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
73 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
74 after reception to before a subsequent reception. This should
75 mean slightly faster delivery, and also confirmation of reception
78 JH/02 Move from using the pcre library to pcre2. The former is no longer
79 being developed or supported (by the original developer).
81 JH/03 Constification work in the filters module required a major version
82 bump for the local-scan API. Specifically, the "headers_charset"
83 global which is visible via the API is now const and may therefore
84 not be modified by local-scan code.
86 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
87 sendfile() didi not account for the way the ClamAV driver code called it.
89 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
90 time check was being done for every character; replace that with one
93 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
94 sent was prefixed with a length byte.
96 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
97 RFC 5321. Previously Dovecot (at least) would log errors during
100 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
101 in 4.61 without a reason logged. Binaries will be bigger, which might
102 matter on diskspace-constrained systems, but debug is easier.
104 JH/09 Fix macro-definition during "-be" expansion testing. The move to
105 write-protected store for macros had not accounted for these runtime
106 additions; fix by removing this protection for "-be" mode.
108 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
109 handing out large-numbered file descriptors, violating the usual Unix
110 assumption (and required by Posix) that the lowest possible number will be
111 allocated by the kernel when a new one is needed. In the daemon, and any
112 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
113 are not useable for FD_SET() [and hence select()] and overwrite the stack.
114 Assorted crashes happen.
116 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
117 main-section options or in a connect ACL, the value from the first ever
118 connection was never replaced for subsequent connections. Found by
121 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
122 though only once PCRE2 was introduced: the memory accounting used under
123 debug offset allocations by an int, giving a hard trap in early startup.
124 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
126 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
127 with underbars is given. The write-protection of configuration introduced
128 in 4.95 trapped when normalisation was applied to an option not needing
131 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
133 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
134 failing to close the certificates directory, every hour or any time it
137 JH/16 Debugging initiated by an ACL control now continues through into routing
138 and transport processes. Previously debugging stopped any time Exim
139 re-execs, or for processing a queued message.
141 JH/17 The "expand" debug selector now gives more detail, specifically on the
142 result of expansion operators and items.
144 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
145 bad comparison between the option value and the name of the file to
146 be included was done, and a mismatch was wrongly identified.
147 4.88 to 4.95 are affected.
149 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
151 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
152 supplied as the db file-pair basename. Previously, if a directory
153 path was given, for example via the autoreply "once" option, the DB
154 file.pag and file.dir files would be created in that directory's
157 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
158 "taint" log_selector. These were previously deprecated.
160 JH/22 Fix static address-list lookups to properly return the matched item.
161 Previously only the domain part was returned.
163 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
164 the call into OpenSSL to send a TLS Close was being repeated; this
165 resulted in the library waiting for the peer's Close. If that was never
166 sent we waited forever. Fix by tracking send calls.
168 JH/24 The ${run} expansion item now expands its command string elements after
169 splitting. Previously it was before; the new ordering makes handling
170 zero-length arguments simpler. The old ordering can be obtained by
171 appending a new option "preexpand", after a comma, to the "run".
173 JH/25 Taint-check exec arguments for transport-initiated external processes.
174 Previously, tainted values could be used. This affects "pipe", "lmtp" and
175 "queryprogram" transport, transport-filter, and ETRN commands.
176 The ${run} expansion is also affected: in "preexpand" mode no part of
177 the command line may be tainted, in default mode the executable name
180 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
181 the the facility was not passed across execs, and only the first message
182 passed over a connection could use BDAT; any further ones using DATA.
184 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
185 uses $sending_ip_address and an interface is specified.
186 Previously any use of the local address in the EHLO name disabled
187 PIPECONNECT, the common case being to use the rDNS of it.
189 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
190 resumption. Previously verify failed because no certificate status is
191 passed on the wire for the restarted session. Fix by using the recorded
192 ocsp status of the stored session for the new connection.
194 JH/29 TLS resumption: the key for session lookup in the client now includes
195 more info that a server could potentially use in configuring a TLS
196 session, avoiding oferring mismatching sessions to such a server.
197 Previously only the server IP was used.
199 JH/30 Fix string_copyn() for limit greater than actual string length.
200 Previously the copied amount was the limit, which could result in a
201 overlapping memcpy for newly allocated destination soon after a
202 source string shorter than the limit. Found/investigated by KM.
204 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
205 close; it may be needed for a subsequent connection. This caused a
206 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
208 JH/32 Fix CHUNKING for a second message on a connection when the first was
209 rejected. Previously we did not reset the chunking-offered state, and
210 erroneously rejected the BDAT command. Investigation help from
213 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
214 an empty address. Previously the expansion returned an error.
216 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
217 proxy. Previously these were misparsed, leading to paniclog entries.
223 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
224 related applications. Previously an "H" was used where available info
225 says that "M" should be, so change to match.
227 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
228 as arguments, so an implementation trying to copy these into a local
229 buffer was taking a taint-enforcement trap. Fix by using dynamically
230 created buffers. Similar fix for radius expansion condition.
232 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
233 reasonable, eg. to count headers. Fix by using dynamically created
234 buffers rather than a local. Do similar fixes for ACL actions "dcc",
235 "log_reject_target", "malware" and "spam"; the arguments are expanded
236 so could be handling tainted values.
238 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
239 broken the (no-op) support for this sendmail command. Restore it
240 to doing nothing, silently, and returning good status.
242 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
243 record path was given (or the default used) without a leading directory
244 path, an error occurred on trying to open it. Use the transport's working
247 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
248 transport. Previously it was the name on the DNS A-record; use instead
249 the head of the CNAME chain leading there (if there is one). This seems
250 to align better with RFC 6125.
252 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
253 smtp_accept_max_per_host allocated resources which were not released
254 when the limit was exceeded. This eventually crashed the daemon. Fix
255 by adding a release action in that path.
257 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
258 expanded; previously using tainted values was rejected. Fix by using
259 dynamically-created buffers.
261 JH/09 Relax restrictions on ACL verify condition needing access to message
262 headers. Previously they were only permitted in data and non-smtp ACLs;
263 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
264 not_blind, header_sender and header_names_ascii verification.
266 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
267 Previously a macro used one argument twice; when called with the
268 argument as an expression having side-effects, incorrect operation
269 resulted. Use an inlineable function.
271 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
272 held open for a verify callout. Previously this wan not accounted for
273 and a corrupt onward SMTP conversation resulted.
275 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
276 Previously they were embedded naively in the constructed address; when
277 needed, strip the quoting and quote the entire local_part.
278 Also make the inbound_srs expansion condition handle quoting.
280 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
281 excluded, not matching the documentation.
283 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
284 was given for the sqlite_dbfile a trap resulted.
286 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
287 "name" argument resulted in a trap. There is no reason to disallow such;
288 this was a coding error.
290 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
291 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
292 spent suspended, ignoring the POSIX definition. Previously we assumed
293 it did and a constant offset from real time could be used as a correction.
294 Change to using the same clock source for the start-of-message and the
295 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
296 exists, just to get a clock slightly more aligned to reality.
298 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
299 RFC says it is optional some validators care. The missing char was not
300 intended but triggered by a line-wrap alignment. Discovery and fix by
301 Guillaume Outters, hacked on by JH.
303 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
304 name being quoted was tainted a trap would be taken. Fix by using
305 dynamically created buffers. The routine could have been called by a
306 rewrite with the "h" flag, by using the "-F" command-line option, or
307 by using a "name=" option on a control=submission ACL modifier.
309 JH/19 SPF: change the Authentication-Results expansion component to give
310 smtp.helo when the sender domain is empty. Previously it gave
313 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
314 not in 127.0.0.0/8 to help in spotting list domains taken over by a
315 domain-parking registrar.
317 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
318 Previously when a whitespace character was specified it was not inserted
319 after removing the newline.
321 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
322 the domain part of the recipient address. This overrides any tls_sni
323 option set, which was previously used.
325 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
328 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
329 is_tainted() had an off-by-one error in the overenthusiastic direction.
330 Find and fix by Gavan. Although NetBSD is not a supported platform for
331 4.94 this bug could affect other platforms.
333 PP/01 Fix default prime selection to be consistent.
334 One path used ike23 still, instead of exim.dev.20160529.3; now both
335 execution flows will use the same DH primes (currently
336 exim.dev.20160529.3).
338 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
339 option in smtp transport, to match the documentation. Previously
340 verification was not being done in some cases where it should have been.
342 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
343 than one server was defined and depending on the platform memory layout
344 details, an internal consistency trap could be hit while walking the list
347 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
348 files. The value is used by the authresults expansion item. Previously
349 if this was used in a router or transport, a crash could result.
351 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
352 have a sufficient incidence of EINTR returns from select that an
353 interaction with other operations done by the main daemon loop exposed
354 a bug in the error-handling. This was benign apart from the log
357 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
358 for consistency with delivered "=>" and failed "**" lines. While we're
359 there, handle PRX and TFO.
361 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
362 applied. This resulted, if any header-line rewrite rules were configured,
363 in a panic-log triggerable by sending a message with a long address in
364 a header. Fix by increasing the arbitrary limit to larger than a single
365 (dewrapped) 5322 header line maximum size.
367 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
368 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
369 RFC 6648 which deprecates X- options in protocols as a general practice.
370 Changeover between the implementations is handled by the mechanisms
373 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
374 a remote which uses callout/hold. Previously the local was queued.
376 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
379 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
380 log line. The asterisk is now consistently appended to the remote IP
381 (and port, if given), and will also be provided on defer and fail log
382 lines. Previously it could be placed on the local IP if that was being
383 logged, and was only provided on delivery lines.
385 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
387 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
388 authenticator client_send option. Previously the next char, after a pair
389 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
390 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
391 documentation. There is still no way to get a leading ^ immediately
392 after a NUL (ie. for the password of a PLAIN method authenticator.
394 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
395 files. For bad sizes read, delete the record and whine to paniclog.
397 JH/38 When logging an AUTH failure, as server, do not include sensitive
398 information. Previously, the credentials would be included if given
399 as part of the AUTH command line and an ACL denied authentication.
401 JH/39 Bug 2691: fix $local_part_data. When the matching list element
402 referred to a file, bad data was returned. This likely also affected
405 JH/40 The gsasl authenticator now supports caching of the salted password
406 generated by the client-side implementation. This required the addition
407 of a new variable: $auth4.
409 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
410 left undeleted; the attempt to re-create it then failed - resulting in
411 the usual "SIGHUP tp have daemon reload configuration" to not work.
412 This affected any platform not supporting "abstract" Unix-domain
413 sockets (i.e. not Linux).
415 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
416 recipients" response to RCPT in a later response, with a 250. The
417 previous coding assumed this would not happen, and under PIPELINING
418 would result in both lost and duplicate recipients for a message.
420 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
421 Previously the weighting was incorrectly applied. Similar fix for socks
422 proxies. Found and fixed by Heiko Schlichting.
424 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
425 not handle sub-lists included using the +namedlist syntax. While
426 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
427 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
430 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
431 Previously the permanent pool was used, so the sore could not be freed.
432 This meant a connection with many messages would use continually-growing
435 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
436 per-pool so as not to waste too much space. Previously a constant size
437 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
438 DOS attacks harder. The cost is wasted memory use in the larger blocks.
440 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
441 same space repeatedly, and a smaller process footprint.
443 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
444 for GnuTLS. Previously the name started "%s" which, while not a bug,
445 looked as if if might be one.
447 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
448 "continued connection") make the $proxy_* variables available. Previously
449 the information was not passed across the exec() call for subsequent
450 transport executions. This also mean that the log lines for the
451 messages can show the proxy information.
453 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
454 receive time. With modern systems the difference is significant.
455 The historical behaviour can be restored by disabling (a new) log_selector
456 "queue_time_exclusive".
458 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
459 content) were specifically tested for. Now, also cover expansions
460 resulting in ACL names and inline ACL content.
462 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
463 making it unusable in complex expressions.
465 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
466 fail with a taint-check on the spoolfile name, and leave the message
469 HS/01 Enforce absolute PID file path name.
471 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
473 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
475 PP/02 Bug 2643: Correct TLS DH constants.
476 A missing NUL termination in our code-generation tool had led to some
477 incorrect Diffie-Hellman constants in the Exim source.
478 Reported by kylon94, code-gen tool fix by Simon Arlott.
480 PP/03 Impose security length checks on various command-line options.
481 Fixes CVE-2020-SPRSS reported by Qualys.
483 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
484 better. Reported by Qualys.
486 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
487 providing a particularly obnoxious sender full name.
490 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
492 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
495 PP/08 Change default for recipients_max from unlimited to 50,000.
497 PP/09 Fix security issue with too many recipients on a message (to remove a
498 known security problem if someone does set recipients_max to unlimited,
499 or if local additions add to the recipient list).
500 Fixes CVE-2020-RCPTL reported by Qualys.
502 PP/10 Fix security issue in SMTP verb option parsing
503 Fixes CVE-2020-EXOPT reported by Qualys.
505 PP/11 Fix security issue in BDAT state confusion.
506 Ensure we reset known-good where we know we need to not be reading BDAT
507 data, as a general case fix, and move the places where we switch to BDAT
508 mode until after various protocol state checks.
509 Fixes CVE-2020-BDATA reported by Qualys.
511 HS/03 Die on "/../" in msglog file names
513 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
514 the Exim runtime user are allowed to create files.
516 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
519 QS/03 When reading the output from interpreted forward files we do not
520 pass the pipe between the parent and the interpreting process to
521 executed child processes (if any).
523 QS/04 Always die if requested from internal logging, even is logging is
526 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
527 the API; compilation noo longer completes with DMARC support included.
528 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
529 been reported on other platforms.
531 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
533 JH/56 Make the majority of info read from config files readonly, for defence-in-
534 depth against exploits. Suggestion by Qualys.
535 Not supported on Solaris 10.
537 JH/57 Fix control=fakreject for a custom message containing tainted data.
538 Previously this resulted in a log complaint, due to a re-expansion present
539 since fakereject was originally introduced.
541 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
542 modifier was given, a loop resulted.
544 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
545 pipelined SMTP command followed the BDAT LAST then it would be
546 incorrectly treated as part of the message body, causing a verification
549 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
550 require looser validation rules than those for 821-level addresses,
551 which only permit IP addresses.
557 JH/01 Avoid costly startup code when not strictly needed. This reduces time
558 for some exim process initialisations. It does mean that the logging
559 of TLS configuration problems is only done for the daemon startup.
561 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
563 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
564 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
567 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
568 DKIM signing is being done. Previously a transport_filter would always
569 disable CHUNKING, falling back to traditional DATA.
571 JH/05 Regard command-line recipients as tainted.
573 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
575 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
576 PAM library frees one of the arguments given to it, despite the
577 documentation. Therefore a plain malloc must be used.
579 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
580 on-stack buffers were used, resulting in a taint trap when DSN information
581 copied from a received message was written into the buffer.
583 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
584 the ordering of its ARC headers. This caused a crash.
586 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
587 a new record was being constructed with information from the peer, a trap
590 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
591 installation would get error messages from DMARC verify, when it hit the
592 nonexistent file indicated by the default. Distros wanting DMARC enabled
593 should both provide the file and set the option.
594 Also enforce no DMARC verification for command-line sourced messages.
596 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
597 could, depending on the platform, hang at the STARTTLS response.
599 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
600 message on a connection. Previously if one message had ARC headers and
601 the following one did not, a crash could result when adding an
602 Authentication-Results: header.
604 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
605 Exim main code and Exim-related utities. The introduction of taint
606 tracking also did many adjustments to string handling. Since then, eximon
607 frequently terminated with an assert failure.
609 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
610 check for 452 responses. This slightly helps the inefficieny of doing
611 a large alias-expansion into a recipient-limited target. The max_rcpt
612 transport option still applies (and at the current default, will override
613 the new feature). The check is done for either cause of synch, and forces
614 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
615 connection. The new facility is not tunable at this time.
617 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
618 library live data was being used, so the results became garbage. Make
619 copies while it is still usable.
621 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
622 on delivery deferred (==) and failed (**) lines (if a delivery was
623 attemtped). Previously it was only on completion (=>) lines.
625 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
626 for the expansion of the server_scram_iter and server_scram_salt options.
628 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
629 are now specifically given a NO_DATA response without hitting the system
630 resolver. The library goes on to do the now-standard TXT lookup.
631 Use of dnsdb lookups is not affected.
633 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
634 only retrieve the errormessage once. Previously two calls to dlerror()
635 were used, and the second one (for mainlog/paniclog) retrieved null
638 JH/20 Taint checking: disallow use of tainted data for
639 - the appendfile transport file and directory options
640 - the pipe transport command
641 - the autoreply transport file, log and once options
642 - file names used by the redirect router (including filter files)
644 - paths used by single-key lookups
645 Previously this was permitted.
647 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
648 adjusted the size of a major service buffer; this failed because the
649 buffer was in use at the time. Change to a compile-time increase in the
650 buffer size, when this authenticator is compiled into exim.
652 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
653 previous fast-mode was untenable in the face of glibs using mmap to
654 support larger malloc requests.
656 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
657 New values supported, if defined on system where compiled:
658 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
659 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
661 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
662 running a limited number of proceses in parallel, a benefit is gained. The
663 amount varies with the platform hardware and load. The use of the option
664 queue_run_in_order means we cannot do this, as ordering becomes
667 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
668 had introduced a string-copy (for ensuring NUL-termination) which was not
669 appropriate for that case, which can include embedded NUL bytes in the
670 block of data. Investigation showed the copy to actually be needless, the
671 data being length-specified.
673 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
674 done during a receiving connection, and both used TLS, global info was
675 used rather than per-connection info for tracking the state of data
676 queued for transmission. This could result in a connection hang.
678 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
679 Previously, when delivering serveral messages down a single connection
680 only the first would provide a SIZE. This was due to the size information
681 not being properly tracked.
683 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
684 TAI (at 37 seconds currently), pretend to be in UTC for time-related
685 expansion and logging. Previously, spurious values such as a future
686 minute could be seen.
688 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
689 it could crash from a null-deref. This could also affect the
690 ${addresses: } operator and ${readsock } item.
692 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
693 message following a mime one, the variable was not reset.
695 JH/30 When an pipelined-connect fails at the first response, assume incorrect
696 cached capability (perhaps the peer reneged?) and immediately retry in
699 JH/31 Fix spurious detection of timeout while writing to transport filter.
701 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
702 an attempt to copy the string was made before checking it.
704 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
705 taint of the lookup key was maintained; we now regard the presence in the
706 filesystem as sufficient validation.
708 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
709 argument is supplied.
711 JH/35 The dsearch lookup now requires that the directory is an absolute path.
712 Previously this was not checked, and nonempty relative paths made an
713 access under Exim's current working directory.
715 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
716 Previously no event was raised.
718 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
719 parameter supplied by the sender MAIL FROM command. Previously it was
720 ignored, and only the check_spool_space option value for the required
723 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
724 the size of the signing public-key. Previously it was instead giving
725 the size of the signature hash.
727 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
728 the default. See the (new) dkim_verify_min_keysizes option.
730 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
731 and an ACL use a lookup for checking either the local_part or domain,
732 stale data could be accessed. Ensure that variable references are
733 dropped between messages.
735 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
736 by the client was not checked as pointing within response data before
737 being used. A malicious client could thus cause an out-of-bounds read and
738 possibly gain authentication. Fix by adding the check.
740 JH/42 Internationalisation: change the default for downconversion in the smtp
741 transport to be "if needed". Previously it was "as previously set" for
742 the message, which usually meant "if needed" for message-submission but
743 "no" for everything else. However, MTAs have been seen using SMTPUTF8
744 even when the envelope addresses did not need it, resulting in forwarding
745 failures to non-supporting MTAs. A downconvert in such cases will be
746 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
747 transport. The change does mean that addresses needing conversion will
748 be converted when previously a delivery failure would occur.
750 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
751 response was received it would be used unchecked in a fail-DSN, violating
752 standards on line-length limits. Truncate if needed.
754 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
755 doesn't work. (Additionally add a new main config option to configure the
762 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
763 side, to decode a TLS 1.3 packet capture.
765 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
766 Previously the default library behaviour applied, sending two, each in
769 JH/03 Debug output for ACL now gives the config file name and line number for
772 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
774 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
776 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
777 buffer overrun for (non-chunking) other transports.
779 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
780 TLS1.3, means that a server rejecting a client certificate is not visible
781 to the client until the first read of encrypted data (typically the
782 response to EHLO). Add detection for that case and treat it as a failed
783 TLS connection attempt, so that the normal retry-in-clear can work (if
784 suitably configured).
786 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
787 and/or domain. Found and fixed by Jason Betts.
789 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
790 configuration). If a CNAME target was not a wellformed name pattern, a
793 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
794 the OS reports them interleaved with other addresses.
796 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
797 used both for input and for a verify callout, both encrypted, SMTP
798 responses being sent by the server could be lost. This resulted in
799 dropped connections and sometimes bounces generated by a peer sending
802 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
803 string. Previously it was possible to cause undefined behaviour in a
804 library routine (usually a crash). Found by "zerons".
806 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
809 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
810 API was removed, so update to use the newer ones.
812 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
813 any timeout set, is taking a long time. Previously we would hang on to a
814 rotated logfile "forever" if the input was arriving with long gaps
815 (a previous attempt to fix addressed lack, for a long time, of initial
818 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
819 shared (NFS) environment. The length of the tempfile name is now
820 4 + 16 ("hdr.$message_exim_id") which might break on file
821 systems which restrict the file name length to lower values.
824 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
825 shared (NFS) environment.
827 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
828 did for all versions <4.90). Notably -M, -m, --invert, -I may be
831 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
832 on some platforms for bit 31.
834 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
835 to changes apparently associated with TLS1.3 handling some of the APIs
836 previously used were either nonfunctional or inappropriate. Strings
837 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
838 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
839 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
840 This affects log line X= elements, the $tls_{in,out}_cipher variables,
841 and the use of specific cipher names in the encrypted= ACL condition.
843 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
845 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
846 verification result was not updated unless hosts_require_ocsp applied.
848 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
849 queue_list_requires_admin set to false, non-admin users were denied the
852 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
853 directory-of-certs mode. Previously they were advertised despite the
856 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
857 A single TCP connection by a client will now hold a TLS connection open
858 for multiple message deliveries, by default. Previously the default was to
861 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
862 default. If built with the facility, DANE will be used. The facility
863 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
865 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
866 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
867 must be defined and you must still, unless you define DISABLE_TLS, manage
868 the the include-dir and library-file requirements that go with that
869 choice. Non-TLS builds are still supported.
871 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
874 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
875 default. If the platform supports and has the facility enabled, it will
876 be requested on all coneections.
878 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
879 controlled by the build-time option SUPPORT_PIPE_CONNECT.
881 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
883 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
884 success-DSN messages. Previously the From: header was always the default
885 one for these; the option was ignored.
887 JH/28 Fix the timeout on smtp response to apply to the whole response.
888 Previously it was reset for every read, so a teergrubing peer sending
889 single bytes within the time limit could extend the connection for a
890 long time. Credit to Qualsys Security Advisory Team for the discovery.
892 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
893 delivery address, which leaked information of the results of local
894 forwarding. Change to the original envelope recipient address, per
897 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
898 requested. Previously not bounce was generated and a log entry of
899 error ignored was made.
901 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
903 JH/32 Introduce a general tainting mechanism for values read from the input
904 channel, and values derived from them. Refuse to expand any tainted
905 values, to catch one form of exploit.
907 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
908 was unused and the unexpanded text used for the test. Found and
909 fixed by Ruben Jenster.
911 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
912 an attempt to use a TLS library read routine dereffed a nul pointer,
915 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
916 them in our smtp response.
918 JH/36 Have the general router option retry_use_local_part default to true when
919 any of the restrictive preconditions are set (to anything). Previously it
920 was only for check_local user. The change removes one item of manual
921 configuration which is required for proper retries when a remote router
922 handles a subset of addresses for a domain.
924 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
925 link count into consideration.
927 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
928 caused the extension of big_buffer, the following lines were ignored.
930 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
931 accordance with RFC 2308. Previously there was no expiry, so a longlived
932 receive process (eg. due to ACL delays) versus a short SOA value could
935 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
937 JH/39 Promote DMARC support to mainline.
939 JH/40 Bug 2452: Add a References: header to DSNs.
941 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
942 parameters. The relevant library call is documented as "Deprecated: This
943 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
944 3.6.0, DH parameters are negotiated following RFC7919."
946 HS/06 Change the default of dnssec_request_domains to "*"
948 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
949 carried on and emitted a BDAT command, even when PIPELINING was not
952 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
953 buffer was used for the filename, resulting in a trap when tainted
954 arguments (eg. $domain) were used.
956 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
957 recommended to avoid a possible server-load attack. The feature can be
958 re-enabled via the openssl_options main cofiguration option.
960 JH/45 local_scan API: documented the current smtp_printf() call. This changed
961 for version 4.90 - adding a "more data" boolean to the arguments.
962 Bumped the ABI version number also, this having been missed previously;
963 release versions 4.90 to 4.92.3 inclusive were effectively broken in
964 respect of usage of smtp_printf() by either local_scan code or libraries
965 accessed via the ${dlfunc } expansion item. Both will need coding
966 adjustment for any calls to smtp_printf() to match the new function
967 signature; a FALSE value for the new argument is always safe.
969 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
970 the file-offset (which the Linux syscall does, and exim expects); this
971 resulted in an indefinite loop.
973 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
974 to do ARC verification. The Authentication-Results: header line added
975 by the configuration then had no ARC item.
977 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
978 an incorrect mode was used for reading the result, resulting in it being
985 JH/01 Remove code calling the customisable local_scan function, unless a new
986 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
988 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
989 non-signal-safe functions being used.
991 JH/03 Bug 2269: When presented with a received message having a stupidly large
992 number of DKIM-Signature headers, disable DKIM verification to avoid
993 a resource-consumption attack. The limit is set at twenty.
995 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
996 report of oldest_pass in ${authres } in consequence, and separate out
997 some descriptions of reasons for verification fail.
999 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1000 files in the spool were present and unlocked. A queue-runner could spot
1001 them, resulting in a duplicate delivery. Fix that by doing the unlock
1002 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1003 add more error-checking on spoolfile handling while that code is being
1006 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1007 No known attacks, no CVE, this is defensive hardening.
1009 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1010 a queue-runner could start a delivery while other operations were ongoing.
1011 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1012 Found and investigated by Tim Stewart. Fix by using the open message data
1013 file handle rather than opening another, and not locally closing it (which
1014 releases a lock) for that case, while creating the temporary .eml format
1015 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1017 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1018 $sender_verify_failure/$recipient_verify_failure to "random".
1020 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1023 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1024 Previously this would segfault.
1026 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1029 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1030 like zero, since the resolver should be doing this for us, But we need one
1031 as a CNAME but no MX presence gets the CNAME returned; we need to check
1032 that doesn't point to an MX to declare it "no MX returned" rather than
1033 "error, loop". A new main option is added so the older capability of
1034 following some limited number of chain links is maintained.
1036 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1038 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1039 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1040 now-deprecated earlier definitions used only specified the range up to TLS
1041 1.2 (in the older-version library docs).
1043 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1045 JH/15 Rework TLS client-side context management. Stop using a global, and
1046 explicitly pass a context around. This enables future use of TLS for
1047 connections to service-daemons (eg. malware scanning) while a client smtp
1048 connection is using TLS; with cutthrough connections this is quite likely.
1050 JH/16 Fix ARC verification to do AS checks in reverse order.
1052 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1054 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1055 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1056 Previously the "utf8" would be re-prepended for every additional message.
1058 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1059 Previously thery were accepted, resulting in issues when attempting to
1060 forward messages to a non-supporting MTA.
1062 PP/02 Let -n work with printing macros too, not just options.
1064 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1065 one parent address was copied, and bogus data was used at delivery-logging
1066 time. Either a crash (after delivery) or bogus log data could result.
1067 Discovery and analysis by Tim Stewart.
1069 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1070 Previously if the string ended mid-character, we did not insert the
1071 promised '?' replacement.
1073 PP/04 Documentation: current string operators work on bytes, not codepoints.
1075 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1076 should pack well giving a smaller memory footprint so better caching and
1077 therefore performance. Group the declarations where this can't be done so
1078 that the byte-sized flag variables are not interspersed among pointer
1079 variables, giving a better chance of good packing by the compiler.
1081 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1082 non-null, to avoid issues with sites running BATV. Previously reports were
1083 sent with an empty envelope sender so looked like bounces.
1085 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1086 The ignore_error flag wasn't being returned from the filter subprocess so
1087 was not set for later routers. Investigation and fix by Matthias Kurz.
1089 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1090 and a msg:complete for the whole, when a message is manually removed using
1091 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1093 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1094 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1095 hope that is portable enough.
1097 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1098 requiring a known-CA anchor certificate; make it now rely entirely on the
1099 TLSA as an anchor. Checking the name on the leaf cert against the name
1100 on the A-record for the host is still done for TA (but not for EE mode).
1102 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1103 included in delivery lines for non-proxied connections, when compiled with
1104 SUPPORT_SOCKS and running with proxy logging enabled.
1106 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1107 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1108 move the existing event to fire before the normal logging of message
1109 failure so that custom logging is bracketed by normal logging.
1111 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1112 msg:fail:internal event. Developement by Matthias Kurz.
1114 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1115 far too small for todays use of crypto signatures stored there. Go all
1116 the way to the max DNS message size of 64kB, even though this might be
1117 overmuch for IOT constrained device use.
1119 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1120 copy a string over itself. The library routine is documented as not
1121 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1123 JH/32 For main options check_spool_space and check_inode_space, where the
1124 platform supports 64b integers, support more than the previous 2^31 kB
1125 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1126 the previous G, M, k.
1128 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1129 $authenticated_fail_id variable on authentication failure. Previously
1132 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1133 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1134 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1135 GNUTLS_SEC_PARAM_MEDIUM.
1137 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1138 side. Previously we would continue as if no SNI had been received.
1140 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1141 "<" character, which should be a list-separator specification, we walked
1142 off past the nul-terimation.
1144 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1145 causes) even when the retry time is not yet met. Previously they were
1146 not, meaning that when (say) an account was over-quota and temp-rejecting,
1147 and multiple senders' messages were queued, only one sender would get
1148 notified on each configured delay_warning cycle.
1150 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1152 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1153 cipher-suites, an error can be left on the stack even for a succeeding
1154 accept; this results in impossible error messages when a later operation
1157 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1158 return error codes indicating retry. Under TLS1.3 this becomes required.
1160 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1161 it only wrote the new authenticators, resulting in a lack of tracking of
1162 peer changes of ESMTP extensions until the next cache flush.
1164 JH/41 Fix the loop reading a message header line to check for integer overflow,
1165 and more-often against header_maxsize. Previously a crafted message could
1166 induce a crash of the recive process; now the message is cleanly rejected.
1168 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1169 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1175 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1176 When redis_servers is set to a list of > 1 element, and the Redis servers
1177 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1178 case of MOVED into a DEFER case instead, thus moving the query onto the
1179 next server in the list. For a cluster of N elements, all N servers must
1180 be defined in redis_servers.
1182 GF/02 Catch and remove uninitialized value warning in exiqsumm
1183 Check for existence of @ARGV before looking at $ARGV[0]
1185 JH/01 Replace the store_release() internal interface with store_newblock(),
1186 which internalises the check required to safely use the old one, plus
1187 the allocate and data copy operations duplicated in both (!) of the
1188 extant use locations.
1190 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1191 modifier. This matches the restriction on the commandline.
1193 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1194 Previously only the last row was returned.
1196 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1197 we assumed that tags in the header were well-formed, and parsed the
1198 element content after inspecting only the first char of the tag.
1199 Assumptions at that stage could crash the receive process on malformed
1202 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1203 While running the DKIM ACL we operate on the Permanent memory pool so that
1204 variables created with "set" persist to the DATA ACL. Also (at any time)
1205 DNS lookups that fail create cache records using the Permanent pool. But
1206 expansions release any allocations made on the current pool - so a dnsdb
1207 lookup expansion done in the DKIM ACL releases the memory used for the
1208 DNS negative-cache, and bad things result. Solution is to switch to the
1209 Main pool for expansions.
1210 While we're in that code, add checks on the DNS cache during store_reset,
1211 active in the testsuite.
1212 Problem spotted, and debugging aided, by Wolfgang Breyha.
1214 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1215 When none of the hosts presented to a transport match an already-open
1216 connection, close it and proceed with the list. Previously we would
1217 queue the message. Spotted by Lena with Yahoo, probably involving
1220 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1221 Previously a spurious "250 OK id=" response was appended to the proper
1224 JH/08 The "support for" informational output now, which built with Content
1225 Scanning support, has a line for the malware scanner interfaces compiled
1226 in. Interface can be individually included or not at build time.
1228 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1229 by the template makefile "src/EDITME". The "STREAM" support for an older
1230 ClamAV interface method is removed.
1232 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1233 rows affected is given instead).
1235 JH/11 The runtime Berkeley DB library version is now additionally output by
1236 "exim -d -bV". Previously only the compile-time version was shown.
1238 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1239 SMTP connection. Previously, when one had more recipients than the
1240 first, an abortive onward connection was made. Move to full support for
1241 multiple onward connections in sequence, handling cutthrough connection
1242 for all multi-message initiating connections.
1244 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1245 routers. Previously, a multi-recipient message would fail to match the
1246 onward-connection opened for the first recipient, and cause its closure.
1248 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1249 a timeout on read on a GnuTLS initiating connection, resulting in the
1250 initiating connection being dropped. This mattered most when the callout
1251 was marked defer_ok. Fix to keep the two timeout-detection methods
1254 JH/15 Relax results from ACL control request to enable cutthrough, in
1255 unsupported situations, from error to silently (except under debug)
1256 ignoring. This covers use with PRDR, frozen messages, queue-only and
1259 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1261 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1262 metadata, resulting in a crash in free().
1264 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1265 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1266 Broken also in d185889f4, with init system revamp.
1268 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1269 Previously we abruptly closed the connection after reading a malware-
1270 found indication; now we go on to read the "scan ok" response line,
1273 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1274 ACL. Previously, a crash would result.
1276 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1277 macro text after a replacement (previously it was only once per line) and
1278 by skipping builtin macros when searching for an uppercase lead character.
1280 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1281 for the build is renamed.
1283 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1284 was allocated for every new TLS startup, meaning one per message. Fix
1285 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1287 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1288 reported the original. Fix to report (as far as possible) the ACL
1289 result replacing the original.
1291 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1292 OpenSSL. Certificate information is loaded for every new TLS startup,
1293 and the resources needed to be freed.
1295 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1297 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1300 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1301 DATA response info to the (existing) per-recipient response info for
1302 the "C=" log element. It can have useful tracking info from the
1303 destination system. Patch from Simon Arlott.
1305 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1306 length value. Previously this would segfault.
1308 HS/02 Support Avast multiline protoocol, this allows passing flags to
1309 newer versions of the scanner.
1311 JH/28 Ensure that variables possibly set during message acceptance are marked
1312 dead before release of memory in the daemon loop. This stops complaints
1313 about them when the debug_store option is enabled. Discovered specifically
1314 for sender_rate_period, but applies to a whole set of variables.
1315 Do the same for the queue-runner and queue-list loops, for variables set
1316 from spool message files. Do the same for the SMTP per-message loop, for
1317 certain variables indirectly set in ACL operations.
1319 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1320 as a multi-recipient message from a mailinglist manager). The coding had
1321 an arbitrary cutoff number of characters while checking for more input;
1322 enforced by writing a NUL into the buffer. This corrupted long / fast
1323 input. The problem was exposed more widely when more pipelineing of SMTP
1324 responses was introduced, and one Exim system was feeding another.
1325 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1326 receiving system, and refused recipients seen by the sending system
1327 (propating to people being dropped from mailing lists).
1328 Discovered and pinpointed by David Carter.
1330 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1331 replaced by the ${authresults } expansion.
1333 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1335 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1336 allows proper process termination in container environments.
1338 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1339 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1341 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1342 and "err_perm", deprecated since 4.83 when the RFC-defined words
1343 "temperror" and "permerror" were introduced.
1345 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1346 transport-filters or DKIM-signing. The restriction was lost in the
1347 consolidation of verify-callout and delivery SMTP handling.
1348 Extend the restriction to also cover ARC-signing.
1350 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1351 in defer=pass mode supply a 450 to the initiator. Previously the message
1354 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1355 tls_require_ciphers is used as before.
1357 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1358 "pass_unscanned". Only tmpfails from the scanner are written to
1359 the paniclog, as they may require admin intervention (permission
1360 denied, license issues). Other scanner errors (like decompression
1361 bombs) do not cause a paniclog entry.
1363 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1364 Previously it was possible to log spurious information in receive log
1367 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1368 triggered odd behaviour from Outlook Express clients.
1370 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1373 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1374 since the IETF WG has not yet settled on that versus the original
1375 "bare" representation.
1377 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1378 Previously the millisecond value corrupted the output.
1379 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1380 corrupted the output.
1386 JH/01 Rework error string handling in TLS interface so that the caller in
1387 more cases is responsible for logging. This permits library-sourced
1388 string to be attached to addresses during delivery, and collapses
1389 pairs of long lines into single ones.
1391 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1392 during configuration. Wildcards are allowed and expanded.
1394 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1395 This permits better logging.
1397 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1398 do not close it down and have the child transport start it up again on
1399 the passed-on TCP connection. Instead, proxy the child (and any
1400 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1401 affected: the continued delivery log lines do not have any DNSSEC, TLS
1402 Certificate or OCSP information. TLS cipher information is still logged.
1404 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1405 identical IP addresses on different listening ports. Will also affect
1408 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1409 add noisy ifdef guards to special-case this sillyness.
1410 Patch from Bernd Kuhls.
1412 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1413 than 255 are no longer allowed.
1415 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1416 Disable the session-cache too, which might reduce our load. Since we
1417 currrectly use a new context for every connection, both as server and
1418 client, there is no benefit for these.
1419 GnuTLS appears to not support tickets server-side by default (we don't
1420 call gnutls_session_ticket_enable_server()) but client side is enabled
1421 by default on recent versions (3.1.3 +) unless the PFS priority string
1424 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1425 <https://reproducible-builds.org/specs/source-date-epoch/>.
1427 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1428 the check for any unsuccessful recipients did not notice the limit, and
1429 erroneously found still-pending ones.
1431 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1432 MSG_MORE. Only in-clear (not on TLS connections).
1434 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1435 transport-filter is involved we can buffer the headers in memory for
1436 creating the signature, and read the spool data file once for the
1437 signature and again for transmission.
1439 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1440 4.77 as the kernel support then wasn't solid, having issues in 64bit
1441 mode. Now, it's been long enough. Add support for FreeBSD also.
1443 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1444 case where the routing stage had gathered several addresses to send to
1445 a host before calling the transport for the first, we previously failed
1446 to close down TLS in the old transport process before passing the TCP
1447 connection to the new process. The new one sent a STARTTLS command
1448 which naturally failed, giving a failed delivery and bloating the retry
1449 database. Investigation and fix prototype from Wolfgang Breyha.
1451 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1452 false-negatives in the check that the sender had not preempted a response
1453 or prompt from Exim (running as a server), due to that code's lack of
1454 awareness of the SMTP input buffering.
1456 PP/04 Add commandline_checks_require_admin option.
1457 Exim drops privileges sanely, various checks such as -be aren't a
1458 security problem, as long as you trust local users with access to their
1459 own account. When invoked by services which pass untrusted data to
1460 Exim, this might be an issue. Set this option in main configuration
1461 AND make fixes to the calling application, such as using `--` to stop
1464 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1465 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1466 responses to those, into a single TLS record each way (this usually means
1467 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1470 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1471 affects you only if you're dancing at the edge of the param size limits.
1472 If you are, and this message makes sense to you, then: raise the
1473 configured limit or use OpenSSL 1.1. Nothing we can do for older
1476 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1477 cmdline element to get the documented default one. Previously it was
1480 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1481 are unsuable for non-message-specific reasons - eg. connection timeout,
1482 banner-time rejection.
1484 JH/16 Fix logging of delivery remote port, when specified by router, under
1487 PP/06 Repair manualroute's ability to take options in any order, even if one
1488 is the name of a transport.
1491 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1493 JH/17 Change the list-building routines interface to use the expanding-string
1494 triplet model, for better allocation and copying behaviour.
1496 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1497 Previously it was constructed the first time a possibly-matching string
1498 was met in the configuration file input during startup; now it is done
1501 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1502 compatible one, to avoid the (poorly documented) possibility of a config
1503 file in the working directory redirecting the DB files, possibly correpting
1504 some existing file. CVE-2017-10140 assigned for BDB.
1506 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1507 cache-hot. Previously, although the result was properly cached, the
1508 initial verify call returned a defer.
1510 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1511 the main verify for receipient in uncached-mode.
1513 JH/22 Retire historical build files to an "unsupported" subdir. These are
1514 defined as "ones for which we have no current evidence of testing".
1516 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1517 if present. Previously it was ignored.
1519 JH/24 Start using specified-initialisers in C structure init coding. This is
1520 a C99 feature (it's 2017, so now considered safe).
1522 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1523 if was a fixed-sized field and bitmask ops via macros; it is now more
1526 PP/07 GitHub PR 56: Apply MariaDB build fix.
1527 Patch provided by Jaroslav Škarvada.
1529 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1530 during Coverity cleanups [4.87 JH/47]
1531 Diagnosis and fix provided by Michael Fischer v. Mollard.
1533 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1534 the right size to place the terminating semicolon on its own folded
1535 line, the header hash was calculated to an incorrect value thanks to
1536 the (relaxed) space the fold became.
1538 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1539 and confused the parent.
1541 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1542 which could crash as a result. This could lead to undeliverable messages.
1544 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1547 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1548 matters on fast-turnover and PID-randomising systems, which were getting
1549 out-of-order delivery.
1551 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1552 a possibly-overlapping copy. The symptom was that "Remote host closed
1553 connection in response to HELO" was logged instead of the actual 4xx
1556 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1557 Previously only that bufferd was discarded, resulting in SYMTP command
1560 JH/32 DKIM: when a message has multiple signatures matching an identity given
1561 in dkim_verify_signers, run the dkim acl once for each. Previously only
1562 one run was done. Bug 2189.
1564 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1565 "panic the current process" to "deliberately defer". The panic log is
1566 still written with the problem list name; the mail and reject logs now
1567 get a temp-reject line for the message that was being handled, saying
1568 something like "domains check lookup or other defer". The SMTP 451
1569 message is still "Temporary local problem".
1571 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1572 A crafted sequence of BDAT commands could result in in-use memory beeing
1573 freed. CVE-2017-16943.
1575 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1576 from SMTP input. Previously it was always done; now only done for DATA
1577 and not BDAT commands. CVE-2017-16944.
1579 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1580 to the message (such as an overlong header line). Previously this was
1581 not done and we did not exit BDAT mode. Followon from the previous item
1582 though a different problem.
1588 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1589 than -2003 did; needs libidn2 in addition to libidn.
1591 JH/02 The path option on a pipe transport is now expanded before use.
1593 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1594 Patch provided by "Björn", documentation fix added too.
1596 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1597 missing a wire-to-host endian conversion.
1599 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1600 close after a BDAT command line could be taken as a following command,
1601 giving a synch failure. Fix by only checking for synch immediately
1602 before acknowledging the chunk.
1604 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1605 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1606 macro. Patches provided by Josh Soref.
1608 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1609 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1610 by IANA as a service extension. However, John Klensin suggests that we
1613 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1614 the dkim code may be unix-mode line endings rather than smtp wire-format
1615 CRLF, so prepend a CR to any bare LF.
1617 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1618 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1620 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1621 the first were themselves being wrongly included in the feed into dkim
1622 processing; with most chunk sizes in use this resulted in an incorrect
1623 body hash calculated value.
1625 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1626 DKIM signature block, for verification. Although advised against by
1627 standards it is specifically not ruled illegal.
1629 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1631 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1632 missing a body hash (the bh= tag).
1634 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1635 It seems that HAProxy sends the Proxy Protocol information in clear and
1636 only then does a TLS startup, so do the same.
1638 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1639 TCP connections (such as for Spamd) unless the daemon successfully set
1640 Fast Open mode on its listening sockets. This fixes breakage seen on
1641 too-old kernels or those not configured for Fast Open, at the cost of
1642 requiring both directions being enabled for TFO, and TFO never being used
1643 by non-daemon-related Exim processes.
1645 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1646 endings, at least on the first header line. Try to canonify any that get
1647 past that check, despite the cost.
1649 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1650 now limited to an arbitrary five deep, while parsing addresses with the
1651 strip_excess_angle_brackets option enabled.
1653 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1654 instead leave the unprompted TLS handshake in socket buffer for the
1655 TLS library to consume.
1657 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1659 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1661 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1662 region in one operation, for speed, and this leaves assigned pointers
1663 dangling. Add checks run only under the testsuite which checks all
1664 variables at a store-reset and panics on a dangling pointer; add code
1665 explicitly nulling out all the variables discovered. Fixes one known
1666 bug: a transport crash, where a dangling pointer for $sending_ip_address
1667 originally assigned in a verify callout, is re-used.
1669 PP/06 Drop '.' from @INC in various Perl scripts.
1671 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1673 PP/08 Reduce a number of compilation warnings under clang; building with
1674 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1675 should be warning-free.
1677 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1679 HS/01 Fix portability problems introduced by PP/08 for platforms where
1680 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1682 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1683 chunk. This allows us to accept broken chunked messages. We need a more
1684 general solution here.
1686 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1687 already-broken messages in the queue.
1689 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1691 JH/19 Fix reference counting bug in routing-generated-address tracking.
1697 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1698 supports it and a size is available (ie. the sending peer gave us one).
1700 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1701 years of being deprecated). The replacements are the ACLs
1702 acl_smtp_mime and acl_not_smtp_mime.
1704 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1705 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1706 or even an in-clear connection were permitted. Now, if the host lookup
1707 was dnssec and dane was requested then the host is only used if the
1708 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1709 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1710 if one fails this test.
1711 This means that a poorly-configured remote DNS will make it incommunicado;
1712 but it protects against a DNS-interception attack on it.
1714 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1715 non-noisy when a race steals the message being considered.
1717 JH/05 If main configuration option tls_certificate is unset, generate a
1718 self-signed certificate for inbound TLS connections.
1720 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1721 in rewrites and routers.
1723 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1724 and logged a warning sing 4.83; now they are a configuration file error.
1726 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1727 (lacking @domain). Apply the same qualification processing as RCPT.
1729 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1731 JH/10 Support ${sha256:} applied to a string (as well as the previous
1734 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1735 a cutthrough deliver is pending, as we always want to make a connection.
1736 This also avoids re-routing the message when later placing the cutthrough
1737 connection after a verify cache hit.
1738 Do not update it with the verify result either.
1740 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1741 when routing results in more than one destination address.
1743 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1744 signing (which inhibits the cutthrough capability). Previously only
1745 the presence of an option was tested; now an expansion evaluating as
1746 empty is permissible (obviously it should depend only on data available
1747 when the cutthrough connection is made).
1749 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1750 the relevant preceding SMTP command did not note the pipelining mode.
1752 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1753 Previously they were not counted.
1755 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1756 as one having no matching records. Previously we deferred the message
1757 that needed the lookup.
1759 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1760 distinguished as "(=".
1762 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1763 for missing MX records. Previously it only worked for missing A records.
1765 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1767 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1768 after the data-go-ahead and data-ack. Patch from Jason Betts.
1770 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1771 even for a "none" policy. Patch from Tony Meyer.
1773 JH/22 Fix continued use of a connection for further deliveries. If a port was
1774 specified by a router, it must also match for the delivery to be
1777 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1778 When one of the recipients of a message was unsuitable for the connection
1779 (has no matching addresses), we lost track of needing to mark it
1780 deferred. As a result mail would be lost.
1782 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1784 JH/25 Decoding ACL controls is now done using a binary search; the source code
1785 takes up less space and should be simpler to maintain. Merge the ACL
1786 condition decode tables also, with similar effect.
1788 JH/26 Fix problem with one_time used on a redirect router which returned the
1789 parent address unchanged. A retry would see the parent address marked as
1790 delivered, so not attempt the (identical) child. As a result mail would
1793 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1794 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1795 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1798 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1799 Main config options check_{log,spool}_{inodes,space} are now
1800 100 inodes, 10MB unless set otherwise in the configuration.
1802 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1803 Previously it only applied to the main-section connection policy
1806 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1808 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1809 by me. Added RFC7919 DH primes as an alternative.
1811 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1812 are not in the system include path.
1814 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1815 GnuTLS, when a session startup failed (eg because the client disconnected)
1816 Exim did stdio operations after fclose. This was exposed by a recent
1817 change which nulled out the file handle after the fclose.
1819 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1820 signed directly by the cert-signing cert, rather than an intermediate
1821 OCSP-signing cert. This is the model used by LetsEncrypt.
1823 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1825 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1826 an incoming connection.
1828 HS/02 Bug 1802: Do not half-close the connection after sending a request
1831 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1832 fallback to "prime256v1".
1834 JH/34 SECURITY: Use proper copy of DATA command in error message.
1835 Could leak key material. Remotely exploitable. CVE-2016-9963.
1841 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1842 and 3.4.4 - once the server is enabled to respond to an OCSP request
1843 it does even when not requested, resulting in a stapling non-aware
1844 client dropping the TLS connection.
1846 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1847 support variable-length bit vectors. No functional change.
1849 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1850 The I= interface field on outgoing lines is now after the H= remote
1851 host field, same as incoming lines. There is a separate
1852 outgoing_interface log selector which allows you to disable the
1855 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1856 If not running log_selector +smtp_connection the mainlog would be held
1857 open indefinitely after a "too many connections" event, including to a
1858 deleted file after a log rotate. Leave the per net connection logging
1859 leaving it open for efficiency as that will be quickly detected by the
1860 check on the next write.
1862 HS/01 Bug 1671: Fix post transport crash.
1863 Processing the wait-<transport> messages could crash the delivery
1864 process if the message IDs didn't exist for some reason. When
1865 using 'split_spool_directory=yes' the construction of the spool
1866 file name failed already, exposing the same netto behaviour.
1868 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1869 mime_regex ACL conditions.
1871 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1872 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1873 to HELO, local diagnostic string.
1875 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1876 log line to debug. Even when configured with a tls authenticator many
1877 client connections are expected to not authenticate in this way, so
1878 an authenticate fail is not an error.
1880 HS/02 Add the Exim version string to the process info. This way exiwhat
1881 gives some more detail about the running daemon.
1883 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1884 matter for fast-change records such as DNSBLs.
1886 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1887 retry record, even if constant. There may be multiple transports with
1888 different interface settings and the retry behaviour needs to be kept
1891 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1893 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1895 JH/10 Bug 840: fix log_defer_output option of pipe transport
1897 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1898 hosts_randomize. This matters a lot when combined with mua_wrapper.
1900 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1901 ${quote_pgsql:<string>} operator.
1903 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1905 JH/14 Change header file name for freeradius-client. Relevant if compiling
1906 with Radius support; from the Gentoo tree and checked under Fedora.
1908 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1910 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1911 option transport_filter, meaning no filtering.
1913 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1915 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1916 defaults to "*" (all hosts). The variable is now available when not built
1917 with TLS, default unset, mainly to enable keeping the testsuite sane.
1918 If a server certificate is not supplied (via tls_certificate) an error is
1919 logged, and clients will find TLS connections fail on startup. Presumably
1920 they will retry in-clear.
1921 Packagers of Exim are strongly encouraged to create a server certificate
1922 at installation time.
1924 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1925 with the $config_file variable.
1927 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1928 in transport context, after the attempt, and per-recipient. The latter type
1929 is per host attempted. The event data is the error message, and the errno
1930 information encodes the lookup type (A vs. MX) used for the (first) host,
1931 and the trailing two digits of the smtp 4xx response.
1933 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1934 to write to mainlog (or rejectlog, paniclog) in the window between file
1935 creation and permissions/ownership being changed. Particularly affects
1936 installations where exicyclog is run as root, rather than exim user;
1937 result is that the running daemon panics and dies.
1939 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1941 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1942 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1943 "pri" and "weight". Note that the previous implicit priority given by the
1944 list order is no longer honoured.
1946 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1947 for DKIM processing.
1949 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1950 by defining SUPPORT_SOCKS.
1952 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1953 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1954 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1955 variables are renamed to proxy_{local,external}_{address,port}.
1957 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1958 for a build by defining SUPPORT_I18N
1960 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1961 of the query string, and make ${quote_redis:} do that quoting.
1963 JH/29 Move Events support from Experimental to mainline, enabled by default
1964 and removable for a build by defining DISABLE_EVENT.
1966 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1968 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1969 cached by the daemon.
1971 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1972 by defining LOOKUP_REDIS. The libhiredis library is required.
1974 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1975 keys are given for lookup.
1977 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1978 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1979 only supported when built with TLS support. The PolarSSL SHA routines
1980 are still used when the TLS library is too old for convenient support.
1982 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1983 openssl_options), for security. OpenSSL forces this from version 1.1.0
1984 server-side so match that on older versions.
1986 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1987 allocation for $value could be released as the expansion processing
1988 concluded, but leaving the global pointer active for it.
1990 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1991 and to use the domains and local_parts ACL conditions.
1993 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1994 incorrectly not doubled on cutthrough transmission, hence seen as a
1995 body-termination at the receiving system - resulting in truncated mails.
1996 Commonly the sender saw a TCP-level error, and retransmitted the message
1997 via the normal store-and-forward channel. This could result in duplicates
1998 received - but deduplicating mailstores were liable to retain only the
1999 initial truncated version.
2001 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2003 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2005 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2006 we're in there, support oversigning also; bug 1309.
2008 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2010 HS/04 Add support for keep_environment and add_environment options.
2012 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2013 either intentional arithmetic overflow during PRNG, or testing config-
2016 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2017 delivery resulted in actual delivery. Cancel cutthrough before DATA
2020 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2021 rejects a recipient: pass the reject to the originator.
2023 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2024 Many were false-positives and ignorable, but it's worth fixing the
2027 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2028 for the new environment-manipulation done at startup. Move the routines
2029 from being local to tls.c to being global via the os.c file.
2031 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2032 an extract embedded as result-arg for a map, the first arg for extract
2033 is unavailable so we cannot tell if this is a numbered or keyed
2034 extraction. Accept either.
2040 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2043 JH/02 The smtp transport option "multi_domain" is now expanded.
2045 JH/03 The smtp transport now requests PRDR by default, if the server offers
2048 JH/04 Certificate name checking on server certificates, when exim is a client,
2049 is now done by default. The transport option tls_verify_cert_hostnames
2050 can be used to disable this per-host. The build option
2051 EXPERIMENTAL_CERTNAMES is withdrawn.
2053 JH/05 The value of the tls_verify_certificates smtp transport and main options
2054 default to the word "system" to access the system default CA bundle.
2055 For GnuTLS, only version 3.0.20 or later.
2057 JH/06 Verification of the server certificate for a TLS connection is now tried
2058 (but not required) by default. The verification status is now logged by
2059 default, for both outbound TLS and client-certificate supplying inbound
2062 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2065 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2066 Status Notification (bounce) messages are now MIME format per RFC 3464.
2067 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2068 under the control of the dsn_advertise_hosts option, and routers may
2069 have a dsn_lasthop option.
2071 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2072 default, modifiable by a malware= option. The list separator for
2073 the options can now be changed in the usual way. Bug 68.
2075 JH/10 The smtp_receive_timeout main option is now expanded before use.
2077 JH/11 The incoming_interface log option now also enables logging of the
2078 local interface on delivery outgoing connections.
2080 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2081 if the interface and destination host and port all match.
2083 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2086 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2087 Patch from Andrew Lewis.
2089 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2090 now supports optional time-restrictions, weighting, and priority
2091 modifiers per server. Patch originally by <rommer@active.by>.
2093 JH/16 The spamd_address main option now supports a mixed list of local
2094 and remote servers. Remote servers can be IPv6 addresses, and
2095 specify a port-range.
2097 JH/17 Bug 68: The spamd_address main option now supports an optional
2098 timeout value per server.
2100 JH/18 Bug 1581: Router and transport options headers_add/remove can
2101 now have the list separator specified.
2103 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2106 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2109 JH/21 Support for the A6 type of dns record is withdrawn.
2111 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2112 rather than the verbs used.
2114 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2115 from 255 to 1024 chars.
2117 JH/24 Verification callouts now attempt to use TLS by default.
2119 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2120 are generic router options now. The defaults didn't change.
2122 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2123 Original patch from Alexander Shikoff, worked over by JH.
2125 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2126 files if ZCAT_COMMAND is not executable.
2128 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2130 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2132 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2133 Normally benign, it bites when the pair was led to by a CNAME;
2134 modern usage is to not canonicalize the domain to a CNAME target
2135 (and we were inconsistent anyway for A-only vs AAAA+A).
2137 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2139 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2140 when evaluating $sender_host_dnssec.
2142 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2143 $sender_helo_dnssec variable.
2145 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2147 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2149 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2151 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2152 documented as working, but never had. Support all but $spam_report.
2154 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2155 added for tls authenticator.
2157 HS/03 Add perl_taintmode main config option
2163 TL/01 When running the test suite, the README says that variables such as
2164 no_msglog_check are global and can be placed anywhere in a specific
2165 test's script, however it was observed that placement needed to be near
2166 the beginning for it to behave that way. Changed the runtest perl
2167 script to read through the entire script once to detect and set these
2168 variables, reset to the beginning of the script, and then run through
2169 the script parsing/test process like normal.
2171 TL/02 The BSD's have an arc4random API. One of the functions to induce
2172 adding randomness was arc4random_stir(), but it has been removed in
2173 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2174 function when detected.
2176 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2177 cause callback expansion.
2179 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2180 syntax errors in an expansion can be treated as a string instead of
2181 logging or causing an error, due to the internal use of bool_lax
2182 instead of bool when processing it.
2184 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2185 server certificates when making smtp deliveries.
2187 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2189 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2191 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2193 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2194 Merged patch from Sebastian Wiedenroth.
2196 JH/05 Fix results-pipe from transport process. Several recipients, combined
2197 with certificate use, exposed issues where response data items split
2198 over buffer boundaries were not parsed properly. This eventually
2199 resulted in duplicates being sent. This issue only became common enough
2200 to notice due to the introduction of connection certificate information,
2201 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2203 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2204 size buffer was used, resulting in syntax errors when an expansion
2207 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2208 version 3.3.6 or later.
2210 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2211 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2212 both become "event_action", the variables become $event_name, $event_data
2213 and $event_defer_errno. There is a new variable $verify_mode, usable in
2214 routers, transports and related events. The tls:cert event is now also
2215 raised for inbound connections, if the main configuration event_action
2218 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2219 early OCSP support, but no stapling (appears to be less than 1.0.0).
2221 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2222 server certificate names available under the smtp transport option
2223 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2226 JH/10 Time-related extraction expansions from certificates now use the main
2227 option "timezone" setting for output formatting, and are consistent
2228 between OpenSSL and GnuTLS compilations. Bug 1541.
2230 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2231 encoded parameter in the incoming message. Bug 1558.
2233 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2234 include certificate info, eximon was claiming there were spoolfile
2237 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2239 JH/14 Log delivery-related information more consistently, using the sequence
2240 "H=<name> [<ip>]" wherever possible.
2242 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2243 are problematic for Debian distribution, omit them from the release
2246 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2248 JH/16 Fix string representation of time values on 64bit time_t architectures.
2251 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2252 output list separator was used.
2257 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2258 checkers that were complaining about end of non-void function with no
2261 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2262 This was a regression introduced in 4.83 by another bugfix.
2264 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2266 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2267 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2273 TF/01 Correctly close the server side of TLS when forking for delivery.
2275 When a message was received over SMTP with TLS, Exim failed to clear up
2276 the incoming connection properly after forking off the child process to
2277 deliver the message. In some situations the subsequent outgoing
2278 delivery connection happened to have the same fd number as the incoming
2279 connection previously had. Exim would try to use TLS and fail, logging
2280 a "Bad file descriptor" error.
2282 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2283 utilities have not been installed.
2285 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2286 temporary space as the ACL may create new global variables.
2288 TL/01 LDAP support uses per connection or global context settings, depending
2289 upon the detected version of the libraries at build time.
2291 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2292 to extract and use the src ip:port in logging and expansions as if it
2293 were a direct connection from the outside internet. PPv2 support was
2294 updated based on HAProxy spec change in May 2014.
2296 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2298 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2299 Properly escape header and check for NULL return.
2301 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2304 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2306 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2307 characters in header names, implemented as a verify condition.
2308 Contributed by Michael Fischer v. Mollard.
2310 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2311 results permerror and temperror. Previous values are deprecated but
2312 still accepted. In a future release, err_perm and err_temp will be
2313 completely removed, which will be a backward incompatibility if the
2314 ACL tests for either of these two old results. Patch contributed by
2315 user bes-internal on the mailing list.
2317 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2319 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2320 selectors, in both main and reject logs.
2322 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2325 JH/07 Add malware type "sock" for talking to simple daemon.
2327 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2329 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2330 routers/transports under cutthrough routing.
2332 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2333 numbers. Touch up "bool" conditional to keep the same definition.
2335 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2337 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2339 JH/12 Expand items in router/transport headers_add or headers_remove lists
2340 individually rather than the list as a whole. Bug 1452.
2342 Required for reasonable handling of multiple headers_ options when
2343 they may be empty; requires that headers_remove items with embedded
2344 colons must have them doubled (or the list-separator changed).
2346 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2347 view the policy declared in the DMARC record. Currently, $dmarc_status
2348 is a combined value of both the record presence and the result of the
2351 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2353 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2354 dnslookup router and the smtp transport (applying to the forward
2357 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2358 of ldap servers used for a specific lookup. Patch provided by Heiko
2361 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2362 New variable $lookup_dnssec_authenticated for observability.
2364 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2365 Patch submitted by Lars Timman.
2367 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2369 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2370 Requires trusted mode and valid format message id, aborts otherwise.
2371 Patch contributed by Heiko Schlichting.
2373 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2374 certextract with support for various fields. Bug 1358.
2376 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2377 is requested by default, modifiable by smtp transport option
2380 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2381 operate on certificate variables to give certificate fingerprints
2382 Also new ${sha256:cert_variable}.
2384 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2386 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2387 Christian Aistleitner.
2389 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2391 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2392 file. Patch from Wolfgang Breyha.
2394 JH/25 Expand the coverage of the delivery $host and $host_address to
2395 client authenticators run in verify callout. Bug 1476.
2397 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2398 align with daemon_smtp_ports. Bug 72.
2400 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2401 support and error reporting did not work properly.
2403 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2404 and is readable. Patch from Andrew Colin Kissa.
2406 TL/14 Enhance documentation of ${run expansion and how it parses the
2407 commandline after expansion, particularly in the case when an
2408 unquoted variable expansion results in an empty value.
2410 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2412 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2413 by renaming away from it. Addresses GH issue 15, reported by
2416 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2418 TL/15 SECURITY: prevent double expansion in math comparison functions
2419 (can expand unsanitized data). Not remotely exploitable.
2426 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2428 PP/02 Make -n do something, by making it not do something.
2429 When combined with -bP, the name of an option is not output.
2431 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2434 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2435 $sender_host_name and config options to manage this, and basic check
2438 PP/05 DSCP support for outbound connections and control modifier for inbound.
2440 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2441 (Only plugin which currently uses this is kerberos4, which nobody should
2442 be using, but we should make it available and other future plugins might
2443 conceivably use it, even though it would break NAT; stuff *should* be
2444 using channel bindings instead).
2446 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2447 name; added for Sendmail compatibility; requires admin caller.
2448 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2449 just ignore it); requires trusted caller.
2450 Also parse but ignore: -Ac -Am -X<logfile>
2453 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2455 TL/02 Add +smtp_confirmation as a default logging option.
2457 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2458 Patch by Magnus Holmgren from 2007-02-20.
2460 TL/04 Bugzilla 1281 - Spec typo.
2461 Bugzilla 1283 - Spec typo.
2462 Bugzilla 1290 - Spec grammar fixes.
2464 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2466 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2468 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2469 mailing list by by Dmitry Isaikin.
2471 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2473 JH/02 Support "G" suffix to numbers in ${if comparisons.
2475 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2477 NM/01 Bugzilla 1197 - Spec typo
2478 Bugzilla 1196 - Spec examples corrections
2480 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2482 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2483 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2486 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2487 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2489 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2490 "acl {{name}{arg}...}", and optional args on acl condition
2493 JH/05 Permit multiple router/transport headers_add/remove lines.
2495 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2497 JH/07 Avoid using a waiting database for a single-message-only transport.
2498 Performance patch from Paul Fisher. Bugzilla 1262.
2500 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2503 JH/09 Add $headers_added variable, with content from use of ACL modifier
2504 add_header (but not yet added to the message). Bugzilla 199.
2506 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2507 Pulled from Bugzilla 817 by Wolfgang Breyha.
2509 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2511 (nb: this is the same fix as in Exim 4.80.1)
2513 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2516 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2519 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2521 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2522 advertise SMTP AUTH mechanism to us, instead of a generic
2523 protocol violation error. Also, make Exim more robust to bad
2524 data from the Dovecot auth socket.
2526 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2528 When a queue runner is handling a message, Exim first routes the
2529 recipient addresses, during which it prunes them based on the retry
2530 hints database. After that it attempts to deliver the message to
2531 any remaining recipients. It then updates the hints database using
2534 So if a recipient address works intermittently, it can get repeatedly
2535 deferred at routing time. The retry hints record remains fresh so the
2536 address never reaches the final cutoff time.
2538 This is a fairly common occurrence when a user is bumping up against
2539 their storage quota. Exim had some logic in its local delivery code
2540 to deal with this. However it did not apply to per-recipient defers
2541 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2543 This change adds a proper retry rule check during routing so that the
2544 final cutoff time is checked against the message's age. We only do
2545 this check if there is an address retry record and there is not a
2546 domain retry record; this implies that previous attempts to handle
2547 the address had the retry_use_local_parts option turned on. We use
2548 this as an approximation for the destination being like a local
2549 delivery, as in LMTP.
2551 I suspect this new check makes the old local delivery cutoff check
2552 redundant, but I have not verified this so I left the code in place.
2554 TF/02 Correct gecos expansion when From: is a prefix of the username.
2556 Test 0254 submits a message to Exim with the header
2560 When I ran the test suite under the user fanf2, Exim expanded
2561 the header to contain my full name, whereas it should have added
2562 a Resent-Sender: header. It erroneously treats any prefix of the
2563 username as equal to the username.
2565 This change corrects that bug.
2567 GF/01 DCC debug and logging tidyup
2568 Error conditions log to paniclog rather than rejectlog.
2569 Debug lines prefixed by "DCC: " to remove any ambiguity.
2571 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2573 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2574 Bug spotted by Jeremy Harris; was flawed since initial commit.
2575 Would have resulted in OCSP responses post-SNI triggering an Exim
2576 NULL dereference and crash.
2578 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2580 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2581 Bug detection, analysis and fix by Samuel Thibault.
2582 Bugzilla 1331, Debian bug #698092.
2584 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2586 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2587 Server implementation by Todd Lyons, client by JH.
2588 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2589 config variable "prdr_enable" controls whether the server
2590 advertises the facility. If the client requests PRDR a new
2591 acl_data_smtp_prdr ACL is called once for each recipient, after
2592 the body content is received and before the acl_smtp_data ACL.
2593 The client is controlled by both of: a hosts_try_prdr option
2594 on the smtp transport, and the server advertisement.
2595 Default client logging of deliveries and rejections involving
2596 PRDR are flagged with the string "PRDR".
2598 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2599 fclose(). Diagnosis by Todd Lyons.
2601 PP/17 Update configure.default to handle IPv6 localhost better.
2602 Patch by Alain Williams (plus minor tweaks).
2605 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2606 This is now consistent with GnuTLS, and is now documented: the
2607 previous undocumented portable approach to treating the option as
2608 unset was to force an expansion failure. That still works, and
2609 an empty string is now equivalent.
2611 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2612 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2613 not performing validation itself.
2615 PP/20 Added force_command boolean option to pipe transport.
2616 Patch from Nick Koston, of cPanel Inc.
2618 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2621 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2623 PP/21 Fix eximon continuous updating with timestamped log-files.
2624 Broken in a format-string cleanup in 4.80, missed when I repaired the
2625 other false fix of the same issue.
2626 Report and fix from Heiko Schlichting.
2629 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2630 Report from Prashanth Katuri.
2632 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2633 It's SecureTransport, so affects any MacOS clients which use the
2634 system-integrated TLS libraries, including email clients.
2636 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2637 using a MIME ACL for non-SMTP local injection.
2638 Report and assistance in diagnosis by Warren Baker.
2640 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2642 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2644 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2645 last id that failed so it may be referenced in subsequent ACL's.
2647 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2650 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2651 ldap library initialization, allowing self-signed CA's to be
2652 used. Also properly sets require_cert option later in code by
2653 using NULL (global ldap config) instead of ldap handle (per
2654 session). Bug diagnosis and testing by alxgomz.
2656 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2657 the src/util/ subdirectory.
2659 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2660 renamed to Transport Post Delivery Action by Jeremy Harris, as
2663 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2664 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2665 redis_servers = needs to be configured which will be used by the redis
2666 lookup. Patch from Warren Baker, of The Packet Hub.
2668 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2670 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2671 hostname or reverse DNS when processing a host list. Used suggestions
2672 from multiple comments on this bug.
2674 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2676 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2677 Missed a few lines, added it to make the runtest require no keyboard
2680 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2681 contains upper case chars. Make router use caseful_local_part.
2683 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2684 support when GnuTLS has been built with p11-kit.
2690 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2692 This, or similar/improved, will also be change PP/11 of 4.82.
2698 PP/01 Handle short writes when writing local log-files.
2699 In practice, only affects FreeBSD (8 onwards).
2700 Bugzilla 1053, with thanks to Dmitry Isaikin.
2702 NM/01 Bugzilla 949 - Documentation tweak
2704 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2707 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2709 PP/02 Implemented gsasl authenticator.
2711 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2713 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2714 `pkg-config foo` for cflags/libs.
2716 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2717 with rest of GSASL and with heimdal_gssapi.
2719 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2720 `pkg-config foo` for cflags/libs for the TLS implementation.
2722 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2723 properties get this fed in as external SSF. A number of robustness
2724 and debugging improvements to the cyrus_sasl authenticator.
2726 PP/08 cyrus_sasl server now expands the server_realm option.
2728 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2729 Patch by Jeremy Harris.
2731 PP/10 Added dbmjz lookup type.
2733 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2735 PP/12 MAIL args handles TAB as well as SP, for better interop with
2736 non-compliant senders.
2737 Analysis and variant patch by Todd Lyons.
2739 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2740 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2741 Patch from Dirk Mueller <dmueller@suse.com>
2743 PP/13 tls_peerdn now print-escaped for spool files.
2744 Observed some $tls_peerdn in wild which contained \n, which resulted
2745 in spool file corruption.
2747 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2748 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2749 or write after TLS renegotiation, which otherwise led to messages
2752 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2753 as a tracking header (ie: a signed header comes before the signature).
2754 Patch from Wolfgang Breyha.
2756 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2757 comma-sep list; embedded commas doubled.
2759 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2761 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2763 Report and patch from Dmitry Banschikov.
2765 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2766 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2767 failures. We appear to now support TLS1.1+ with Exim.
2769 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2770 lets Exim select keys and certificates based upon TLS SNI from client.
2771 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2772 before an outbound SMTP session. New log_selector, +tls_sni.
2774 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2775 NULL dereference. Report and patch from Alun Jones.
2777 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2778 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2779 Not seeing resolver debug output on NetBSD, but suspect this is a
2780 resolver implementation change.
2782 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2783 Left warnings. Added "eximon gdb" invocation mode.
2785 PP/21 Defaulting "accept_8bitmime" to true, not false.
2787 PP/22 Added -bw for inetd wait mode support.
2789 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2790 locate the relevant includes and libraries. Made this the default.
2792 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2793 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2795 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2796 This may cause build issues on older platforms.
2798 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2799 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2800 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2801 Added SNI support via GnuTLS too.
2802 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2804 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2806 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2807 Applied second patch from Janne, implementing suggestion to default
2808 multiple-strings-in-record handling to match SPF spec.
2810 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2812 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2813 read-only, out of scope).
2814 Patch from Wolfgang Breyha, report from Stuart Northfield.
2816 PP/29 Fix three issues highlighted by clang analyser static analysis.
2817 Only crash-plausible issue would require the Cambridge-specific
2818 iplookup router and a misconfiguration.
2819 Report from Marcin Mirosław.
2821 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2823 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2824 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2825 As part of this, removing so much warning spew let me fix some minor
2826 real issues in debug logging.
2828 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2829 assignment on my part. Fixed.
2831 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2832 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2833 Janne Snabb (who went above and beyond: thank you).
2835 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2836 string otherwise requires a connection and a bunch more work and it's
2837 relatively easy to get wrong. Should also expose TLS library linkage
2840 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2841 64-bit ${eval} (JH/03).
2843 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2844 GNU libc to support some of the 64-bit stuff, should not lead to
2845 conflicts. Defined before os.h is pulled in, so if a given platform
2846 needs to override this, it can.
2848 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2849 protection layer was required, which is not implemented.
2850 Bugzilla 1254, patch from Wolfgang Breyha.
2852 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2853 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2854 tls_dhparam take prime identifiers. Also unbreak combination of
2855 OpenSSL+DH_params+TLSSNI.
2857 PP/39 Disable SSLv2 by default in OpenSSL support.
2863 PP/01 Solaris build fix for Oracle's LDAP libraries.
2864 Bugzilla 1109, patch from Stephen Usher.
2866 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2868 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2871 TF/02 Fix a couple more cases where we did not log the error message
2872 when unlink() failed. See also change 4.74-TF/03.
2874 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2875 lock up or crash if it happened to be inside a call to libc when it
2876 got a SIGUSR1 from exiwhat.
2878 The SIGUSR1 handler appends the current process status to the process
2879 log which is later printed by exiwhat. It used to use the general
2880 purpose logging code to do this, but several functions it calls are
2881 not safe for signals.
2883 The new output code in the SIGUSR1 handler is specific to the process
2884 log, and simple enough that it's easy to inspect for signal safety.
2885 Removing some special cases also simplifies the general logging code.
2886 Removing the spurious timestamps from the process log simplifies
2889 TF/04 Improved ratelimit ACL condition.
2891 The /noupdate option has been deprecated in favour of /readonly which
2892 has clearer semantics. The /leaky, /strict, and /readonly update modes
2893 are mutually exclusive. The update mode is no longer included in the
2894 database key; it just determines when the database is updated. (This
2895 means that when you upgrade Exim will forget old rate measurements.)
2897 Exim now checks that the per_* options are used with an update mode that
2898 makes sense for the current ACL. For example, when Exim is processing a
2899 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2900 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2901 must specify per_mail/readonly. If you omit the update mode it defaults to
2902 /leaky where that makes sense (as before) or /readonly where required.
2904 The /noupdate option is now undocumented but still supported for
2905 backwards compatibility. It is equivalent to /readonly except that in
2906 ACLs where /readonly is required you may specify /leaky/noupdate or
2907 /strict/noupdate which are treated the same as /readonly.
2909 A useful new feature is the /count= option. This is a generalization
2910 of the per_byte option, so that you can measure the throughput of other
2911 aggregate values. For example, the per_byte option is now equivalent
2912 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2914 The per_rcpt option has been generalized using the /count= mechanism
2915 (though it's more complicated than the per_byte equivalence). When it is
2916 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2917 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2918 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2919 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2920 Note that using per_rcpt with a non-readonly update mode in more than
2921 one ACL will cause the recipients to be double-counted. (The per_mail
2922 and per_byte options don't have this problem.)
2924 The handling of very low rates has changed slightly. If the computed rate
2925 is less than the event's count (usually one) then this event is the first
2926 after a long gap. In this case the rate is set to the same as this event's
2927 count, so that the first message of a spam run is counted properly.
2929 The major new feature is a mechanism for counting the rate of unique
2930 events. The new per_addr option counts the number of different
2931 recipients that someone has sent messages to in the last time period. It
2932 behaves like per_rcpt if all the recipient addresses are different, but
2933 duplicate recipient addresses do not increase the measured rate. Like
2934 the /count= option this is a general mechanism, so the per_addr option
2935 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2936 example, measure the rate that a client uses different sender addresses
2937 with the options per_mail/unique=$sender_address. There are further
2938 details in the main documentation.
2940 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2942 TF/06 Removed a few PCRE remnants.
2944 TF/07 Automatically extract Exim's version number from tags in the git
2945 repository when doing development or release builds.
2947 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2948 Bugzilla 879. Patch from Paul Fisher.
2950 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2951 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2954 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2956 PP/05 Variable $av_failed, true if the AV scanner deferred.
2957 Bugzilla 1078. Patch from John Horne.
2959 PP/06 Stop make process more reliably on build failure.
2960 Bugzilla 1087. Patch from Heiko Schlittermann.
2962 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2963 Bugzilla 1089. Patch from Heiko Schlittermann.
2965 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2966 Bugzilla 1131. Patch from Holger Weiß.
2968 PP/09 Handle IPv6 addresses with SPF.
2969 Bugzilla 860. Patch from Wolfgang Breyha.
2971 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2973 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2976 PP/11 match_* no longer expand right-hand-side by default.
2977 New compile-time build option, EXPAND_LISTMATCH_RHS.
2978 New expansion conditions, "inlist", "inlisti".
2980 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2982 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2984 PP/14 fix log_write() format string regression from TF/03.
2985 Bugzilla 1152. Patch from Dmitry Isaikin.
2991 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2993 PP/02 Harmonised TLS library version reporting; only show if debugging.
2994 Layout now matches that introduced for other libraries in 4.74 PP/03.
2996 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2998 PP/04 New "dns_use_edns0" global option.
3000 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3003 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3004 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3006 TK/01 Updated PolarSSL code to 0.14.2.
3007 Bugzilla 1097. Patch from Andreas Metzler.
3009 PP/07 Catch divide-by-zero in ${eval:...}.
3010 Fixes bugzilla 1102.
3012 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3015 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3016 format-string attack -- SECURITY: remote arbitrary code execution.
3018 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3019 time unintentionally subject to list matching rules, letting the header
3020 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3021 arbitrary string expansion). This allowed for information disclosure.
3023 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3024 INT_MIN/-1 -- value coerced to INT_MAX.
3030 NM/01 Workaround for PCRE version dependency in version reporting
3033 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3034 This fixes portability to compilers other than gcc, notably
3035 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3037 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3038 makefiles for portability to HP-UX and POSIX correctness.
3040 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3041 Also via indented variable definition in the Makefile.
3042 (Debugging by Oliver Heesakkers).
3044 PP/02 Restore caching of spamd results with expanded spamd_address.
3045 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3047 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3048 Improves build reliability. Fix from: Frank Elsner
3050 NM/02 Fix wide character breakage in the rfc2047 coding
3051 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3053 NM/03 Allow underscore in dnslist lookups
3054 Fixes bug 1026. Patch from Graeme Fowler
3056 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3057 Code patches from Adam Ciarcinski of NetBSD.
3059 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3062 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3063 is logged, to avoid truncation. Patch from John Horne.
3065 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3066 Patch from Jakob Hirsch.
3068 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3069 SQL string expansion failure details.
3070 Patch from Andrey Oktyabrski.
3072 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3073 Patch from Simon Arlott.
3075 PP/09 New lookups functionality failed to compile on old gcc which rejects
3076 extern declarations in function scope.
3077 Patch from Oliver Fleischmann
3079 PP/10 Use sig_atomic_t for flags set from signal handlers.
3080 Check getgroups() return and improve debugging.
3081 Fixed developed for diagnosis in bug 927 (which turned out to be
3084 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3085 Patch from Mark Zealey.
3087 PP/12 Bugzilla 1056: Improved spamd server selection.
3088 Patch from Mark Zealey.
3090 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3091 Based on patch from Heiko Schlittermann.
3093 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3094 Patch from Uwe Doering, sign-off by Michael Haardt.
3096 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3097 variable declaration deep within a block. Bug and patch from
3100 PP/15 lookups-Makefile IRIX compatibility coercion.
3102 PP/16 Make DISABLE_DKIM build knob functional.
3104 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3105 Patch by Simon Arlott
3107 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3108 variable argument macros. Our copy now differs from upstream.
3114 TF/01 Failure to get a lock on a hints database can have serious
3115 consequences so log it to the panic log.
3117 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3118 controlled using the smtp_confirmation log selector.
3120 TF/03 Include the error message when we fail to unlink a spool file.
3122 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3123 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3124 for maintaining out-of-tree patches for some time.
3126 PP/01 Bugzilla 139: Documentation and portability issues.
3127 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3128 Handle per-OS dynamic-module compilation flags.
3130 PP/02 Let /dev/null have normal permissions.
3131 The 4.73 fixes were a little too stringent and complained about the
3132 permissions on /dev/null. Exempt it from some checks.
3133 Reported by Andreas M. Kirchwitz.
3135 PP/03 Report version information for many libraries, including
3136 Exim version information for dynamically loaded libraries. Created
3137 version.h, now support a version extension string for distributors
3138 who patch heavily. Dynamic module ABI change.
3140 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3141 privilege escalation vulnerability whereby the Exim run-time user
3142 can cause root to append content of the attacker's choosing to
3145 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3148 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3149 If dropping privileges for untrusted macros, we disabled normal logging
3150 on the basis that it would fail; for the Exim run-time user, this is not
3151 the case, and it resulted in successful deliveries going unlogged.
3152 Fixed. Reported by Andreas Metzler.
3158 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3159 only prepend for the Resent-* case. Fixes regression introduced in
3160 Exim 4.70 by NM/22 for Bugzilla 607.
3162 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3163 increasing numbers of administrators be bitten by this.
3165 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3167 PP/03 Bugzilla 994: added openssl_options main configuration option.
3169 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3171 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3173 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3175 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3176 without a peer certificate, leading to a segfault because of an
3177 assumption that peers always have certificates. Be a little more
3178 paranoid. Problem reported by Martin Tscholak.
3180 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3181 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3182 NB: ClamAV planning to remove STREAM in "middle of 2010".
3183 CL also introduces -bmalware, various -d+acl logging additions and
3184 more caution in buffer sizes.
3186 PP/09 Implemented reverse_ip expansion operator.
3188 PP/10 Bugzilla 937: provide a "debug" ACL control.
3190 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3192 PP/12 Bugzilla 973: Implement --version.
3194 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3196 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3198 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3200 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3201 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3202 ignore trailing whitespace.
3204 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3206 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3209 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3210 Notification from Dr Andrew Aitchison.
3212 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3213 ExtendedDetectionInfo response format.
3214 Notification from John Horne.
3216 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3219 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3220 XSL and documented dependency on system catalogs, with examples of how
3223 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3226 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3227 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3228 configuration file which is writeable by the Exim user or group.
3230 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3231 of configuration files to cover files specified with the -C option if
3232 they are going to be used with root privileges, not just the default
3235 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3236 option (effectively making it always true).
3238 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3239 files to be used while preserving root privileges.
3241 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3242 that rogue child processes cannot use them.
3244 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3245 run-time user, instead of root.
3247 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3248 Exim run-time user without dropping privileges.
3250 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3251 result string, instead of calling string_vformat() twice with the same
3254 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3255 for other users. Others should always drop root privileges if they use
3256 -C on the command line, even for a whitelisted configure file.
3258 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3260 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3266 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3267 $data_path, and $header_path variables; fixed documentation bugs and
3270 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3271 exipick to access non-standard spools, including the "frozen" queue
3274 NM/01 Bugzilla 965: Support mysql stored procedures.
3275 Patch from Alain Williams
3277 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3279 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3280 Patch from Andreas Metzler
3282 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3283 Patch from Kirill Miazine
3285 NM/05 Bugzilla 671: Added umask to procmail example.
3287 JJ/03 installed exipick 20100323.0, fixing doc bug
3289 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3290 directory. Notification and patch from Dan Rosenberg.
3292 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3294 TK/02 Improve log output when DKIM signing operation fails.
3296 MH/01 Treat the transport option dkim_domain as a colon separated
3297 list, not as a single string, and sign the message with each element,
3298 omitting multiple occurences of the same signer.
3300 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3301 Bugzilla 985, 986. Patch by Simon Arlott
3303 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3304 Patch by Simon Arlott
3306 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3307 MBX locking. Notification from Dan Rosenberg.
3313 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3315 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3317 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3319 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3321 NM/04 Bugzilla 915: Flag broken perl installation during build.
3327 TK/01 Added patch by Johannes Berg that expands the main option
3328 "spamd_address" if it starts with a dollar sign.
3330 TK/02 Write list of recipients to X-Envelope-Sender header when building
3331 the mbox-format spool file for content scanning (suggested by Jakob
3334 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3335 (http://www.dcc-servers.net/) support via dccifd. Activated by
3336 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3338 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3339 by Mark Daniel Reidel <mr@df.eu>.
3341 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3342 When building exim an external PCRE library is now needed -
3343 PCRE is a system library on the majority of modern systems.
3344 See entry on PCRE_LIBS in EDITME file.
3346 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3347 conversation. Added nologin parameter to request.
3348 Patch contributed by Kirill Miazine.
3350 TF/01 Do not log submission mode rewrites if they do not change the address.
3352 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3354 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3355 log files in place. Contributed by Roberto Lima.
3357 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3359 TF/03 Bugzilla 615: When checking the local_parts router precondition
3360 after a local_part_suffix or local_part_prefix option, Exim now
3361 does not use the address's named list lookup cache, since this
3362 contains cached lookups for the whole local part.
3364 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3365 Robert Millan. Documentation is in experimental-spec.txt.
3367 TF/04 Bugzilla 668: Fix parallel build (make -j).
3369 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3371 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3372 Patch provided by Jan Srzednicki.
3374 TF/05 Leading white space used to be stripped from $spam_report which
3375 wrecked the formatting. Now it is preserved.
3377 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3378 that they are available at delivery time.
3380 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3382 TF/08 TLS error reporting now respects the incoming_interface and
3383 incoming_port log selectors.
3385 TF/09 Produce a more useful error message if an SMTP transport's hosts
3386 setting expands to an empty string.
3388 NM/06 Bugzilla 744: EXPN did not work under TLS.
3389 Patch provided by Phil Pennock.
3391 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3392 Patch provided by Richard Godbee.
3394 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3395 acl_smtp_notquit, added index entry.
3397 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3398 Patch provided by Eugene Bujak.
3400 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3401 accept(). Patch provided by Maxim Dounin.
3403 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3404 Patch provided by Phil Pennock.
3406 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3408 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3409 Patch provided by Brad "anomie" Jorsch.
3411 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3412 Patch provided by Dean Brooks.
3414 TK/05 Add native DKIM support (does not depend on external libraries).
3416 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3417 Patch provided by Graeme Fowler.
3419 NM/16 Bugzilla 851: Documentation example syntax fix.
3421 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3423 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3426 NM/19 Bugzilla 745: TLS version reporting.
3427 Patch provided by Phil Pennock.
3429 NM/20 Bugzilla 167: bool: condition support.
3430 Patch provided by Phil Pennock.
3432 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3433 clients. Patch provided by Phil Pennock.
3435 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3436 Patch provided by Brad "anomie" Jorsch.
3438 NM/23 Bugzilla 687: Fix misparses in eximstats.
3439 Patch provided by Heiko Schlittermann.
3441 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3442 Patch provided by Heiko Schlittermann.
3444 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3445 plus update to original patch.
3447 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3449 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3450 Patch provided by David Brownlee.
3452 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3454 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3456 NM/30 Bugzilla 888: TLS documentation bugfixes.
3458 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3460 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3461 Unlike the original bugzilla I have changed all shell scripts in src tree.
3463 NM/33 Bugzilla 898: Transport filter timeout fix.
3464 Patch by Todd Rinaldo.
3466 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3467 Patch by Serge Demonchaux.
3469 NM/35 Bugzilla 39: Base64 decode bug fixes.
3470 Patch by Jakob Hirsch.
3472 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3474 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3476 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3478 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3484 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3485 ALT-N's libdkim that I have put here:
3486 http://duncanthrax.net/exim-experimental/
3488 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3489 were called 'true' and it seems that C99 defines that as a reserved
3490 keyword to be used with 'bool' variable types. That means you could
3491 not include C99-style headers which use bools without triggering
3492 build errors in sieve.c.
3494 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3495 as mailq or other aliases. Changed the --help handling significantly
3496 to do whats expected. exim_usage() emits usage/help information.
3498 SC/01 Added the -bylocaldomain option to eximstats.
3500 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3502 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3504 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3510 PH/01 Another patch from the Sieve maintainer.
3512 PH/02 When an IPv6 address is converted to a string for single-key lookup
3513 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3514 dots are used instead of colons so that keys in lsearch files need not
3515 contain colons. This was done some time before quoting was made available
3516 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3517 (notated using the quote facility) so as to distinguish them from IPv4
3518 keys. This meant that lookups for IP addresses in host lists did not work
3519 for iplsearch lookups.
3521 This has been fixed by arranging for IPv6 addresses to be expressed with
3522 colons if the lookup type is iplsearch. This is not incompatible, because
3523 previously such lookups could never work.
3525 The situation is now rather anomalous, since one *can* have colons in
3526 ordinary lsearch keys. However, making the change in all cases is
3527 incompatible and would probably break a number of configurations.
3529 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3532 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3533 conversion specification without a maximum field width, thereby enabling
3534 a rogue spamd server to cause a buffer overflow. While nobody in their
3535 right mind would setup Exim to query an untrusted spamd server, an
3536 attacker that gains access to a server running spamd could potentially
3537 exploit this vulnerability to run arbitrary code as the Exim user.
3539 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3540 $primary_hostname instead of what libspf2 thinks the hosts name is.
3542 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3543 a directory entry by the name of the lookup key. Previously, if a
3544 symlink pointed to a non-existing file or a file in a directory that
3545 Exim lacked permissions to read, a lookup for a key matching that
3546 symlink would fail. Now it is enough that a matching directory entry
3547 exists, symlink or not. (Bugzilla 503.)
3549 PH/03 The body_linecount and body_zerocount variables are now exported in the
3552 PH/04 Added the $dnslist_matched variable.
3554 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3555 This means they are set thereafter only if the connection becomes
3558 PH/06 Added the client_condition to authenticators so that some can be skipped
3559 by clients under certain conditions.
3561 PH/07 The error message for a badly-placed control=no_multiline_responses left
3562 "_responses" off the end of the name.
3564 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3566 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3567 (without spaces) instead of just copying the configuration text.
3569 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3571 PH/11 Added $max_received_linelength.
3573 PH/12 Added +ignore_defer and +include_defer to host lists.
3575 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3576 way in which PCRE > 7.0 is built.
3578 PH/14 Implemented queue_only_load_latch.
3580 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3581 MAIL command. The effect was to mangle the value on 64-bit systems.
3583 PH/16 Another patch from the Sieve maintainer.
3585 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3587 PH/18 If a system quota error occurred while trying to create the file for
3588 a maildir delivery, the message "Mailbox is full" was not appended to the
3589 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3590 only to a quota excession during the actual writing of the file.
3592 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3593 characters?) which causes problems in log lines. The DN values are now
3594 passed through string_printing() before being added to log lines.
3596 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3597 and InterBase are left for another time.)
3599 PH/21 Added message_body_newlines option.
3601 PH/22 Guard against possible overflow in moan_check_errorcopy().
3603 PH/23 POSIX allows open() to be a macro; guard against that.
3605 PH/24 If the recipient of an error message contained an @ in the local part
3606 (suitably quoted, of course), incorrect values were put in $domain and
3607 $local_part during the evaluation of errors_copy.
3613 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3614 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3617 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3618 issue a MAIL command.
3620 PH/02 In an ACL statement such as
3622 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3624 if a client was not listed at all, or was listed with a value other than
3625 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3626 the condition was not true (as it should be), so access was not denied.
3627 The bug was that the ! inversion was incorrectly passed on to the second
3628 item. This has been fixed.
3630 PH/03 Added additional dnslists conditions == and =& which are different from
3631 = and & when the dns lookup returns more than one IP address.
3633 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3634 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3636 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3637 FSYNC, which compiles an option called disable_fsync that allows for
3638 bypassing fsync(). The documentation is heavily laced with warnings.
3640 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3642 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3643 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3644 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3645 including adding "make clean"; (3) Added -fPIC when compiling the test
3646 dynamically loaded module, to get rid of a warning.
3648 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3649 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3650 = 0s. The bug is otherwise harmless.
3652 PH/07 There was a bug in the dovecot authenticator such that the value of
3653 $auth1 could be overwritten, and so not correctly preserved, after a
3654 successful authentication. This usually meant that the value preserved by
3655 the server_setid option was incorrect.
3657 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3659 PH/09 Installed PCRE release 7.0.
3661 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3662 run for batched SMTP input. It is now run at the start of every message
3663 in the batch. While fixing this I discovered that the process information
3664 (output by running exiwhat) was not always getting set for -bs and -bS
3665 input. This is fixed, and it now also says "batched" for BSMTP.
3667 PH/11 Added control=no_pipelining.
3669 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3670 patch, slightly modified), and move the expansion of helo_data till after
3671 the connection is made in the smtp transport (so it can use these
3674 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3676 PH/14 Added log_selector = +pid.
3678 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3680 PH/16 Add ${if forany and ${if forall.
3682 PH/17 Added dsn_from option to vary the From: line in DSNs.
3684 PH/18 Flush SMTP output before performing a callout, unless control =
3685 no_callout_flush is set.
3687 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3688 was true (the default) a successful delivery failed to delete the retry
3689 item, thus causing premature timeout of the address. The bug is now
3692 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3694 PH/21 Long custom messages for fakedefer and fakereject are now split up
3695 into multiline responses in the same way that messages for "deny" and
3696 other ACL rejections are.
3698 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3699 with slight modification.
3701 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3702 draft, changing the syntax and factoring some duplicate code".
3704 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3705 for deliveries of the second and subsequent messages over the same SMTP
3708 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3709 ${reduce, with only minor "tidies".
3711 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3713 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3714 expansion side effects.
3716 PH/27 When a message times out after an over-quota error from an Exim-imposed
3717 quota, the bounce message says "mailbox is full". This message was not
3718 being given when it was a system quota that was exceeded. It now should
3721 MH/03 Made $recipients available in local_scan(). local_scan() already has
3722 better access to the recipient list through recipients_list[], but
3723 $recipients can be useful in postmaster-provided expansion strings.
3725 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3726 in the case of a MAIL command with additional options following the
3727 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3728 were accidentally chopped off.
3730 PH/29 SMTP synchronization checks are implemented when a command is read -
3731 there is a check that no more input is waiting when there shouldn't be
3732 any. However, for some commands, a delay in an ACL can mean that it is
3733 some time before the response is written. In this time, more input might
3734 arrive, invalidly. So now there are extra checks after an ACL has run for
3735 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3736 pipelining has not been advertised.
3738 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3740 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3741 returned in the SMTP response when smtp_return_error_details was false.
3742 This has been fixed.
3744 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3745 instead of the C I/O that was originally supplied, because problems were
3746 reported on Solaris.
3748 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3749 Exim which did not show up earlier: it was assuming that a call to
3750 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3751 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3752 was a macro that became an assignment, so it seemed to work. This has
3753 changed to a proper function call with a void return, hence the compile
3754 error. Exim's code has been fixed.
3756 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3759 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3761 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3763 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3764 because it turns out to be too big or there is a timeout) there is no
3765 "Completed" line in the log. When some messages of this type were
3766 selected by exigrep, they were listed as "not completed". Others were
3767 picked up by some special patterns. I have improved the selection
3768 criteria to be more general.
3770 PH/38 The host_find_failed option in the manualroute router can now be set
3771 to "ignore", to completely ignore a host whose IP address cannot be
3772 found. If all hosts are ignored, the behaviour is controlled by the new
3773 host_all_ignored option.
3775 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3776 homing or because of multiple MX records with /mx) generated more than
3777 one IP address, and the following item turned out to be the local host,
3778 all the secondary addresses of the first item were incorrectly removed
3779 from the list, along with the local host and any following hosts (which
3780 is what is supposed to happen).
3782 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3783 whoever called Exim into the -H file. In the case of the daemon it was
3784 behaving confusingly. When first started, it used values for whoever
3785 started the daemon, but after a SIGHUP it used the Exim user (because it
3786 calls itself on a restart). I have changed the code so that it now always
3789 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3790 message are rejected with the same error (e.g. no authentication or bad
3791 sender address), and a DATA command is nevertheless sent (as can happen
3792 with PIPELINING or a stupid MUA), the error message that was given to the
3793 RCPT commands is included in the rejection of the DATA command. This is
3794 intended to be helpful for MUAs that show only the final error to their
3797 PH/42 Another patch from the Sieve maintainer.
3799 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3800 Eximstats - Fixed some broken HTML links and added missing column headers
3802 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3803 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3805 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3807 PH/43 Yet another patch from the Sieve maintainer.
3809 PH/44 I found a way to check for a TCP/IP connection going away before sending
3810 the response to the final '.' that terminates a message, but only in the
3811 case where the client has not sent further data following the '.'
3812 (unfortunately, this is allowed). However, in many cases there won't be
3813 any further data because there won't be any more messages to send. A call
3814 to select() can be used: if it shows that the input is "ready", there is
3815 either input waiting, or the socket has been closed. An attempt to read
3816 the next input character can distinguish the two cases. Previously, Exim
3817 would have sent an OK response which the client would never have see.
3818 This could lead to message repetition. This fix should cure that, at
3819 least in a lot of common cases.
3821 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3822 advertised in response to EHLO.
3828 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3829 fixed by 4.65/MH/01 (is this a record?) are fixed:
3831 (i) An empty string was always treated as zero by the numeric comparison
3832 operators. This behaviour has been restored.
3834 (ii) It is documented that the numeric comparison operators always treat
3835 their arguments as decimal numbers. This was broken in that numbers
3836 starting with 0 were being interpreted as octal.
3838 While fixing these problems I realized that there was another issue that
3839 hadn't been noticed. Values of message_size_limit (both the global option
3840 and the transport option) were treated as octal if they started with 0.
3841 The documentation was vague. These values are now always treated as
3842 decimal, and I will make that clear in the documentation.
3848 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3849 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3852 MH/01 Don't check that the operands of numeric comparison operators are
3853 integers when their expansion is in "skipping" mode (fixes bug
3854 introduced by 4.64-PH/07).
3856 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3857 child addresses, Exim now panics and dies. Previously, because the count
3858 is held in a short int, deliveries were likely to be lost. As such a
3859 large number of recipients for a single message is ridiculous
3860 (performance will be very, very poor), I have chosen to impose a limit
3861 rather than extend the field.
3867 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3868 leftover -K file (the existence of which was triggered by #402).
3869 While we were at it, introduced process PID as part of the -K
3870 filename. This should rule out race conditions when creating
3873 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3874 processing considerably. Previous code took too long for large mails,
3875 triggering a timeout which in turn triggers #401.
3877 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3878 in the DK code in transports.c. sendfile() is not really portable,
3879 hence the _LINUX specificness.
3881 TF/01 In the add_headers option to the mail command in an Exim filter,
3882 there was a bug that Exim would claim a syntax error in any
3883 header after the first one which had an odd number of characters
3886 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3887 callout verification, Exim cached a "reject" for the entire domain. This
3888 is correct for most verifications, but it is not correct for a recipient
3889 verification with use_sender or use_postmaster set, because in that case
3890 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3891 case of MAIL FROM:<> rejection from other early rejections (e.g.
3892 rejection of HELO). When verifying a recipient using a non-null MAIL
3893 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3894 Whatever the result of the callout, the value of the domain cache is
3895 left unchanged (for any other kind of callout, getting as far as trying
3896 RCPT means that the domain itself is ok).
3898 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3901 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3902 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3903 session, the daemon ignores ECONNECT errors and logs others; it now
3904 ignores EPIPE as well.
3906 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3907 (quoted-printable decoding).
3909 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3910 later the small subsequent patch to fix an introduced bug.
3912 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3914 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3916 PH/08 An error is now given if message_size_limit is specified negative.
3918 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3919 to be given (somewhat) arbitrary names.
3921 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3924 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3925 miscellaneous code fixes
3927 PH/10 Added the log_reject_target ACL modifier to specify where to log
3930 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3931 hostname. This is wrong, because it relates to the incoming message (and
3932 probably the interface on which it is arriving) and not to the outgoing
3933 callout (which could be using a different interface). This has been
3934 changed to use the value of the helo_data option from the smtp transport
3935 instead - this is what is used when a message is actually being sent. If
3936 there is no remote transport (possible with a router that sets up host
3937 addresses), $smtp_active_hostname is used.
3939 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3940 tweaks were necessary in order to get it to work (see also 21 below):
3941 (a) The code assumed that strncpy() returns a negative number on buffer
3942 overflow, which isn't the case. Replaced with Exim's string_format()
3944 (b) There were several signed/unsigned issues. I just did the minimum
3945 hacking in of casts. There is scope for a larger refactoring.
3946 (c) The code used strcasecmp() which is not a standard C function.
3947 Replaced with Exim's strcmpic() function.
3948 (d) The code set only $1; it now sets $auth1 as well.
3949 (e) A simple test gave the error "authentication client didn't specify
3950 service in request". It would seem that Dovecot has changed its
3951 interface. Fortunately there's a specification; I followed it and
3952 changed what the client sends and it appears to be working now.
3954 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3957 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3958 address A is aliased to B and C, where B exists and C does not. Without
3959 -v the output is "A verified" because verification stops after a
3960 successful redirection if more than one address is generated. However,
3961 with -v the child addresses are also verified. Exim was outputting "A
3962 failed to verify" and then showing the successful verification for C,
3963 with its parentage. It now outputs "B failed to verify", showing B's
3964 parentage before showing the successful verification of C.
3966 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3967 look up a TXT record in a specific list after matching in a combined
3970 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3971 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3972 they consult the DNS. I had assumed they would set it the way they
3973 wanted; and indeed my experiments on Linux seem to show that in some
3974 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3975 To be on the safe side, however, I have now made the interface to
3976 host_find_byname() similar to host_find_bydns(), with an argument
3977 containing the DNS resolver options. The host_find_byname() function now
3978 sets these options at its start, just as host_find_bydns() does. The smtp
3979 transport options dns_qualify_single and dns_search_parents are passed to
3980 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3981 of host_find_byname() use the default settings of RES_DEFNAMES
3982 (qualify_single) but not RES_DNSRCH (search_parents).
3984 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3985 spool_read_header() do less string testing, by means of a preliminary
3986 switch on the second character of optional "-foo" lines. (This is
3987 overdue, caused by the large number of possibilities that now exist.
3988 Originally there were few.) While I was there, I also converted the
3989 str(n)cmp tests so they don't re-test the leading "-" and the first
3990 character, in the hope this might squeeze out yet more improvement.
3992 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3993 flag allowing group syntax was set by the header_syntax check but not
3994 turned off, possible causing trouble later; (2) The flag was not being
3995 set at all for the header_verify test, causing "group"-style headers to
3996 be rejected. I have now set it in this case, and also caused header_
3997 verify to ignore an empty address taken from a group. While doing this, I
3998 came across some other cases where the code for allowing group syntax
3999 while scanning a header line wasn't quite right (mostly, not resetting
4000 the flag correctly in the right place). These bugs could have caused
4001 trouble for malformed header lines. I hope it is now all correct.
4003 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4004 with the "reply" argument non-NULL. The code, however (which originally
4005 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4006 but it didn't always do it. This confused somebody who was copying the
4007 code for some other use. I have removed all the tests.
4009 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4010 feature that was used to support insecure browsers during the U.S. crypto
4011 embargo. It requires special client support, and Exim is probably the
4012 only MTA that supported it -- and would never use it because real RSA is
4013 always available. This code has been removed, because it had the bad
4014 effect of slowing Exim down by computing (never used) parameters for the
4015 RSA_EXPORT functionality.
4017 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4018 authenticator to fail if there's a tab character in the incoming data
4019 (there should never be unless someone is messing about, as it's supposed
4020 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4021 if the connection is using TLS or if the remote IP is the same as the
4022 local IP, and the "valid-client-cert option" if a client certificate has
4025 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4026 authenticators. This can be used for authorization after authentication
4027 succeeds. (In the case of plaintext, it servers for both authentication
4030 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4031 if any retry times were supplied.
4033 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4034 connection, where there is no client IP address to check. In this
4035 situation, the verify now always succeeds.
4037 PH/25 Applied John Jetmore's -Mset patch.
4039 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4041 PH/27 In a string expansion for a processed (not raw) header when multiple
4042 headers of the same name were present, leading whitespace was being
4043 removed from all of them, but trailing whitespace was being removed only
4044 from the last one. Now trailing whitespace is removed from each header
4045 before concatenation. Completely empty headers in a concatenation (as
4046 before) are ignored.
4048 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4049 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4051 PH/29 [Removed. This was a change that I later backed out, and forgot to
4052 correct the ChangeLog entry (that I had efficiently created) before
4053 committing the later change.]
4055 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4056 address errors (4xx response to RCPT) over the same connection as other
4057 messages routed to the same hosts. Such deliveries are always "forced",
4058 so retry times are not inspected. This resulted in far too many retries
4059 for the affected addresses. The effect occurred only when there were more
4060 hosts than the hosts_max_try setting in the smtp transport when it had
4061 the 4xx errors. Those hosts that it had tried were not added to the list
4062 of hosts for which the message was waiting, so if all were tried, there
4063 was no problem. Two fixes have been applied:
4065 (i) If there are any address or message errors in an SMTP delivery, none
4066 of the hosts (tried or untried) are now added to the list of hosts
4067 for which the message is waiting, so the message should not be a
4068 candidate for sending over the same connection that was used for a
4069 successful delivery of some other message. This seems entirely
4070 reasonable: after all the message is NOT "waiting for some host".
4071 This is so "obvious" that I'm not sure why it wasn't done
4072 previously. Hope I haven't missed anything, but it can't do any
4073 harm, as the worst effect is to miss an optimization.
4075 (ii) If, despite (i), such a delivery is accidentally attempted, the
4076 routing retry time is respected, so at least it doesn't keep
4077 hammering the server.
4079 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4080 in ${readsocket because some servers need this prod.
4082 PH/32 Added some extra debug output when updating a wait-xxx database.
4084 PH/33 The hint "could be header name not terminated by colon", which has been
4085 given for certain expansion errors for a long time, was not being given
4086 for the ${if def:h_colon_omitted{... case.
4088 PH/34 The spec says: "With one important exception, whenever a domain list is
4089 being scanned, $domain contains the subject domain." There was at least
4090 one case where this was not true.
4092 PH/35 The error "getsockname() failed: connection reset by peer" was being
4093 written to the panic log as well as the main log, but it isn't really
4094 panic-worthy as it just means the connection died rather early on. I have
4095 removed the panic log writing for the ECONNRESET error when getsockname()
4098 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4099 runs only) independently of the message's sender address. This meant
4100 that, if the 4xx error was in fact related to the sender, a different
4101 message to the same recipient with a different sender could confuse
4102 things. In particular, this can happen when sending to a greylisting
4103 server, but other circumstances could also provoke similar problems.
4104 I have changed the default so that the retry time for these errors is now
4105 based a combination of the sender and recipient addresses. This change
4106 can be overridden by setting address_retry_include_sender=false in the
4109 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4110 remote server are returned as part of bounce messages. This was not
4111 happening for LMTP over a pipe (the lmtp transport), but now it is the
4112 same for both kinds of LMTP.
4114 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4115 in header lines that were in fact CNAMEs. This is no longer the case.
4117 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4118 and queue runs started by the daemon processed all messages. This has
4119 been fixed so that -R and -S can now usefully be given with -q<time>.
4121 PH/40 Import PCRE release 6.7 (fixes some bugs).
4123 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4125 PH/42 Give an error if -q is specified more than once.
4127 PH/43 Renamed the variables $interface_address and $interface_port as
4128 $received_ip_address and $received_port, to make it clear that these
4129 values apply to message reception, and not to the outgoing interface when
4130 a message is delivered. (The old names remain recognized, of course.)
4132 PH/44 There was no timeout on the connect() call when using a Unix domain
4133 socket in the ${readsocket expansion. There now is.
4135 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4136 be meaningful with "accept".
4138 SC/01 Eximstats V1.43
4139 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4141 SC/02 Eximstats V1.44
4142 Use a glob alias rather than an array ref in the generated
4143 parser. This improves both readability and performance.
4145 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4146 Collect SpamAssassin and rejection statistics.
4147 Don't display local sender or destination tables unless
4148 there is data to show.
4149 Added average volumes into the top table text output.
4151 SC/04 Eximstats V1.46
4152 Collect data on the number of addresses (recipients)
4153 as well as the number of messages.
4155 SC/05 Eximstats V1.47
4156 Added 'Message too big' to the list of mail rejection
4157 reasons (thanks to Marco Gaiarin).
4159 SC/06 Eximstats V1.48
4160 Mainlog lines which have GMT offsets and are too short to
4161 have a flag are now skipped.
4163 SC/07 Eximstats V1.49 (Alain Williams)
4164 Added the -emptyok flag.
4166 SC/08 Eximstats V1.50
4167 Fixes for obtaining the IP address from reject messages.
4169 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4170 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4171 whitespace changes from 4.64-PH/27
4173 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4176 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4177 are found, allow negative numbers in numeric criteria)
4179 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4181 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4182 to match changes made in 4.64-PH/43
4184 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4186 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4187 "domain" is actually an IP address, and force a failure. This locks out
4188 those revolvers/nameservers that support "A-for-A" lookups, in
4189 contravention of the specifications.
4191 PH/48 When a host name was looked up from an IP address, and the subsequent
4192 forward lookup of the name timed out, the host name was left in
4193 $sender_host_name, contrary to the specification.
4195 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4196 restricted to single-key lookups, Exim was not diagnosing an error if
4197 * or *@ was used with a query-style lookup.
4199 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4201 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4202 long ago, but noone interested enough thought of it. Let's just say that
4203 the "1.1" means that there are some new functions that weren't there at
4204 some point in the past.
4206 PH/51 Error processing for expansion failure of helo_data from an smtp
4207 transport during callout processing was broken.
4209 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4210 tested/used via the -bh/-bhc/-bs options.
4212 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4213 bug, fixed in subsequent PCRE releases).
4215 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4216 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4218 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4224 SC/01 Use a glob alias rather than an array ref in eximstats generated
4225 parser. This improves both readability and performance.
4227 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4228 Don't display local sender or destination tables in eximstats unless
4229 there is data to show.
4230 Added average volumes into the eximstats top table text output.
4232 SC/03 Collect data on the number of addresses (recipients) as well
4233 as the number of messages in eximstats.
4235 TF/01 Correct an error in the documentation for the redirect router. Exim
4236 does (usually) call initgroups() when daemonizing.
4238 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4239 with consistent privilege compared to when running as a daemon.
4241 TF/03 Note in the spec that $authenticated_id is not set for local
4242 submissions from trusted users.
4244 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4245 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4247 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4248 by adding some example configuration directives to the default
4249 configuration file. A little bit of work is required to uncomment the
4250 directives and define how usernames and passwords are checked, but
4251 there is now a framework to start from.
4253 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4254 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4255 without this. I don't know how relevant this is to other LDAP libraries.
4257 PH/02 Add the verb name to the "unknown ACL verb" error.
4259 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4261 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4263 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4264 directory not expanded when it should be if an expanded home directory
4265 was set for the address (which is overridden by the transport).
4267 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4270 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4271 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4272 because it is too late at that time, and has no effect.
4274 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4275 security issue with \' (bugzilla #107). I could not use the
4276 PQescapeStringConn() function, because it needs a PGconn value as one of
4279 PH/08 When testing addresses using -bt, indicate those final addresses that
4280 are duplicates that would not cause an additional delivery. At least one
4281 person was confused, thinking that -bt output corresponded to deliveries.
4282 (Suppressing duplicates isn't a good idea as you lose the information
4283 about possibly different redirections that led to the duplicates.)
4285 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4286 systems where poll() doesn't work, in particular OS X.
4288 PH/10 Added more information to debugging output for retry time not reached.
4290 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4291 operations in malware.c.
4293 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4296 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4297 syslog_duplication set false, log lines that would normally be written
4298 both the the main log and to the reject log were not written to syslog at
4301 PH/14 In the default configuration, change the use of "message" in ACL warn
4302 statements to "add_header".
4304 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4305 not followed by a command (e.g. "seen endif").
4307 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4308 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4311 PH/17 Added extra conditions to the default value of delay_warning_condition
4315 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4316 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4317 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4320 The Auto-Submitted: and various List- headers are standardised, whereas I
4321 don't think Precedence: ever was.
4323 PH/18 Refactored debugging code in route_finduser() to show more information,
4324 in particular, the error code if getpwnam() issues one.
4326 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4327 This is apparently needed in addition to the PH/07 change above to avoid
4328 any possible encoding problems.
4330 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4331 but not after initializing Perl.
4333 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4334 output them only if debugging. By default they are written stderr,
4335 apparently, which is not desirable.
4337 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4340 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4343 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4345 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4346 authenticated or an ident call has been made. Suppress the default
4347 values for $authenticated_id and $authenticated_sender (but permit -oMai
4348 and -oMas) when testing with -bh.
4350 PH/24 Re-jigged the order of the tests in the default configuration so that the
4351 tests for valid domains and recipients precede the DNS black list and CSA
4352 tests, on the grounds that those ones are more expensive.
4354 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4355 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4356 This bug exists in every version of Exim that I still have, right back to
4359 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4360 However, an attempt to turn on case-sensitivity in a regex key by
4361 including (?-i) didn't work because the subject string was already
4362 lowercased, and the effects were non-intuitive. It turns out that a
4363 one-line patch can be used to allow (?-i) to work as expected.
4369 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4370 other effects) broke the use of negated acl sub-conditions.
4372 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4375 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4376 "Deny" causes Exim to reject the incoming connection with a 554 error.
4377 Unfortunately, if there is a major crisis, such as a disk failure,
4378 tcp-wrappers gives "deny", whereas what one would like would be some
4379 kind of temporary error. A kludge has been added to help with this.
4380 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4381 554 error is used if errno is still zero or contains ENOENT (which occurs
4382 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4385 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4387 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4388 errors. Otherwise a message that provokes a temporary error (when other
4389 messages do not) can cause a whole host to time out.
4391 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4392 addresses were routed directly to files or pipes from a redirect router.
4393 File deliveries just didn't batch; pipe deliveries might have suffered
4396 PH/06 A failure to get a lock for a hints database would erroneously always say
4397 "Failed to get write lock", even when it was really a read lock.
4399 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4400 of 0600. This has been changed to use the value of the lockfile_mode
4401 option (which defaults to 0600).
4403 PH/08 Applied small patch from the Sieve maintainer.
4405 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4406 folder from quota calculations, a direct delivery into this folder messed
4407 up the contents of the maildirsize file. This was because the regex was
4408 used only to exclude .Trash (or whatever) when the size of the mailbox
4409 was calculated. There was no check that a delivery was happening into an
4410 excluded directory. This bug has been fixed by ignoring all quota
4411 processing for deliveries into excluded directories.
4413 PH/10 Added the maildirfolder_create_regex option to appendfile.
4419 PH/01 The code for finding all the local interface addresses on a FreeBSD
4420 system running IPv6 was broken. This may well have applied to all BSD
4421 systems, as well as to others that have similar system calls. The broken
4422 code found IPv4 interfaces correctly, but gave incorrect values for the
4423 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4424 that it would not match correctly against @[] and not recognize the IPv6
4427 PH/02 The ipliteral router was not recognizing addresses of the form user@
4428 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4430 PH/03 Added disable_ipv6.
4432 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4433 decoded form, because it is most often used to construct To: headers
4434 lines in autoreplies, and the decoded form may well be syntactically
4435 invalid. However, $reply_address has leading white space removed, and all
4436 newlines turned into spaces so that the autoreply transport does not
4439 PH/05 If group was specified without a user on a router, and no group or user
4440 was specified on a transport, the group from the router was ignored.
4442 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4443 for visible compile-time settings that can be used to change these
4444 numbers, for those that want even more. Backwards compatibility with old
4445 spool files has been maintained. However, going back to a previous Exim
4446 release will lost any variables that are in spool files.
4448 PH/07 Two small changes when running in the test harness: increase delay when
4449 passing a TCP/IP connection to a new process, in case the original
4450 process has to generate a bounce, and remove special handling of
4451 127.0.0.2 (sic), which is no longer necessary.
4453 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4454 be the same on different OS.
4456 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4459 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4460 whether --show-vars was specified or not
4462 JJ/02 exipick: Added support for new ACL variable spool format introduced
4465 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4466 syntactically invalid From: or Reply-to: line, and a filter used this to
4467 generate an autoreply, and therefore failed to obtain an address for the
4468 autoreply, Exim could try to deliver to a non-existent relative file
4469 name, causing unrelated and misleading errors. What now happens is that
4470 it logs this as a hard delivery error, but does not attempt to create a
4473 PH/11 The exinext utility has a -C option for testing purposes, but although
4474 the given file was scanned by exinext itself; it wasn't being passed on
4475 when Exim was called.
4477 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4478 an end-of-file indication when reading a command response.
4480 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4481 compiled. In many other places in Exim, IPv6 addresses are always
4482 recognized, so I have changed this. It also means that IPv4 domain
4483 literals of the form [IPV4:n.n.n.n] are now always recognized.
4485 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4486 used if the router is not running as root, for example, when verifying at
4487 ACL time, or when using -bh. The debugging output from this situation was
4488 non-existent - all you got was a failure to exec. I have made two
4491 (a) Failures to set uid/gid, the current directory, or a process leader
4492 in a subprocess such as that created by queryprogram now generate
4493 suitable debugging output when -d is set.
4495 (b) The queryprogram router detects when it is not running as root,
4496 outputs suitable debugging information if -d is set, and then runs
4497 the subprocess without attempting to change uid/gid.
4499 PH/15 Minor change to Makefile for building test_host (undocumented testing
4502 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4503 additional section of a DNS packet that returns MX or SRV records.
4504 Instead, it always explicitly searches for A/AAAA records. This avoids
4505 major problems that occur when a DNS server includes only records of one
4506 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4507 fixed another bug: if SRV records were looked up and the corresponding
4508 address records were *not* found in the additional section, the port
4509 values from the SRV records were lost.
4511 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4512 using the correct key (the original address) when searching the retry
4513 rules in order to find which one to use for generating the retry hint.
4515 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4516 adding the default one. Similarly, if it contains a Reply-To: header, the
4517 errors_reply_to option, if set, is not used.
4519 PH/19 When calculating a retry time, Exim used to measure the "time since
4520 failure" by looking at the "first failed" field in the retry record. Now
4521 it does not use this if it is later than than the arrival time of the
4522 message. Instead it uses the arrival time. This makes for better
4523 behaviour in cases where some deliveries succeed, thus re-setting the
4524 "first failed" field. An example is a quota failure for a huge message
4525 when small messages continue to be delivered. Without this change, the
4526 "time since failure" will always be short, possible causing more frequent
4527 delivery attempts for the huge message than are intended.
4528 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4530 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4531 $1, $2, $3) because the numerical variables can be reset during some
4532 expansion items (e.g. "match"), thereby losing the authentication data.
4534 PH/21 Make -bV show the size of off_t variables so that the test suite can
4535 decide whether to run tests for quotas > 2G.
4537 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4538 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4539 filecount value is greater than 2G or if a quota value is greater than 2G
4540 on a system where the size of off_t is not greater than 4, a panic error
4543 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4544 never match. The debug and -bh output now contains an explicit error
4545 message indicating a malformed IPv4 address or mask.
4547 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4548 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4549 PH/23 above applies.
4551 PH/25 Do not write to syslog when running in the test harness. The only
4552 occasion when this arises is a failure to open the main or panic logs
4553 (for which there is an explicit test).
4555 PH/26 Added the /no_tell option to "control=freeze".
4557 PH/27 If a host name lookup failed very early in a connection, for example, if
4558 the IP address matched host_lookup and the reverse lookup yielded a name
4559 that did not have a forward lookup, an error message of the form "no IP
4560 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4561 could be logged. Now it outputs the IP address instead of "NULL".
4563 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4564 allows the sender and the authenticated sender to be set when
4565 submitting a message from within Exim. Since child_open_exim() is
4566 documented for local_scan(), the new function should be too.
4568 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4569 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4570 results in an empty string is now treated as unset.
4572 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4574 PH/31 Added sender_verify_fail logging option.
4576 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4577 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4578 tidied the source and removed it altogether.
4580 PH/33 When a queue run was abandoned because the load average was too high, a
4581 log line was always written; now it is written only if the queue_run log
4582 selector is set. In addition, the log line for abandonment now contains
4583 information about the queue run such as the pid. This is always present
4584 in "start" and "stop" lines but was omitted from the "abandon" line.
4586 PH/34 Omit spaces between a header name and the colon in the error message that
4587 is given when verify = headers_syntax fails (if there are lots of them,
4588 the message gets confusing).
4590 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4591 names, as there are now some PTR records that contain slashes. This check
4592 is only to protect against broken name servers that fall over on strange
4593 characters, so the fact that it applies to all lookups doesn't matter.
4595 PH/36 Now that the new test suite is complete, we can remove some of the
4596 special code in Exim that was needed for the old test suite. For example,
4597 sorting DNS records because real resolvers return them in an arbitrary
4598 order. The new test suite's fake resolver always returns records in the
4601 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4602 bounces) except when queue_only is set, to avoid logging races between
4603 the different processes.
4605 PH/38 Panic-die if .include specifies a non-absolute path.
4607 PH/39 A tweak to the "H" retry rule from its user.
4609 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4610 a label. They prevented compilation on older perls.
4612 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4613 a warning to be raised on newish perls.
4615 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4616 on queue. Changes to match documented behaviour of showing count of
4617 messages matching specified criteria.
4619 PH/40 Changed the default ident timeout from 30s to 5s.
4621 PH/41 Added support for the use of login_cap features, on those BSD systems
4622 that have them, for controlling the resources used by pipe deliveries.
4624 PH/42 The content-scanning code uses fopen() to create files in which to put
4625 message data. Previously it was not paying any attention to the mode of
4626 the files. Exim runs with umask(0) because the rest of the code creates
4627 files with open(), and sets the required mode explicitly. Thus, these
4628 files were ending up world-writeable. This was not a big issue, because,
4629 being within the spool directory, they were not world-accessible. I have
4630 created a function called modefopen, which takes an additional mode
4631 argument. It sets umask(777), creates the file, chmods it to the required
4632 mode, then resets the umask. All the relevant calls to fopen() in the
4633 content scanning code have been changed to use this function.
4635 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4636 to 24 hours. This avoids potential overflow problems when processing G
4637 and H retry rules. I suspect nobody ever tinkers with this value.
4639 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4641 PH/45 When the plaintext authenticator is running as a client, the server's
4642 challenges are checked to ensure they are valid base64 strings. By
4643 default, the authentication attempt is cancelled if an invalid string is
4644 received. Setting client_ignore_invalid_base64 true ignores these errors.
4645 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4646 they are received. Thus, the responses can be made to depend on the
4647 challenges. If an invalid string is ignored, an empty string is placed in
4650 PH/46 Messages that are created by the autoreply transport now contains a
4651 References: header, in accordance with RFCs 2822 and 3834.
4653 PH/47 Added authenticated_sender_force to the smtp transport.
4655 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4657 PH/49 Installed latest patch from the Sieve maintainer.
4659 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4660 was also set, the appendfile transport was unnecessarily scanning a
4661 directory of message files (e.g. for maildir delivery) to find the count
4662 of files (along with the size), even though it did not need this
4663 information. It now does the scan only if it needs to find either the
4664 size of the count of files.
4666 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4668 PH/52 Two bugs concerned with error handling when the smtp transport is
4671 (i) Exim was not creating retry information for temporary errors given
4672 for individual recipients after the DATA command when the smtp transport
4673 was used in LMTP mode. This meant that they could be retried too
4674 frequently, and not timed out correctly.
4676 (ii) Exim was setting the flag that allows error details to be returned
4677 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4678 recipients that were returned after the DATA command.
4680 PH/53 This is related to PH/52, but is more general: for any failing address,
4681 when detailed error information was permitted to be returned to the
4682 sender, but the error was temporary, then after the final timeout, only
4683 "retry timeout exceeded" was returned. Now it returns the full error as
4684 well as "retry timeout exceeded".
4686 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4687 do this, and (what is worse) MTAs that accept it.
4689 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4690 will now be deprecated.
4692 PH/56 New os.c-cygwin from the Cygwin maintainer.
4694 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4695 formats (previously only available in exim formats via -bpr, -bpru,
4696 and -bpra. Now also available in native and exiqgrep formats)
4698 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4699 with very large, slow to parse queues
4701 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4703 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4705 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4706 responses to authentication challenges, though it was showing the
4707 challenges; (ii) I've removed the CR characters from the debug output for
4710 PH/58 Allow for the insertion of a newline as well as a space when a string
4711 is turned into more than one encoded-word during RFC 2047 encoding. The
4712 Sieve code now uses this.
4714 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4715 data_4xx, lost_connection, tls_required.
4717 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4718 message was being sent as an SMTP response.
4720 PH/61 Add -l and -k options to exicyclog.
4722 PH/62 When verifying, if an address was redirected to one new address, so that
4723 verification continued, and the new address failed or deferred after
4724 having set something in $address_data, the value of $address_data was not
4725 passed back to the ACL. This was different to the case when no
4726 redirection occurred. The value is now passed back in both cases.
4728 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4729 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4730 use login_cap.h, so on its own it isn't the distinguishing feature. The
4731 new name refers directly to the setclassresources() function.
4733 PH/65 Added configuration files for NetBSD3.
4735 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4737 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4738 is preferred over IPv4.
4740 PH/68 The bounce_return_message and bounce_return_body options were not being
4741 honoured for bounces generated during the reception of non-SMTP messages.
4742 In particular, this applied to messages rejected by the ACL. This bug has
4743 been fixed. However, if bounce_return_message is true and bounce_return_
4744 body is false, the headers that are returned for a non-SMTP message
4745 include only those that have been read before the error was detected.
4746 (In the case of an ACL rejection, they have all been read.)
4748 PH/69 The HTML version of the specification is now built in a directory called
4749 spec_html instead of spec.html, because the latter looks like a path with
4750 a MIME-type, and this confuses some software.
4752 PH/70 Catch two compiler warnings in sieve.c.
4754 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4755 function verify_get_ident() calls ip_connect() to connect a socket, but
4756 if the "connect()" function timed out, ip_connect() used to close the
4757 socket. However, verify_get_ident() also closes the socket later, and in
4758 between Exim writes to the log, which may get opened at this point. When
4759 the socket was closed in ip_connect(), the log could get the same file
4760 descriptor number as the socket. This naturally causes chaos. The fix is
4761 not to close the socket in ip_connect(); the socket should be closed by
4762 the function that creates it. There was only one place in the code where
4763 this was missing, in the iplookup router, which I don't think anybody now
4764 uses, but I've fixed it anyway.
4766 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4767 well as to direct DNS lookups. Otherwise the handling of names in host
4768 lists is inconsistent and therefore confusing.
4774 PH/01 Two changes to the default runtime configuration:
4776 (1) Move the checks for relay_from_hosts and authenticated clients from
4777 after to before the (commented out) DNS black list checks.
4779 (2) Add control=submission to the relay_from_hosts and authenticated
4780 clients checks, on the grounds that messages accepted by these
4781 statements are most likely to be submissions.
4783 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4785 (1) Generate an error if the third argument for the ${prvs expansion is
4788 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4791 (3) Reset the variables that are obtained from the first argument of
4792 ${prvscheck and used in the second argument before leaving the code,
4793 because their memory is reclaimed, so using them afterwards may do
4796 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4797 one (it's much easier than Tom thought :-).
4799 (5) Because of (4), we can now allow for the use of $prvscheck_result
4800 inside the third argument.
4802 PH/03 For some reason, the default setting of PATH when running a command from
4803 a pipe transport was just "/usr/bin". I have changed it to
4806 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4807 anything to be listed in the output from -bV.
4809 PH/05 When a filter generated an autoreply, the entire To: header line was
4810 quoted in the delivery log line, like this:
4812 => >A.N.Other <ano@some.domain> <original@ddress> ...
4814 This has been changed so that it extracts the operative address. There
4815 may be more than one such address. If so, they are comma-separated, like
4818 => >ano@some.domain,ona@other.domain <original@ddress> ...
4820 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4821 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4822 not being looked up in the rDNS to get a host name, Exim was showing the
4823 IP address twice in Received: lines, even though the IP addresses were
4824 identical. For example:
4826 Received: from [1.2.3.4] (helo=[1.2.3.4])
4828 However, if the real host name was known, it was omitting the HELO data
4829 if it matched the actual IP address. This has been tidied up so that it
4830 doesn't show the same IP address twice.
4832 PH/07 When both +timestamp and +memory debugging was on, the value given by
4833 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4834 called by the debug printing, thereby overwriting the timestamp buffer.
4835 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4837 PH/08 When the original message was included in an autoreply transport, it
4838 always said "this is a copy of the message, including all the headers",
4839 even if body_only or headers_only was set. It now gives an appropriate
4842 PH/09 Applied a patch from the Sieve maintainer which:
4844 o fixes some comments
4845 o adds the (disabled) notify extension core
4846 o adds some debug output for the result of if/elsif tests
4847 o points to the current vacation draft in the documentation
4848 and documents the missing references header update
4852 o fixes a bug in processing the envelope test (when testing
4853 multiple envelope elements, the last element determined the
4856 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4857 Electronic Mail") by including:
4859 Auto-submitted: auto-generated
4861 in the messages that it generates (bounce messages and others, such as
4862 warnings). In the case of bounce messages for non-SMTP messages, there was
4863 also a typo: it was using "Auto_submitted" (underscore instead of
4864 hyphen). Since every message generated by Exim is necessarily in response
4865 to another message, thes have all been changed to:
4867 Auto-Submitted: auto-replied
4869 in accordance with these statements in the RFC:
4871 The auto-replied keyword:
4873 - SHOULD be used on messages sent in direct response to another
4874 message by an automatic process,
4876 - MUST NOT be used on manually-generated messages,
4878 - MAY be used on Delivery Status Notifications (DSNs) and Message
4879 Disposition Notifications (MDNs),
4881 - MUST NOT be used on messages generated by automatic or periodic
4882 processes, except for messages which are automatic responses to
4885 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4886 to the default Received: header definition.
4888 PH/12 Added log selector acl_warn_skipped (default on).
4890 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4891 variables because (a) they are in the wrong storage pool and (b) even if
4892 they were copied, it wouldn't work properly because of the caching.
4894 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4895 checking when decoding. Apparently there are clients that generate
4896 overlong encoded strings. Why am I not surprised?
4898 PH/15 If the first argument of "${if match_address" was not empty, but did not
4899 contain an "@" character, Exim crashed. Now it writes a panic log message
4900 and treats the condition as false.
4902 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4904 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4905 extension "envelope-auth". The code is finished and in agreement with
4906 other implementations, but there is no documentation so far and in fact,
4907 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4908 not changing the active code.
4910 Print executed "if" and "elsif" statements when debugging is used. This
4911 helps a great deal to understand what a filter does.
4913 Document more things not specified clearly in RFC3028. I had all this
4914 sorted out, when out of a sudden new issues came to my mind. Oops."
4916 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4919 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4920 canonical form (as documented). However, after a host name lookup from
4921 the IP address, check_host() was doing a simple string comparison with
4922 addresses acquired from the DNS when checking that the found name did
4923 have the original IP as one of its addresses. Since any found IPv6
4924 addresses are likely to be in abbreviated form, the comparison could
4925 fail. Luckily, there already exists a function for doing the comparison
4926 by converting both addresses to binary, so now that is used instead of
4927 the text comparison.
4929 PH/20 There was another similar case to PH/19, when a complete host name was
4930 given in a host list; looking up its IP address could give an abbreviated
4931 form, whereas the current host's name might or might not be abbreviated.
4932 The same fix has been applied.
4938 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4939 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4942 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4943 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4945 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4947 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4948 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4949 Also, for "Auto-Submitted", treat anything other than "no" as
4950 non-personal, in accordance with RFC 3834. (Previously it treated
4951 anything starting "auto-" as non-personal.)
4953 TF/01 The control=submission/name=... option had a problem with syntax
4954 errors if the name included a slash character. The /name= option
4955 now slurps the rest of the string, so it can include any characters
4956 but it must come last in the list of options (after /sender_retain
4959 PH/05 Some modifications to the interface to the fake nameserver for the new
4967 TK/01 Added the "success_on_redirect" address verification option. See
4968 NewStuff for rationale and an example.
4970 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4972 PH/02 Patch to exigrep to allow it to work on syslog lines.
4974 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4975 fread() to skip over the body file's header line, because in Cygwin the
4976 header line is locked and is inaccessible.
4978 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4979 co-exist for some time) to make it clear that it is the Exim ID that is
4980 referenced, not the Message-ID: header line.
4982 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4983 string_format() function, because snprintf() does not exist on all
4986 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4987 ${stat: expansion item.
4989 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4990 protocol synchronization error", to keep the pedants happy.
4992 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4993 well as for IRIX systems, when gcc is being used. See the host.c source
4996 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4998 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5001 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5002 transport and to the smtp transport in LMTP mode.
5004 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5006 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5007 run a filter in a subprocess. This could lead to confusion in subsequent
5008 lookups in the parent process. There should also be a search_tidyup() at
5009 the end of the subprocess.
5011 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5012 only if the host matched helo_try_verify_hosts, which caused the
5013 verification to occur when the EHLO/HELO command was issued. The ACL just
5014 tested the remembered result. Now, if a previous verification attempt has
5015 not happened, "verify = helo" does it there and then.
5017 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5019 TK/03 Fix log output including CR from clamd.
5021 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5022 exist provoked a memory error which could cause a segfault.
5024 PH/15 Installed PCRE 6.2
5026 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5028 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5029 of the problem. Specifically, suggested +O2 rather than +O1 for the
5032 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5034 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5035 record, the port was not correctly logged when the outgoing_port log
5036 selector was set (it logged the transort's default port).
5038 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5039 fallback_hosts, and "hosts" in the smtp transport.
5041 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5042 host errors such as "Connection refused".
5044 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5045 authentication with radiusclient 0.4.9:
5047 - Error returned from rc_read_config was caught wrongly
5048 - Username/password not passed on to radius server due to wrong length.
5050 The presumption is that some radiusclient API changes for 4.51/PH/17
5051 were not taken care of correctly. The code is still untested by me (my
5052 Linux distribution still has 0.3.2 of radiusclient), but it was
5053 contributed by a Radius user.
5055 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5056 expanding the "port" option of the smtp transport.
5058 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5059 while reading a MIME header. Thanks to Tom Hughes for a patch.
5061 PH/24 Include config.h inside local_scan.h so that configuration settings are
5064 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5065 that in an ACL for RCPT (for example), you can examine exactly what was
5068 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5069 commands, but it was not correctly comparing the address with the actual
5070 client host address. Thus, it would show the EHLO address in Received:
5071 header lines when this was not necessary.
5073 PH/27 Added the % operator to ${eval:}.
5075 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5076 it should be ignoring failures (because with -C, for example, it has lost
5077 privilege). It wasn't ignoring creation failures other than "already
5080 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5083 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5084 because an input line was too long, either on its own, or by virtue of
5085 too many continuations, the temporary file was not being removed, and the
5086 return code was incorrect.
5088 PH/31 Missing "BOOL" in function definition in filtertest.c.
5090 PH/32 Applied Sieve patches from the maintainer.
5092 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5094 PH/33 Added "verify = not_blind".
5096 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5097 Local/Makefile (with some defaults set). These are used in built scripts
5098 such as exicyclog, but they have never been used in the exim_install
5099 script (though there are many overriding facilities there). I have
5100 arranged that the exim_install script now takes note of these two
5103 PH/35 Installed configuration files for Dragonfly.
5105 PH/36 When a locally submitted message by a trusted user did not contain a
5106 From: header, and the sender address was obtained from -f or from an SMTP
5107 MAIL command, and the trusted user did not use -F to supply a sender
5108 name, $originator_name was incorrectly used when constructing a From:
5109 header. Furthermore, $originator_name was used for submission mode
5110 messages from external hosts without From: headers in a similar way,
5111 which is clearly wrong.
5113 PH/37 Added control=suppress_local_fixups.
5115 PH/38 When log_selector = +received_sender was set, and the addition of the
5116 sender made the log line's construction buffer exactly full, or one byte
5117 less than full, an overflow happened when the terminating "\n" was
5120 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5121 when the result of a list match is failure because a DNS lookup failed.
5123 PH/40 RM_COMMAND is now used in the building process.
5125 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5126 the "build-* directories that it finds.
5128 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5129 address in a domain literal was a prefix of an interface address.
5131 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5132 when verifying a sender address, unless rewrite_headers is false.
5134 PH/44 Wrote a long comment about why errors_to addresses are verified as
5135 recipients, not senders.
5137 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5138 the ratelimit ACL was added.
5140 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5142 PH/46 Added extra information about PostgreSQL errors to the error string.
5144 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5145 avoiding the need to install special zones in a real server. This is
5146 backwards compatible; if it can't find the fake resolver, it drops back.
5147 Thus, both old and new test suites can be run.
5149 TF/02 Added util/ratelimit.pl
5151 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5152 clock is set back in time.
5154 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5155 Candler <B.Candler@pobox.com>.
5157 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5158 OK for addresses that are the result of redirections.
5160 PH/48 A number of further additions for the benefit of the new test suite,
5161 including a fake gethostbyname() that interfaces to the fake DNS resolver
5164 TF/06 The fix for widen_domains has also been applied to qualify_single and
5165 search_parents which are the other dnslookup options that can cause
5168 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5171 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5173 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5174 service" if it encountered an MX record with an empty target hostname.
5175 The message is now "an MX or SRV record indicated no SMTP service".
5177 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5178 if the DNS of the sending site is misconfigured. This is quite a
5179 common situation. This change restores the behaviour of treating a
5180 helo verification defer as a failure.
5182 PH/51 If self=fail was set on a router, the bounce message did not include the
5183 actual error message.
5189 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5191 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5192 command itself ended in error, the underlying message about the transport
5193 filter timeout was being overwritten with the pipe command error. Now the
5194 underlying error message should be appended to the second error message.
5196 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5198 PH/02 Reduce the amount of output that "make" produces by default. Full output
5199 can still be requested.
5201 PH/03 The warning log line about a condition test deferring for a "warn" verb
5202 was being output only once per connection, rather than after each
5203 occurrence (because it was using the same function as for successful
5204 "warn" verbs). This seems wrong, so I have changed it.
5206 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5207 it should not have, which might have caused a crash in the right
5208 circumstances, but probably never did.
5210 PH/04 Installed a modified version of Tony Finch's patch to make submission
5211 mode fix the return path as well as the Sender: header line, and to
5212 add a /name= option so that you can make the user's friendly name appear
5215 TF/03 Added the control = fakedefer ACL modifier.
5217 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5218 Mark Lowes for thorough testing.
5220 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5222 TK/03 Merged latest SRS patch from Miles Wilton.
5224 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5225 with the definition in sysexits.h (which is #included earlier).
5226 Fortunately, Exim does not actually use EX_OK. The code used to try to
5227 preserve the sysexits.h value, by assuming that macro definitions were
5228 scanned for macro replacements. I have been disabused of this notion,
5229 so now the code just undefines EX_OK before #including unistd.h.
5231 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5232 in the smtp transport. When a block could not be written in a single
5233 write() function, the timeout was being re-applied to each part-write.
5234 This seems wrong - if the receiver was accepting one byte at a time it
5235 would take for ever. The timeout is now adjusted when this happens. It
5236 doesn't have to be particularly precise.
5238 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5239 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5241 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5242 without a domain if the check to <postmaster@domain> fails.
5244 SC/01 Eximstats: added -xls and the ability to specify output files
5245 (patch written by Frank Heydlauf).
5247 SC/02 Eximstats: use FileHandles for outputting results.
5249 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5251 SC/04 Eximstats: fixed display of large numbers with -nvr option
5253 SC/05 Eximstats: fixed merging of reports with empty tables.
5255 SC/06 Eximstats: added the -include_original_destination flag
5257 SC/07 Eximstats: removed tabs and trailing whitespace.
5259 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5261 TK/06 MBOX spool code: Add real "From " MBOX separator line
5262 so the .eml file is really in mbox format (even though
5263 most programs do not really care). Patch from Alex Miller.
5265 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5266 The latter is generated from $received_to and is only set if the
5267 message has one envelope recipient. SA can use these headers,
5268 obviously out-of-the-box. Patch from Alex Miller.
5270 PH/08 The ${def test on a variable was returning false if the variable's
5271 value was "0", contrary to what the specification has always said!
5272 The result should be true unless the variable is empty.
5274 PH/09 The syntax error of a character other than { following "${if
5275 def:variable_name" (after optional whitespace) was not being diagnosed.
5276 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5277 accidental colon was present, for example, could give incorrect results.
5279 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5280 result is used (not including appendfile, where other changes are about
5283 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5284 This involved changing a lot of size variables from int to off_t. It
5285 should work with maildirs and everything.
5287 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5288 spamd dying while we are connected to it.
5290 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5293 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5294 function for BDB 4.3.
5296 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5298 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5299 the sources for PCRE have been re-arranged and more files are now
5302 PH/15 The code I had for printing potentially long long variables in PH/11
5303 above was not the best (it lost precision). The length of off_t variables
5304 is now inspected at build time, and an appropriate printing format (%ld
5305 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5306 to be "long long int" or "long int". This is needed for the internal
5307 formatting function string_vformat().
5309 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5310 the configuration file to be ":syslog", then the script "guesses" where
5311 the logs files are, rather than using the compiled in default. In our
5312 case the guess is not the same as the compiled default, so the script
5313 suddenly stopped working when I started to use syslog. The patch checks
5314 to see if log_file_path is "". If so, it attempts to read it from exim
5315 with no configuration file to get the compiled in version, before it
5316 falls back to the previous guessing code."
5318 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5319 implementing BATV in an Exim configuration. See NewStuff for the gory
5322 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5323 Makefile that are specific to HP-UX.
5325 PH/18 If the "use_postmaster" option was set for a recipient callout together
5326 with the "random" option, the postmaster address was used as the MAIL
5327 FROM address for the random test, but not for the subsequent recipient
5328 test. It is now used for both.
5330 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5331 patch removes a few documentation additions to RFC 3028, because the
5332 latest draft now contains them. It adds the new en;ascii-case comparator
5333 and a new error check for 8bit text in MIME parts. Comparator and
5334 require names are now matched exactly. I enabled the subaddress
5335 extension, but it is not well tested yet (read: it works for me)."
5337 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5338 rework some of the code of TK/09 above to avoid the hardwired use of
5339 "%lld" and "long long". Replaced the call to snprintf() with a call to
5342 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5343 records point to non-existent hosts", "retry timeout exceeded", and
5344 "retry time not reached for any host after a long failure period".
5346 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5347 experimental DomainKeys support:
5349 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5350 (2) On an error such as an illegally used "control", the wrong name for
5351 the control was given.
5353 These problems did NOT occur unless DomainKeys support was compiled.
5355 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5357 PH/24 Added ${if match_ip condition.
5359 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5360 databases so that it will be absolutely obvious if a crash occurs in the
5361 DB library. This is a regular occurrence (often caused by mis-matched
5364 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5365 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5366 picked up on a user's system that detects such things. There doesn't seem
5367 to be a gcc warning option for this - only an attribute that has to be
5368 put on the function's prototype. It seems that in Fedora Core 4 they have
5369 set this on a number of new functions. No doubt there will be more in due
5372 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5373 specify a transport. However, if an address that was verified by such a
5374 router was the subject of a callout, Exim crashed because it tried to
5375 read the rcpt_include_affixes from the non-existent transport. Now it
5376 just assumes that the setting of that option is false. This bug was
5377 introduced by 4.51/PH/31.
5379 PH/28 Changed -d+all to exclude +memory, because that information is very
5380 rarely of interest, but it makes the output a lot bigger. People tend to
5381 do -d+all out of habit.
5383 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5384 code in os-type was giving problems when libc.so lives in lib64, like on
5387 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5388 aren't the modern standard, and it seems that some systems' include files
5389 don't always have them. Exim was already checking for some of the newer
5390 ones like T_AAAA, and defining it itself. I've added checks for all the
5391 record types that Exim uses.
5393 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5394 not automatically generating a new one, as it is supposed to. This
5395 prevented TLS from working. If the file did exist, but contained invalid
5396 data, a new version was generated, as expected. It was only the case of a
5397 non-existent file that was broken.
5399 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5400 with a change in libdomainkeys > 0.64.
5402 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5403 from DNS. If the selector record carries the flag, it now has
5404 precedence over the domain-wide flag.
5406 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5408 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5409 the use of a port name that isn't defined in /etc/services) occurred, the
5410 message was deferred as in a normal delivery, and thus remained on the
5411 spool, instead of being failed because of the mua_wrapper setting. This
5412 is now fixed, and I tidied up some of the mua_wrapper messages at the
5415 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5416 the messages in a hash of arrays rather than using individual hashes.
5417 This is a bit cleaner and results in dramatic memory savings, albeit
5418 at a slight CPU cost.
5420 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5421 as requested by Marc Sherman.
5423 SC/10 Eximstats: added histograms for user specified patterns as requested
5426 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5428 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5429 fopen() in the content-scanning modules that did not already have it.
5435 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5436 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5438 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5440 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5442 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5443 to test Sieve filters that use "vacation".
5445 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5446 that changes the way the GnuTLS parameters are stored in the cache file.
5447 The new format can be generated externally. For backward compatibility,
5448 if the data in the cache doesn't make sense, Exim assumes it has read an
5449 old-format file, and it generates new data and writes a new file. This
5450 means that you can't go back to an older release without removing the
5453 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5454 work if there are any delivery delays because "one_time" forces the
5455 parent to be marked "delivered", so its unseen clone is never tried
5456 again. For this reason, Exim now forbids the simultaneous setting of
5459 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5460 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5461 entry if you want to know the details. The fix, however, neglected to
5462 consider the case when local delivery batching is involved. The test for
5463 "previously delivered" was not happening when checking to see if an
5464 address could be batched with a previous (undelivered) one; under
5465 certain circumstances this could lead to multiple deliveries to the same
5468 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5469 in its include files, and this causes problems building Exim.
5471 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5472 header_syntax) but Exim was just ignoring anything given after a slash.
5473 In particular, this caused confusion with an attempt to use "verify =
5474 reverse_host_lookup/defer_ok". An error is now given when options are
5475 supplied for verify items that do not have them. (Maybe reverse_host_
5476 lookup should have a defer_ok option, but that's a different point.)
5478 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5479 defined by RFC 821) to 2048, because there were problems with some AUTH
5480 commands, and RFC 1869 says the size should be increased for extended
5481 SMTP commands that take arguments.
5483 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5486 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5487 "unknown" error; now it says that the functionality isn't in the binary.
5489 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5490 an address' error message when a string expansion fails (syntax or
5491 whatever). Otherwise the password may appear in the log. Following change
5492 PH/42 below, there is no longer a chance of it appearing in a bounce
5495 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5497 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5498 of its addresses was ever tried. (Bugzilla bug #2.)
5500 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5501 the result incorrectly in the debug output. (It correctly added a newline
5502 to what was transported.)
5504 TF/01 Added $received_time.
5506 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5507 commented out examples of how to interface to a virus scanner and to
5508 SpamAssassin. Also added commented examples of av_scanner and
5509 spamd_address settings.
5511 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5512 and controls are allowed in which ACLs. There were a couple of minor
5513 errors. Some of the entries in the conditions table (which is a table of
5514 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5515 negation of where the condition IS allowed.
5517 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5519 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5520 header file does not have a version number, so I've had to invent a new
5521 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5522 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5523 radiusclient), but it was contributed by a Radius user.
5525 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5526 files or directories, for OpenSSL.
5528 PH/19 When an Exim process that is running as root has to create an Exim log
5529 file, it does so in a subprocess that runs as exim:exim so as to get the
5530 ownership right at creation (otherwise, other Exim processes might see
5531 the file with the wrong ownership). There was no test for failure of this
5532 fork() call, which would lead to the process getting stuck as it waited
5533 for a non-existent subprocess. Forks do occasionally fail when resources
5534 run out. I reviewed all the other calls to fork(); they all seem to check
5537 PH/20 When checking for unexpected SMTP input at connect time (before writing
5538 the banner), Exim was not dealing correctly with a non-positive return
5539 from the read() function. If the client had disconnected by this time,
5540 the result was a log entry for a synchronization error with an empty
5541 string after "input=" when read() returned zero. If read() returned -1
5542 (an event I could not check), uninitialized data bytes were printed.
5543 There were reports of junk text (parts of files, etc) appearing after
5546 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5548 PH/22 Added support for macro redefinition, and (re)definition in between
5549 driver and ACL definitions.
5551 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5552 forgetting to use the resulting value; it was using the unexpanded value.
5554 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5555 hadn't been configured. The fix is from Juergen Kreileder, who
5556 understands it better than I do:
5558 "Here's what I see happening with three configured cyrus_sasl
5559 authenticators configured (plain, login, cram-md5):
5561 On startup auth_cyrus_sasl_init() gets called for each of these.
5562 This means three calls to sasl_listmech() without a specified mech_list.
5563 => SASL tests which mechs of all available mechs actually work
5564 => three warnings about OTP not working
5565 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5567 With the patch, sasl_listmech() also gets called three times. But now
5568 SASL's mech_list option is set to the server_mech specified in the the
5569 authenticator. Or in other words, the answer from sasl_listmech()
5570 gets limited to just the mech you're testing for (which is different
5572 => the return list contains just 'plain' or 'login', 'cram-md5' or
5573 nothing depending on the value of ob->server_mech.
5575 I've just tested the patch: Authentication still works fine,
5576 unavailable mechs specified in the exim configuration are still
5577 caught, and the auth.log warnings about OTP are gone."
5579 PH/25 When debugging is enabled, the contents of the command line are added
5580 to the debugging output, even when log_selector=+arguments is not
5583 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5584 answer is "GNU", and only if the return is "GNU/something" is the answer
5587 PH/27 $acl_verify_message is now set immediately after the failure of a
5588 verification in an ACL, and so is available in subsequent modifiers. In
5589 particular, the message can be preserved by coding like this:
5591 warn !verify = sender
5592 set acl_m0 = $acl_verify_message
5594 Previously, $acl_verify_message was set only while expanding "message"
5595 and "log_message" when a very denied access.
5597 PH/28 Modified OS/os.c-Linux with
5599 -#ifndef OS_LOAD_AVERAGE
5600 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5602 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5603 nomenclature these days.)
5605 PH/29 Installed patch from the Sieve maintainer that adds the options
5606 sieve_useraddress and sieve_subaddress to the redirect router.
5608 PH/30 In these circumstances:
5609 . Two addresses routed to the same list of hosts;
5610 . First host does not offer TLS;
5611 . First host accepts first address;
5612 . First host gives temporary error to second address;
5613 . Second host offers TLS and a TLS session is established;
5614 . Second host accepts second address.
5615 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5616 and peerdn, if requested) that were in fact used only for the second
5619 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5620 attention to any local part prefix or suffix that was matched by the
5621 router that accepted the address. It now behaves in the same way as it
5622 does for delivery: the affixes are removed from the local part unless
5623 rcpt_include_affixes is set on the transport.
5625 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5626 timeout during the DATA phase of an incoming message.
5628 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5629 applied a patch sanctioned by the Sieve maintainer.
5631 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5632 the uid or gid is negative. A case of a negative gid caused this to be
5633 noticed. The fix allows for either to be negative.
5635 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5636 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5639 PH/36 The change PH/12 above was broken. Fixed it.
5641 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5642 the grounds that routing the same address twice would always produce the
5643 same answer. This might have been true once, but it is certainly no
5644 longer true now. Routing a child address may depend on the previous
5645 routing that produced that child. Some complicated redirection strategies
5646 went wrong when messages had multiple recipients, and made Exim's
5647 behaviour dependent on the order in which the addresses were given.
5649 I have moved the duplicate checking until after the routing is complete.
5650 Exim scans the addresses that are assigned to local and remote
5651 transports, and removes any duplicates. This means that more work will be
5652 done, as duplicates will always all be routed, but duplicates are
5653 presumably rare, so I don't expect this is of any significance.
5655 For deliveries to pipes, files, and autoreplies, the duplicate checking
5656 still happens during the routing process, since they are not going to be
5659 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5660 It corrects a timeout issue with spamd. This is Ian's comment: "The
5661 background is that sometimes spamd either never reads data from a
5662 connection it has accepted, or it never writes response data. The exiscan
5663 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5664 blindly assumes that writes won't block so it may never time out."
5666 PH/39 Allow G after quota size as well as K and M.
5668 PH/40 The value set for $authenticated_id in an authenticator may not contain
5669 binary zeroes or newlines because the value is written to log lines and
5670 to spool files. There was no check on this. Now the value is run through
5671 the string_printing() function so that such characters are converted to
5672 printable escape sequences.
5674 PH/41 $message_linecount is a new variable that contains the total number of
5675 lines in the message. Compare $body_linecount, which is the count for the
5678 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5679 bounce and delay warning messages, except in certain special cases, which
5682 (a) An SMTP error message from a remote host;
5683 (b) A message specified in a :fail: redirection;
5684 (c) A message specified in a "fail" command in a system filter;
5685 (d) A message specified in a FAIL return from the queryprogram router;
5686 (e) A message specified by the cannot_route_message router option.
5688 In these cases only, Exim does include the error details in bounce and
5689 warning messages. There are also a few cases where bland messages such
5690 as "unrouteable address" or "local delivery error" are given.
5692 PH/43 $value is now also set for the "else" part of a ${run expansion.
5694 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5695 being worked on, but at least Exim now implements the latest version to
5698 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5699 process to complete was treated as a delivery failure, a timeout while
5700 writing the message to the pipe was logged, but erroneously treated as a
5701 successful delivery. Such timeouts include transport filter timeouts. For
5702 consistency with the overall process timeout, these timeouts are now
5703 treated as errors, giving rise to delivery failures by default. However,
5704 there is now a new Boolean option for the pipe transport called
5705 timeout_defer, which, if set TRUE, converts the failures into defers for
5706 both kinds of timeout. A transport filter timeout is now identified in
5709 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5710 systems where "make" and "gmake" are different, calling "gmake" at top
5711 level broke things. I've arranged for the value of $(MAKE) to be passed
5712 from the Makefile to this script so that it can call the same version of
5716 A note about Exim versions 4.44 and 4.50
5717 ----------------------------------------
5719 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5720 changes of various kinds. As a consequence, a big documentation update was
5721 needed. This delayed the release for rather longer than seemed good, especially
5722 in the light of a couple of (minor) security issues. Therefore, the changes
5723 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5724 release. So 4.44 and 4.50 are in effect two different branches that both start
5727 I have left the 4.50 change log unchanged; it contains all the changes since
5728 4.43. The change log for 4.44 is below; many of its items are identical to
5729 those for 4.50. This seems to be the most sensible way to preserve the
5730 historical information.
5736 1. Minor wording change to the doc/README.SIEVE file.
5738 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5739 computation of the current number of files was incorrect.
5741 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5742 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5745 4. Give more explanation in the error message when the command for a transport
5746 filter fails to execute.
5748 5. There are several places where Exim runs a non-Exim command in a
5749 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5750 was being done only for the command run by the queryprogram router. It is
5751 now done for all such subprocesses. The other cases are: ${run, transport
5752 filters, and the commands run by the lmtp and pipe transports.
5754 6. Added CONFIGURE_GROUP build-time option.
5756 7. Some older OS have a limit of 256 on the maximum number of file
5757 descriptors. Exim was using setrlimit() to set 1000 as a large value
5758 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5759 systems. I've change it so that if it can't get 1000, it tries for 256.
5761 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5762 was an oversight, and furthermore, ever since the addition of extra
5763 controls (e.g. 4.43/32), the checks on when to allow different forms of
5764 "control" were broken. There should now be diagnostics for all cases when a
5765 control that does not make sense is encountered.
5767 9. Added the /retain_sender option to "control=submission".
5769 10. $recipients is now available in the predata ACL (oversight).
5771 11. Tidy the search cache before the fork to do a delivery from a message
5772 received from the command line. Otherwise the child will trigger a lookup
5773 failure and thereby defer the delivery if it tries to use (for example) a
5774 cached ldap connection that the parent has called unbind on.
5776 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5777 of $address_data from the recipient verification was clobbered by the
5778 sender verification.
5780 13. The value of address_data from a sender verification is now available in
5781 $sender_address_data in subsequent conditions in the ACL statement.
5783 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5785 15. Added a new option "connect=<time>" to callout options, to set a different
5788 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5789 was its contents. (It was OK if the option was not defined at all.)
5791 17. A "Completed" log line is now written for messages that are removed from
5792 the spool by the -Mrm option.
5794 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5795 information about exactly what failed.
5797 19. Added -dd to debug only the daemon process.
5799 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5800 handles timeouts, both on the server side and network timeouts. Renamed the
5801 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5803 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5804 It is now set to "smtps".
5806 22. $host_address is now set to the target address during the checking of
5807 ignore_target_hosts.
5809 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5810 being passed; this would have caused $sender_host_name to have been used if
5811 matching the list had actually called for a host name (not very likely,
5812 since this list is usually IP addresses). A host name is now passed as
5815 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5816 code that specifies a non-restarting handler (typically sigaction() in
5817 modern systems) in an attempt to fix a rare and obscure crash bug.
5819 25. Narrowed the window for a race in the daemon that could cause it to ignore
5820 SIGCHLD signals. This is not a major problem, because they are used only to
5821 wake it up if nothing else does.
5823 26. A malformed maildirsize file could cause Exim to calculate negative values
5824 for the mailbox size or file count. Odd effects could occur as a result.
5825 The maildirsize information is now recalculated if the size or filecount
5828 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5829 support for a long time. Removed HAVE_SYS_VFS_H.
5831 28. Installed the latest version of exipick from John Jetmore.
5833 29. In an address list, if the pattern was not a regular expression, an empty
5834 subject address (from a bounce message) matched only if the pattern was an
5835 empty string. Non-empty patterns were not even tested. This was the wrong
5836 because it is perfectly reasonable to use an empty address as part of a
5837 database query. An empty address is now tested by patterns that are
5838 lookups. However, all the other forms of pattern expect the subject to
5839 contain a local part and a domain, and therefore, for them, an empty
5840 address still always fails if the pattern is not itself empty.
5842 30. Exim went into a mad DNS loop when attempting to do a callout where the
5843 host was specified on an smtp transport, and looking it up yielded more
5844 than one IP address.
5846 31. Re-factored the code for checking spool and log partition space into a
5847 function that finds that data and another that does the check. The former
5848 is then used to implement four new variables: $spool_space, $log_space,
5849 $spool_inodes, and $log_inodes.
5851 32. The RFC2047 encoding function was originally intended for short strings
5852 such as real names; it was not keeping to the 75-character limit for
5853 encoded words that the RFC imposes. It now respects the limit, and
5854 generates multiple encoded words if necessary. To be on the safe side, I
5855 have increased the buffer size for the ${rfc2047: expansion operator from
5858 33. It is now permitted to omit both strings after an "if" condition; if the
5859 condition is true, the result is "true". As before, when the second string
5860 is omitted, a false condition yields an empty string. This makes it less
5861 cumbersome to write custom ACL and router conditions.
5863 34. Failure to deliver a bounce message always caused it to be frozen, even if
5864 there was an errors_to setting on the router. The errors_to setting is now
5867 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5868 canonical form (fully expanded) before being placed in
5869 $sender_host_address.
5871 36. The table in the code that translates DNS record types into text (T_A to
5872 "A" for instance) was missing entries for NS and CNAME. It is just possible
5873 that this could have caused confusion if both these types were looked up
5874 for the same domain, because the text type is used as part of Exim's
5875 per-process caching. But the chance of anyone hitting this buglet seems
5878 37. The dnsdb lookup has been extended in a number of ways.
5880 (1) There is a new type, "zns", which walks up the domain tree until it
5881 finds some nameserver records. It should be used with care.
5883 (2) There is a new type, "mxh", which is like "mx" except that it returns
5884 just the host names, not the priorities.
5886 (3) It is now possible to give a list of domains (or IP addresses) to be
5887 looked up. The behaviour when one of the lookups defers can be
5888 controlled by a keyword.
5890 (4) It is now possible to specify the separator character for use when
5891 multiple records are returned.
5893 38. The dnslists ACL condition has been extended: it is now possible to supply
5894 a list of IP addresses and/or domains to be looked up in a particular DNS
5897 39. Added log_selector=+queue_time_overall.
5899 40. When running the queue in the test harness, wait just a tad after forking a
5900 delivery process, to get repeatability of debugging output.
5902 41. Include certificate and key file names in error message when GnuTLS fails
5903 to set them up, because the GnuTLS error message doesn't include the name
5904 of the failing file when there is a problem reading it.
5906 42. Allow both -bf and -bF in the same test run.
5908 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5910 44. The "Exiscan patch" is now merged into the mainline Exim source.
5912 45. Sometimes the final signoff response after QUIT could fail to get
5913 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5914 < 0 before doing a fflush(). This bug looks as though it goes back to the
5915 introduction of TLS in release 3.20, but "sometimes" must have been rare
5916 because the tests only now provoked it.
5918 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5919 (this can affect the format of dates).
5921 47. exim_tidydb, when checking for the continued existence of a message for
5922 which it has found a message-specific retry record, was not finding
5923 messages that were in split spool directories. Consequently, it was
5924 deleting retry records that should have stayed in existence.
5926 48. Steve fixed some bugs in eximstats.
5928 49. The SPA authentication driver was not abandoning authentication and moving
5929 on to the next authenticator when an expansion was forced to fail,
5930 contradicting the general specification for all authenticators. Instead it
5931 was generating a temporary error. It now behaves as specified.
5933 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5934 (the order specifies the preference for clients). The order is now AES256,
5935 AES128, 3DES, ARCFOUR128.
5937 51. Small patch to Sieve code - explicitly set From: when generating an
5940 52. Exim crashed if a remote delivery caused a very long error message to be
5941 recorded - for instance if somebody sent an entire SpamAssassin report back
5942 as a large number of 550 error lines. This bug was coincidentally fixed by
5943 increasing the size of one of Exim's internal buffers (big_buffer) that
5944 happened as part of the Exiscan merge. However, to be on the safe side, I
5945 have made the code more robust (and fixed the comments that describe what
5948 53. Now that there can be additional text after "Completed" in log lines (if
5949 the queue_time_overall log selector is set), a one-byte patch to exigrep
5950 was needed to allow it to recognize "Completed" as not the last thing in
5953 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5954 patch that reportedly fixes this has been added. I am not expert enough to
5955 create a test for it. This is what the patch creator wrote:
5957 "I found a little strange behaviour of ldap code when working with
5958 Windows 2003 AD Domain, where users was placed in more than one
5959 Organization Units. When I tried to give exim partial DN, the exit code
5960 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5961 But simultaneously result of request was absolutely normal ldap result,
5962 so I produce this patch..."
5964 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5965 REFERENCE, so I have modified the code to exclude the patch when that macro
5968 55. Some experimental protocols are using DNS PTR records for new purposes. The
5969 keys for these records are domain names, not reversed IP addresses. The
5970 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5971 leaves it alone. Component reversal etc. now happens only for IP addresses.
5974 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5976 57. Double the size of the debug message buffer (to 2048) so that more of very
5977 long debug lines gets shown.
5979 58. The exicyclog utility now does better if the number of log files to keep
5980 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5982 59. Two changes related to the smtp_active_hostname option:
5984 (1) $smtp_active_hostname is now available as a variable.
5985 (2) The default for smtp_banner uses $smtp_active_hostname instead
5986 of $primary_hostname.
5988 60. The host_aton() function is supposed to be passed a string that is known
5989 to be a valid IP address. However, in the case of IPv6 addresses, it was
5990 not checking this. This is a hostage to fortune. Exim now panics and dies
5991 if the condition is not met. A case was found where this could be provoked
5992 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5993 components; fortuitously, this particular loophole had already been fixed
5994 by change 4.50/55 above.
5996 If there are any other similar loopholes, the new check in host_aton()
5997 itself should stop them being exploited. The report I received stated that
5998 data on the command line could provoke the exploit when Exim was running as
5999 exim, but did not say which command line option was involved. All I could
6000 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6001 running as the user.
6004 61. There was a buffer overflow vulnerability in the SPA authentication code
6005 (which came originally from the Samba project). I have added a test to the
6006 spa_base64_to_bits() function which I hope fixes it.
6009 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6010 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6012 63. The daemon start-up calls getloadavg() while still root for those OS that
6013 need the first call to be done as root, but it missed one case: when
6014 deliver_queue_load_max is set with deliver_drop_privilege. This is
6015 necessary for the benefit of the queue runner, because there is no re-exec
6016 when deliver_drop_privilege is set.
6018 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6019 This has been fixed.
6021 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6022 was in use, was not putting the data itself into the right store pool;
6023 consequently, it could be overwritten for a subsequent message in the same
6024 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6027 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6029 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6030 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6031 the function were treating the return as a boolean value, which happened to
6032 work because 0=false and not-0=true, but is not correct code.
6034 68. The host_aton() function was not handling scoped IPv6 addresses (those
6035 with, for example, "%eth0" on the end) correctly.
6037 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6038 negated items (that is, ~something) in unsigned ints. Some compilers
6039 apparently mutter when there is no cast.
6041 70. If an address verification called from an ACL failed, and did not produce a
6042 user-specific message (i.e. there was only a "system" message), nothing was
6043 put in $acl_verify_message. In this situation, it now puts the system
6046 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6047 session; change 4.31/43 added the unwanted input to the log line - except
6048 that it did not do this in the start of session case. It now does.
6050 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6051 This is wrong and can cause the other end to generate a synchronization
6052 error if it is another Exim or anything else that does the synchronization
6053 check. A QUIT command is no longer sent after a timeout.
6055 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6056 during host lookups.
6058 74. The defer_ok option of callout verification was not working if it was used
6059 when verifying addresses in header lines, that is, for this case:
6061 verify = header_sender/callout=defer_ok
6063 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6064 those file descriptors could be used for SMTP connections. If anything
6065 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6066 could be sent to the SMTP client, causing chaos. The daemon now opens
6067 stdin, stdout, and stderr to /dev/null when it puts itself into the
6070 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6071 log by default. The user can override this with suitable Perl magic.
6073 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6074 the log message when discard triggers, was not working for the DATA ACL or
6075 for the non-SMTP ACL.
6077 78. Error message wording change in sieve.c.
6079 79. If smtp_accept_max_per_host was set, the number of connections could be
6080 restricted to fewer than expected, because the daemon was trying to set up
6081 a new connection before checking whether the processes handling previous
6082 connections had finished. The check for completed processes is now done
6083 earlier. On busy systems, this bug wouldn't be noticed because something
6084 else would have woken the daemon, and it would have reaped the completed
6087 80. If a message was submitted locally by a user whose login name contained one
6088 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6089 It caused a spool format error. I have fixed the spool reading code. A
6090 related problem was that the "from" clause in the Received: line became
6091 illegal because of the space(s). It is now covered by ${quote_local_part.
6093 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6096 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6098 83. Patch from Sieve maintainer to fix unterminated string problem in
6099 "vacation" handling.
6101 84. Some minor changes to the Linux configuration files to help with other
6102 OS variants using glibc.
6104 85. One more patch for Sieve to update vacation handling to latest spec.
6107 ----------------------------------------------------
6108 See the note above about the 4.44 and 4.50 releases.
6109 ----------------------------------------------------
6115 1. Change 4.43/35 introduced a bug that caused file counts to be
6116 incorrectly computed when quota_filecount was set in an appendfile
6119 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6120 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6123 3. Give more explanation in the error message when the command for a transport
6124 filter fails to execute.
6126 4. There are several places where Exim runs a non-Exim command in a
6127 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6128 was being done only for the command run by the queryprogram router. It is
6129 now done for all such subprocesses. The other cases are: ${run, transport
6130 filters, and the commands run by the lmtp and pipe transports.
6132 5. Some older OS have a limit of 256 on the maximum number of file
6133 descriptors. Exim was using setrlimit() to set 1000 as a large value
6134 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6135 systems. I've change it so that if it can't get 1000, it tries for 256.
6137 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6138 was an oversight, and furthermore, ever since the addition of extra
6139 controls (e.g. 4.43/32), the checks on when to allow different forms of
6140 "control" were broken. There should now be diagnostics for all cases when a
6141 control that does not make sense is encountered.
6143 7. $recipients is now available in the predata ACL (oversight).
6145 8. Tidy the search cache before the fork to do a delivery from a message
6146 received from the command line. Otherwise the child will trigger a lookup
6147 failure and thereby defer the delivery if it tries to use (for example) a
6148 cached ldap connection that the parent has called unbind on.
6150 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6151 of $address_data from the recipient verification was clobbered by the
6152 sender verification.
6154 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6155 was its contents. (It was OK if the option was not defined at all.)
6157 11. A "Completed" log line is now written for messages that are removed from
6158 the spool by the -Mrm option.
6160 12. $host_address is now set to the target address during the checking of
6161 ignore_target_hosts.
6163 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6164 being passed; this would have caused $sender_host_name to have been used if
6165 matching the list had actually called for a host name (not very likely,
6166 since this list is usually IP addresses). A host name is now passed as
6169 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6170 code that specifies a non-restarting handler (typically sigaction() in
6171 modern systems) in an attempt to fix a rare and obscure crash bug.
6173 15. Narrowed the window for a race in the daemon that could cause it to ignore
6174 SIGCHLD signals. This is not a major problem, because they are used only to
6175 wake it up if nothing else does.
6177 16. A malformed maildirsize file could cause Exim to calculate negative values
6178 for the mailbox size or file count. Odd effects could occur as a result.
6179 The maildirsize information is now recalculated if the size or filecount
6182 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6183 support for a long time. Removed HAVE_SYS_VFS_H.
6185 18. Updated exipick to current release from John Jetmore.
6187 19. Allow an empty sender to be matched against a lookup in an address list.
6188 Previously the only cases considered were a regular expression, or an
6191 20. Exim went into a mad DNS lookup loop when doing a callout where the
6192 host was specified on the transport, if the DNS lookup yielded more than
6195 21. The RFC2047 encoding function was originally intended for short strings
6196 such as real names; it was not keeping to the 75-character limit for
6197 encoded words that the RFC imposes. It now respects the limit, and
6198 generates multiple encoded words if necessary. To be on the safe side, I
6199 have increased the buffer size for the ${rfc2047: expansion operator from
6202 22. Failure to deliver a bounce message always caused it to be frozen, even if
6203 there was an errors_to setting on the router. The errors_to setting is now
6206 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6207 canonical form (fully expanded) before being placed in
6208 $sender_host_address.
6210 24. Updated eximstats to version 1.33
6212 25. Include certificate and key file names in error message when GnuTLS fails
6213 to set them up, because the GnuTLS error message doesn't include the name
6214 of the failing file when there is a problem reading it.
6216 26. Expand error message when OpenSSL has problems setting up cert/key files.
6219 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6220 (this can affect the format of dates).
6222 28. exim_tidydb, when checking for the continued existence of a message for
6223 which it has found a message-specific retry record, was not finding
6224 messages that were in split spool directories. Consequently, it was
6225 deleting retry records that should have stayed in existence.
6227 29. eximstats updated to version 1.35
6228 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6229 1.35 - bugfix such that pie charts by volume are generated correctly
6231 30. The SPA authentication driver was not abandoning authentication and moving
6232 on to the next authenticator when an expansion was forced to fail,
6233 contradicting the general specification for all authenticators. Instead it
6234 was generating a temporary error. It now behaves as specified.
6236 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6237 (the order specifies the preference for clients). The order is now AES256,
6238 AES128, 3DES, ARCFOUR128.
6240 31. Small patch to Sieve code - explicitly set From: when generating an
6243 32. Exim crashed if a remote delivery caused a very long error message to be
6244 recorded - for instance if somebody sent an entire SpamAssassin report back
6245 as a large number of 550 error lines. This bug was coincidentally fixed by
6246 increasing the size of one of Exim's internal buffers (big_buffer) that
6247 happened as part of the Exiscan merge. However, to be on the safe side, I
6248 have made the code more robust (and fixed the comments that describe what
6251 33. Some experimental protocols are using DNS PTR records for new purposes. The
6252 keys for these records are domain names, not reversed IP addresses. The
6253 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6254 leaves it alone. Component reversal etc. now happens only for IP addresses.
6257 34. The host_aton() function is supposed to be passed a string that is known
6258 to be a valid IP address. However, in the case of IPv6 addresses, it was
6259 not checking this. This is a hostage to fortune. Exim now panics and dies
6260 if the condition is not met. A case was found where this could be provoked
6261 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6262 components; fortuitously, this particular loophole had already been fixed
6263 by change 4.50/55 or 4.44/33 above.
6265 If there are any other similar loopholes, the new check in host_aton()
6266 itself should stop them being exploited. The report I received stated that
6267 data on the command line could provoke the exploit when Exim was running as
6268 exim, but did not say which command line option was involved. All I could
6269 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6270 running as the user.
6273 35. There was a buffer overflow vulnerability in the SPA authentication code
6274 (which came originally from the Samba project). I have added a test to the
6275 spa_base64_to_bits() function which I hope fixes it.
6278 36. The daemon start-up calls getloadavg() while still root for those OS that
6279 need the first call to be done as root, but it missed one case: when
6280 deliver_queue_load_max is set with deliver_drop_privilege. This is
6281 necessary for the benefit of the queue runner, because there is no re-exec
6282 when deliver_drop_privilege is set.
6284 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6285 was in use, was not putting the data itself into the right store pool;
6286 consequently, it could be overwritten for a subsequent message in the same
6287 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6290 38. Sometimes the final signoff response after QUIT could fail to get
6291 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6292 < 0 before doing a fflush(). This bug looks as though it goes back to the
6293 introduction of TLS in release 3.20, but "sometimes" must have been rare
6294 because the tests only now provoked it.
6300 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6301 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6302 Now it returns an integer. A number of calls were still expecting a T/F
6303 return. Fortuitously, in all cases, the tests worked in OK situations,
6304 which is the norm. However, things would have gone wrong on any write
6305 failures on the smtp file descriptor. This function is used when sending
6306 messages over SMTP and also when doing verify callouts.
6308 2. When Exim is called to do synchronous delivery of a locally submitted
6309 message (the -odf or -odi options), it no longer closes stderr before doing
6312 3. Implemented the mua_wrapper option.
6314 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6316 5. Implemented the functions header_remove(), header_testname(),
6317 header_add_at_position(), and receive_remove_recipient(), and exported them
6320 6. If an ACL "warn" statement specified the addition of headers, Exim already
6321 inserted X-ACL-Warn: at the start if there was no header name. However, it
6322 was not making this test for the second and subsequent header lines if
6323 there were newlines in the string. This meant that an invalid header could
6324 be inserted if Exim was badly configured.
6326 7. Allow an ACL "warn" statement to add header lines at the start or after all
6327 the Received: headers, as well as at the end.
6329 8. Added the rcpt_4xx retry error code.
6331 9. Added postmaster_mailfrom=xxx to callout verification option.
6333 10. Added mailfrom=xxxx to the callout verification option, for verify=
6336 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6337 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6339 12. Inserted some casts to stop certain compilers warning when using pointer
6340 differences as field lengths or precisions in printf-type calls (mostly
6341 affecting debugging statements).
6343 13. Added optional readline() support for -be (dynamically loaded).
6345 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6346 same clock tick as a message's arrival, so that its received time was the
6347 same as the "first fail" time on the retry record, and that message
6348 remained on the queue past the ultimate address timeout, every queue runner
6349 would try a delivery (because it was past the ultimate address timeout) but
6350 after another failure, the ultimate address timeout, which should have then
6351 bounced the address, did not kick in. This was a "< instead of <=" error;
6352 in most cases the first failure would have been in the next clock tick
6353 after the received time, and all would be well.
6355 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6356 being recognized when the domain list was tested by the match_domain
6357 condition in an expansion string.
6359 16. Added the ${str2b64: operator.
6361 17. Exim was always calling setrlimit() to set a large limit for the number of
6362 processes, without checking whether the existing limit was already
6363 adequate. (It did check for the limit on file descriptors.) Furthermore,
6364 errors from getrlimit() and setrlimit() were being ignored. Now they are
6365 logged to the main and panic logs, but Exim does carry on, to try to do its
6366 job under whatever limits there are.
6368 18. Imported PCRE 5.0.
6370 19. Trivial typo in log message " temporarily refused connection" (the leading
6373 20. If the log selector return_path_on_delivery was set and an address was
6374 redirected to /dev/null, the delivery process crashed because it assumed
6375 that a return path would always be set for a "successful" delivery. In this
6376 case, the whole delivery is bypassed as an optimization, and therefore no
6379 21. Internal re-arrangement: the function for sending a challenge and reading
6380 a response while authentication was assuming a zero-terminated challenge
6381 string. It's now changed to take a pointer and a length, to allow for
6382 binary data in such strings.
6384 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6386 23. Exim was not respecting finduser_retries when seeking the login of the
6387 uid under which it was called; it was always trying 10 times. (The default
6388 setting of finduser_retries is zero.) Also, it was sleeping after the final
6389 failure, which is pointless.
6391 24. Implemented tls_on_connect_ports.
6393 25. Implemented acl_smtp_predata.
6395 26. If the domain in control=submission is set empty, Exim assumes that the
6396 authenticated id is a complete email address when it generates From: or
6397 Sender: header lines.
6399 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6400 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6401 chown and chgrp in /bin and hostname in /usr/bin.
6403 28. Exim was keeping the "process log" file open after each use, just as it
6404 does for the main log. This opens the possibility of it remaining open for
6405 long periods when the USR1 signal hits a daemon. Occasional processlog
6406 errors were reported, that could have been caused by this. Anyway, it seems
6407 much more sensible not to leave this file open at all, so that is what now
6410 29. The long-running daemon process does not normally write to the log once it
6411 has entered its main loop, and it closes the log before doing so. This is
6412 so that log files can straightforwardly be renamed and moved. However,
6413 there are a couple of unusual error situations where the daemon does write
6414 log entries, and I had neglected to close the log afterwards.
6416 30. The text of an SMTP error response that was received during a remote
6417 delivery was being truncated at 512 bytes. This is too short for some of
6418 the long messages that one sometimes sees. I've increased the limit to
6421 31. It is now possible to make retry rules that apply only when a message has a
6422 specific sender, in particular, an empty sender.
6424 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6425 it possible to be selective about when SMTP synchronization is enforced.
6427 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6429 32. Implemented hosts_connection_nolog.
6431 33. Added an ACL for QUIT.
6433 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6436 35. Added mailbox_size and mailbox_filecount to appendfile.
6438 36. Added control = no_multiline_responses to ACLs.
6440 37. There was a bug in the logic of the code that waits for the clock to tick
6441 in the case where the clock went backwards by a substantial amount such
6442 that the microsecond fraction of "now" was more than the microsecond
6443 fraction of "then" (but the whole seconds number was less).
6445 38. Added support for the libradius Radius client library this is found on
6446 FreeBSD (previously only the radiusclient library was supported).
6452 1. When certain lookups returned multiple values in the form name=value, the
6453 quoting of the values was not always being done properly. Specifically:
6454 (a) If the value started with a double quote, but contained no whitespace,
6456 (b) If the value contained whitespace other than a space character (i.e.
6457 tabs or newlines or carriage returns) it was not quoted.
6458 This fix has been applied to the mysql and pgsql lookups by writing a
6459 separate quoting function and calling it from the lookup code. The fix
6460 should probably also be applied to nisplus, ibase and oracle lookups, but
6461 since I cannot test any of those, I have not disturbed their existing code.
6463 2. A hit in the callout cache for a specific address caused a log line with no
6464 reason for rejecting RCPT. Now it says "Previous (cached) callout
6465 verification failure".
6467 3. There was an off-by-one bug in the queryprogram router. An over-long
6468 return line was truncated at 256 instead of 255 characters, thereby
6469 overflowing its buffer with the terminating zero. As well as fixing this, I
6470 have increased the buffer size to 1024 (and made a note to document this).
6472 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6473 when Exim is waiting for an SMTP response from a remote server, Exim
6474 restarts its select() call on the socket, thereby resetting its timeout.
6475 This is not a problem when such interrupts are rare. Somebody set up a cron
6476 job to run exiwhat every 2 minutes, which is less than the normal select()
6477 timeout (5 or 10 minutes). This meant that the select() timeout never
6478 kicked in because it was always reset. I have fixed this by comparing the
6479 time when an interrupt arrives with the time at the start of the first call
6480 to select(). If more time than the timeout has elapsed, the interrupt is
6481 treated as a timeout.
6483 5. Some internal re-factoring in preparation for the addition of Sieve
6484 extensions (by MH). In particular, the "personal" test is moved to a
6485 separate function, and given an option for scanning Cc: and Bcc: (which is
6486 not set for Exim filters).
6488 6. When Exim created an email address using the login of the caller as the
6489 local part (e.g. when creating a From: or Sender: header line), it was not
6490 quoting the local part when it contained special characters such as @.
6492 7. Installed new OpenBSD configuration files.
6494 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6495 try to make them clearer.
6497 9. Callout options, other than the timeout value, were being ignored when
6498 verifying sender addresses in header lines. For example, when using
6500 verify = header_sender/callout=no_cache
6502 the cache was (incorrectly) being used.
6504 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6505 only the Cygwin environment.
6507 11. When return_path_on_delivery was set as a log selector, if different remote
6508 addresses in the same message used different return paths and parallel
6509 remote delivery occurred, the wrong values would sometimes be logged.
6510 (Whenever a remote delivery process finished, the return path value from
6511 the most recently started remote delivery process was logged.)
6513 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6514 lines when AUTH and/or TLS are in use. This is the "received protocol"
6515 field. Exim used to use "asmtp" for authenticated SMTP, without any
6516 indication (in the protocol name) for TLS use. Now it follows the RFC and
6517 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6518 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6519 names appear in log lines as well as in Received: header lines.
6521 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6522 extensions, and comparison tests, and to fix some bugs.
6524 14. Changes to the "personal" filter test:
6526 (1) The test was buggy in that it was just doing the equivalent of
6527 "contains" tests on header lines. For example, if a user's address was
6528 anne@some.where, the "personal" test would incorrectly be true for
6530 To: susanne@some.where
6532 This test is now done by extracting each address from the header in turn,
6533 and checking the entire address. Other tests that are part of "personal"
6534 are now done using regular expressions (for example, to check local parts
6535 of addresses in From: header lines).
6537 (2) The list of non-personal local parts in From: addresses has been
6538 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6539 taken from the Sieve specification recommendations.
6541 (3) If the message contains any header line starting with "List-" it is
6542 treated as non-personal.
6544 (4) The test for "circular" in the Subject: header line has been removed
6545 because it now seems ill-conceived.
6547 15. Minor typos in src/EDITME comments corrected.
6549 16. Installed latest exipick from John Jetmore.
6551 17. If headers_add on a router specified a text string that was too long for
6552 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6553 of string_sprintf() is now avoided.
6555 18. $message_body_size was not set (it was always zero) when running the DATA
6556 ACL and the local_scan() function.
6558 19. For the "mail" command in an Exim filter, no default was being set for
6559 the once_repeat time, causing a random time value to be used if "once" was
6560 specified. (If the value happened to be <= 0, no repeat happened.) The
6561 default is now 0s, meaning "never repeat". The "vacation" command was OK
6562 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6563 (I found it when inspecting the code).
6565 20. There is now an overall timeout for performing a callout verification. It
6566 defaults to 4 times the callout timeout, which applies to individual SMTP
6567 commands during the callout. The overall timeout applies when there is more
6568 than one host that can be tried. The timeout is checked before trying the
6569 next host. This prevents very long delays if there are a large number of
6570 hosts and all are timing out (e.g. when the network connections are timing
6571 out). The value of the overall timeout can be changed by specifying an
6572 additional sub-option for "callout", called "maxwait". For example:
6574 verify = sender/callout=5s,maxwait=20s
6576 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6577 to the end before writing, but this should make it even safer).
6579 22. Exim was forgetting that it had advertised PIPELINING for the second and
6580 subsequent messages on an SMTP connection. It was also not resetting its
6581 memory on STARTTLS and an internal HELO.
6583 23. When Exim logs an SMTP synchronization error within a session, it now
6584 records whether PIPELINING has been advertised or not.
6586 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6587 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6588 rather than long int.
6590 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6592 26. Added the never_mail option to autoreply.
6598 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6599 crash if the getsockname() call failed; this can happen if a connection is
6600 closed very soon after it is established. The problem was simply in the
6601 order in which certain operations were done, causing Exim to try to write
6602 to the SMTP stream before it had set up the file descriptor. The bug has
6603 been fixed by making things happen in the correct order.
6609 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6610 before the connection was closed, thus losing the rejection response.
6612 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6613 some early Solaris releases, but causes trouble in current releases where
6614 socklen_t is defined.
6616 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6619 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6622 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6623 that releasing the top of it at the end releases what was used for sub-
6624 expansions (unless the block got too big). However, discard this block if
6625 the first thing is a variable or header, so that we can use its block when
6626 it is dynamic (useful for very large $message_headers, for example).
6628 6. Lookups now cache *every* query, not just the most recent. A new, separate
6629 store pool is used for this. It can be recovered when all lookup caches are
6630 flushed. Lookups now release memory at the end of their result strings.
6631 This has involved some general refactoring of the lookup sources.
6633 7. Some code has been added to the store_xxx() functions to reduce the amount
6634 of flapping under certain conditions.
6636 8. log_incoming_interface used to affect only the <= reception log lines. Now
6637 it causes the local interface and port to be added to several more SMTP log
6638 lines, for example "SMTP connection from", and rejection lines.
6640 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6642 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6644 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6645 could be overwritten at the end of the current message (or the start of a
6646 new message if it was set in a HELO ACL). The value is now preserved for
6647 the duration of the SMTP connection.
6649 12. If a transport had a headers_rewrite setting, and a matching header line
6650 contained an unqualified address, that address was qualified, even if it
6651 did not match any rewriting rules. The underlying bug was that the values
6652 of the flags that permit the existence of unqualified sender and recipient
6653 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6654 non-local messages, and by -bnq for local messages) were not being
6655 preserved with the message after it was received.
6657 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6658 "next input=" as part of the text comprising the host identity instead of
6659 the correct text. The code was using the same buffer for two different
6660 strings. However, depending on which order the printing function evaluated
6661 its arguments, the bug did not always show up. Under Linux, for example, my
6662 test suite worked just fine.
6664 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6665 doesn't work with some older versions of Perl. It has been changed to "my",
6666 which in any case is probably the better facility to use.
6668 15. A really picky compiler found some instances of statements for creating
6669 error messages that either had too many or two few arguments for the format
6672 16. The size of the buffer for calls to the DNS resolver has been increased
6673 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6674 for addresses that have a lot of PTR records. This alleviates a problem; it
6675 does not fully solve it.
6677 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6678 buffer now truncates the list and logs the incident, which is the same
6679 action as happens when Exim is looking up a host name and its aliases.
6680 Previously in this situation something unpredictable would happen;
6681 sometimes it was "internal error: store_reset failed".
6683 18. If a server dropped the connection unexpectedly when an Exim client was
6684 using GnuTLS and trying to read a response, the client delivery process
6685 crashed while trying to generate an error log message.
6687 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6688 string, for example:
6690 warn message = H1: something\nH2: something
6692 the text was added as a single header line from Exim's point of view
6693 though it ended up OK in the delivered message. However, searching for the
6694 second and subsequent header lines using $h_h2: did not work. This has been
6695 fixed. Similarly, if a system filter added multiple headers in this way,
6696 the routers could not see them.
6698 20. Expanded the error message when iplsearch is called with an invalid key to
6699 suggest using net-iplsearch in a host list.
6701 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6702 ACLs are no longer actually imposed (and a message to that effect is
6705 22. If a "gecos" field in a passwd entry contained escaped characters, in
6706 particular, if it contained a \" sequence, Exim got it wrong when building
6707 a From: or a Sender: header from that name. A second bug also caused
6708 incorrect handling when an unquoted " was present following a character
6709 that needed quoting.
6711 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6712 was not being matched caselessly.
6714 24. Arranged for all hyphens in the exim.8 source to be escaped with
6717 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6718 properly. Recipient callout cache records were still being keyed to include
6719 the sender, even when use_sender was set false. This led to far more
6720 callouts that were necessary. The sender is no longer included in the key
6721 when use_sender is false.
6723 26. Added "control = submission" modifier to ACLs.
6725 27. Added the ${base62d: operator to decode base 62 numbers.
6727 28. dnsdb lookups can now access SRV records.
6729 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6730 the configuration file.
6732 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6733 (-v) mode. This makes the output for a verbose queue run more intelligible.
6735 31. Added a use_postmaster feature to recipient callouts.
6737 32. Added the $body_zerocount variable, containing the number of binary zero
6738 bytes in the message body.
6740 33. The time of last modification of the "new" subdirectory is now used as the
6741 "mailbox time last read" when there is a quota error for a maildir
6744 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6746 35. Added +ignore_unknown as a special item in host lists.
6748 36. Code for decoding IPv6 addresses in host lists is now included, even if
6749 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6750 address was recognized as an IP address, but was then not correctly decoded
6751 into binary, causing unexpected and incorrect effects when compared with
6758 1. Very minor rewording of debugging text in manualroute to say "list of
6759 hosts" instead of "hostlist".
6761 2. If verify=header_syntax was set, and a header line with an unqualified
6762 address (no domain) and a large number of spaces between the end of the
6763 name and the colon was received, the reception process suffered a buffer
6764 overflow, and (when I tested it) crashed. This was caused by some obsolete
6765 code that should have been removed. The fix is to remove it!
6767 3. When running in the test harness, delay a bit after writing a bounce
6768 message to get a bit more predictability in the log output.
6770 4. Added a call to search_tidyup() just before forking a reception process. In
6771 theory, someone could use a lookup in the expansion of smtp_accept_max_
6772 per_host which, without the tidyup, could leave open a database connection.
6774 5. Added the variables $recipient_data and $sender_data which get set from a
6775 lookup success in an ACL "recipients" or "senders" condition, or a router
6776 "senders" option, similar to $domain_data and $local_part_data.
6778 6. Moved the writing of debug_print from before to after the "senders" test
6781 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6782 problems for message scanning, either using a data ACL, or using
6783 local_scan() because the Received: header was not generated till after they
6784 were called (in order to set the time as the time of reception completion).
6785 I have revised the way this works. The header is now generated after the
6786 body is received, but before the ACL or local_scan() are called. After they
6787 are run, the timestamp in the header is updated.
6793 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6794 before starting a queue runner without re-exec. This happened only when
6795 deliver_drop_privilege was set or when the Exim user was set to root. The
6796 effect of the bug was that timeouts during subsequent deliveries caused
6797 crashes instead of being properly handled. The handler is now left at its
6798 default (and expected) setting.
6800 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6801 message, again when deliver_drop_privilege is set or Exim is run as root.
6802 The bug described in (1) was not present in this case, but the tidying up
6803 of the other signals was missing. I have made the two cases consistent.
6805 3. The ignore_target_hosts setting on a manualroute router was being ignored
6806 for hosts that were looked up using the /MX notation.
6808 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6811 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6812 operated on the sender address. After changing the $sender_address to <>
6813 for the sender address verify, Exim was re-instated it as the original
6814 (before rewriting) address, but remembering that it had rewritten it, so it
6815 wasn't rewriting it again. This bug also had the effect of breaking the
6816 sender address verification caching when the sender address was rewritten.
6818 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6819 This has been changed so that if the ip literal address matches
6820 ignore_target_hosts, the router declines.
6822 7. Added expansion conditions match_domain, match_address, and match_local_
6823 part (NOT match_host).
6825 8. The placeholder for the Received: header didn't have a length field set.
6827 9. Added code to Exim itself and to exim_lock to test for a specific race
6828 condition that could lead to file corruption when using MBX delivery. The
6829 issue is with the lockfile that is created in /tmp. If this file is removed
6830 after a process has opened it but before that process has acquired a lock,
6831 there is the potential for a second process to recreate the file and also
6832 acquire a lock. This could lead to two Exim processes writing to the file
6833 at the same time. The added code performs the same test as UW imapd; it
6834 checks after acquiring the lock that its file descriptor still refers to
6835 the same named file.
6837 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6838 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6841 11. Added the smtp_active_hostname option. If used, this will typically be made
6842 to depend on the incoming interface address. Because $interface_address is
6843 not set up until the daemon has forked a reception process, error responses
6844 that can happen earlier (such as "too many connections") no longer contain
6847 12. If an expansion in a condition on a "warn" statement fails because a lookup
6848 defers, the "warn" statement is abandoned, and the next ACL statement is
6849 processed. Previously this caused the whole ACL to be aborted.
6851 13. Added the iplsearch lookup type.
6853 14. Added ident_timeout as a log selector.
6855 15. Added tls_certificate_verified as a log selector.
6857 16. Added a global option tls_require_ciphers (compare the smtp transport
6858 option of the same name). This controls incoming TLS connections.
6860 17. I finally figured out how to make tls_require_ciphers do a similar thing
6861 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6862 before starting the TLS session.
6864 18. Tabs are now shown as \t in -bP output.
6866 19. If the log selector return_path_on_delivery was set, Exim crashed when
6867 bouncing a message because it had too many Received: header lines.
6869 20. If two routers both had headers_remove settings, and the first one included
6870 a superfluous trailing colon, the final name in the first list and the
6871 first name in the second list were incorrectly joined into one item (with a
6872 colon in the middle).
6878 1. Added -C and -D options to the exinext utility, mainly to make it easier
6879 to include in the automated testing, but these could be helpful when
6880 multiple configurations are in use.
6882 2. The exinext utility was not formatting the output nicely when there was
6883 an alternate port involved in the retry record key, nor when there was a
6884 message id as well (for retries that were specific to a specific message
6885 and a specific host). It was also confused by IPv6 addresses, because of
6886 the additional colons they contain. I have fixed the IPv4 problem, and
6887 patched it up to do a reasonable job for IPv6.
6889 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6890 delivery, the log line now contains "pipelined" if PIPELINING was used.
6892 4. An SMTP transport process used to panic and die if the bind() call to set
6893 an explicit outgoing interface failed. This has been changed; it is now
6894 treated in the same way as a connect() failure.
6896 5. A reference to $sender_host_name in the part of a conditional expansion
6897 that was being skipped was still causing a DNS lookup. This no longer
6900 6. The def: expansion condition was not recognizing references to header lines
6901 that used bh_ and bheader_.
6903 7. Added the _cache feature to named lists.
6905 8. The code for checking quota_filecount