
8095 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8096 complaints. I've now initialized all the variables, to avoid this.
8098 32. Header lines added by a system filter were not being "seen" during
8099 transport-time rewrites.
8101 33. The info_callback() function passed to OpenSSL is set up with type void
8102 (*)(SSL *, int, int), as described somewhere. However, when calling the
8103 function (actually a macro) that sets it up, the type void(*)() is
8104 expected. I've put in a cast to prevent warnings from picky compilers.
8106 34. If a DNS black list lookup found a CNAME record, but there were no A
8107 records associated with the domain it pointed at, Exim crashed.
8109 35. If a DNS black list lookup returned more than one A record, Exim ignored
8110 all but the first. It now scans all returned addresses if a particular IP
8111 value is being sought. In this situation, the contents of the
8112 $dnslist_value variable are a list of all the addresses, separated by a
8115 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8116 to accept a host name and all its aliases if the forward lookup for any of
8117 them yielded the IP address of the incoming connection. Now it accepts only
8118 those names whose forward lookup yields the correct IP address. Any other
8119 names are discarded. This closes a loophole whereby a rogue DNS
8120 administrator could create reverse DNS records to break through a
8121 wildcarded host restriction in an ACL.
8123 37. If a user filter or a system filter that ran in a subprocess used any of
8124 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8125 the wrong values were passed to the pipe command ($thisaddress had the
8126 value of $0, $0 had the value of $1, etc). This bug was introduced by
8127 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8129 38. Improved the line breaking for long SMTP error messages from ACLs.
8130 Previously, if there was no break point between 40 and 75 characters, Exim
8131 left the rest of the message alone. Two changes have been made: (a) I've
8132 reduced the minimum length to 35 characters; (b) if it can't find a break
8133 point between 35 and 75 characters, it looks ahead and uses the first one
8134 that it finds. This may give the occasional overlong line, but at least the
8135 remaining text gets split now.
8137 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8138 file descriptors might be low, and that setting 1000 would always raise it.
8139 It turns out that in some environments, the limit is already over 1000 and
8140 that lowering it causes trouble. So now Exim takes care not to decrease it.
8142 40. When delivering a message, the value of $return_path is set to $sender_
8143 address at the start of routing (routers may change the value). By an
8144 oversight, this default was not being set up when an address was tested by
8145 -bt or -bv, which affected the outcome if any router or filter referred to
8148 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8149 log only when "message" or "log_message" are set. However, if one of the
8150 conditions was an address verification, or a call to a nested ACL, the
8151 messages generated by the underlying test were being passed through. This
8152 no longer happens. The underlying message is available in $acl_verify_
8153 message for both "message" and "log_message" expansions, so it can be
8154 passed through if needed.
8156 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8157 new expansion $bh_ to give the encoded byte string without charset
8158 translation. Translation happens only if iconv() is available; HAVE_ICONV
8159 indicates this at build time. HEADERS_CHARSET gives the charset to
8160 translate to; headers_charset can change it in the configuration, and
8161 "headers charset" can change it in an individual filter file.
8163 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8164 that creates RFC 2047 encoded "words" labels them as that charset instead
8165 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8166 expansion operator; (ii) when Exim creates a From: line for a local
8167 message; (iii) when a header line is rewritten to include a "phrase" part.
8169 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8170 buggy, causing it to skip the first lines of messages whose message ID
8171 ended in 'D'. This would not have bitten before Exim release 4.14, because
8172 message IDs were unlikely to end in 'D' before then. The effect was to have
8173 incorrect size information for certain domains.
8175 45. #include "config.h" was missing at the start of the crypt16.c module. This
8176 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8179 46. If there was a timeout during a "random" callout check, Exim treated it as
8180 a failure of the random address, and carried on sending RSET and the real
8181 address. If the delay was just some slowness somewhere, the response to the
8182 original RCPT would be taken as a response to RSET and so on, causing
8183 mayhem of various kinds.
8185 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8186 when I implemented it. It didn't allow for the fact that some option values
8187 may legitimately be negative (e.g. size_addition), and it didn't even do
8188 the right test for positive values.
8190 48. Domain names in DNS records are case-independent. Exim always looks them up
8191 in lower case. Some resolvers return domain names in exactly the case they
8192 appear in the zone file, that is, they may contain uppercase letters. Not
8193 all resolvers do this - some return always lower case. Exim was treating a
8194 change of case by a resolver as a change of domain, similar to a widening
8195 of a domain abbreviation. This triggered its re-routing code and so it was
8196 trying to route what was effectively the same domain again. This normally
8197 caused routing to fail (because the router wouldn't handle the domain
8198 twice). Now Exim checks for this case specially, and just changes the
8199 casing of the domain that it ultimately uses when it transmits the message
8202 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8205 50. If a filter generated a file delivery with a non-absolute name (possible if
8206 no home directory exists for the router), the forbid_file option was not
8209 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8210 the existing equality matching.
8212 52. Exim was using ints instead of ino_t variables in some places where it was
8213 dealing with inode numbers.
8215 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8216 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8217 called TMPDIR, and if it finds it is different, it changes its value.
8219 54. The smtp_printf() function is now made available to local_scan() so
8220 additional output lines can be written before returning. There is also an
8221 smtp_fflush() function to enable the detection of a dropped connection.
8222 The variables smtp_input and smtp_batched_input are exported to
8225 55. Changed the default runtime configuration: the message "Unknown user"
8226 has been removed from the ACL, and instead placed on the localuser router,
8227 using the cannot_route_message feature. This means that any verification
8228 failures that generate their own messages won't get overridden. Similarly,
8229 the "Unrouteable address" message that was in the ACL for unverifiable
8230 relay addresses has also been removed.
8232 56. Added hosts_avoid_esmtp to the smtp transport.
8234 57. The exicyclog script was not checking for the esoteric option
8235 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8236 will work only if exicyclog is run under the appropriate euid.
8238 58. Following a discussion on the list, the rules by which Exim recognises line
8239 endings on incoming messages have been changed. The -dropcr and drop_cr
8240 options are now no-ops, retained only for backwards compatibility. The
8241 following line terminators are recognized: LF CRLF CR. However, special
8242 processing applies to CR:
8244 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8245 nor a local message in the state where . is a terminator.
8247 (ii) If a bare CR is encountered in a header line, an extra space is added
8248 after the line terminator so as not to end the header. The reasoning
8249 behind this is that bare CRs in header lines are most likely either
8250 to be mistakes, or people trying to play silly games.
8252 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8253 was being incorrectly given as 18 bytes larger than it should have been.
8254 This is a VOB (very old bug).
8256 60. This may never have affected anything current, but just in case it has:
8257 When the local host is found other than at the start of a list of hosts,
8258 the local host, those with the same MX, and any that follow, are discarded.
8259 When the list in question was part of a longer list of hosts, the following
8260 hosts (not currently being processed) were also being discarded. This no
8261 longer happens. I'm not sure if this situation could ever has previously
8264 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8267 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8268 header. This is something that is recommended in a new Internet Draft, and
8269 is something that is documented as being done by Sendmail. There are two
8270 possible values. For messages generated by the autoreply transport, Exim
8273 Auto-Submitted: auto-replied
8275 whereas for all other generated messages (e.g. bounces) it adds
8277 Auto-Submitted: auto-generated
8279 63. The "personal" condition in filters now includes a test for the
8280 Auto-Submitted: header. If it contains the string "auto-" the message it
8281 not considered personal.
8283 64. Added rcpt_include_affixes as a generic transport option.
8285 65. Added queue_only_override (default true).
8287 66. Added the syslog_duplication option.
8289 67. If what should have been the first header line of a message consisted of
8290 a space followed by a colon, Exim was mis-interpreting it as a header line.
8291 It isn't of course - it is syntactically invalid and should therefore be
8292 treated as the start of the message body. The misbehaviour could have
8293 caused a number of strange effects, including loss of data in subsequent
8294 header lines, and spool format errors.
8296 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8297 client host had authenticated. This control can now be exercised by an ACL
8298 for more flexibility.
8300 69. By default, callouts do not happen when testing with -bh. There is now a
8301 variant, -bhc, which does actually run the callout code, including
8302 consulting and updating the callout cache.
8304 70. Added support for saslauthd authentication, courtesy of Alexander
8307 71. If statvfs() failed on the spool or log directories while checking their
8308 size for availability, Exim confusingly gave the error "space shortage".
8309 Furthermore, in debugging mode it crashed with a floating point exception.
8310 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8311 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8312 serious problem, Exim now writes to the main and panic logs when this
8313 happens, with details of the failure. It then refuses to accept the
8314 incoming message, giving the message "spool directory problem" or "log
8315 directory problem" with a 421 code for SMTP messages.
8317 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8318 0, 1, and 2 exist, because some OS complain for execs without them (see
8319 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8320 descriptors. However, the code omitted to check that the open succeeded,
8321 causing mysterious errors if for some reason the permissions on /dev/null
8322 got screwed. Now Exim writes a message to the main and panic logs, and
8323 bombs out if it can't open /dev/null.
8325 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8326 interact so that it is all more flexible. It is supposed to remain
8327 backwards compatible. Also added extra_local_interfaces.
8329 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8330 to bomb out with an assertion failure - to the client this appears as a
8331 connection drop. This problem occurs in the part of the code that was taken
8332 from the Samba project. Fortunately, the assertion is in a very simple
8333 function, so I have fixed this by reproducing the function inline in the
8334 one place where it is called, and arranging for authentication to fail
8335 instead of killing the process with assert().
8337 75. The SPA client code was not working when the server requested OEM rather
8338 than Unicode encoding.
8340 76. Added code to make require_files with a specific uid setting more usable in
8341 the case where statting the file as root fails - usually a non-root-mounted
8342 NFS file system. When this happens and the failure is EACCES, Exim now
8343 forks a subprocess and does the per-uid checking as the relevant uid.
8345 77. Added process_log_path.
8347 78. If log_file_path was not explicitly set, a setting of check_log_space or
8348 check_log_inodes was ignored.
8350 79. If a space check for the spool or log partitions fails, the incident is now
8351 logged. Of course, in the latter case the data may get lost...
8353 80. Added the %p formatting code to string_format() so that it can be used to
8354 print addresses in debug_print(). Adjusted all the address printing in the
8355 debugging in store.c to use %p rather than %d.
8357 81. There was a concern that a line of code in smtp_in.c could overflow a
8358 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8359 initially expressed, the concern was not well-founded, because trailing
8360 spaces are removed early. However, if the trailing spaces were followed by
8361 a NULL, they did not get removed, so the overflow was possible. Two fixes
8364 (a) I re-wrote the offending code in a cleaner fashion.
8365 (b) If an incoming SMTP command contains a NULL character, it is rejected
8368 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8369 runs initgroups(), so that if the Exim user is in any additional groups,
8370 they will be used during message reception.
8376 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob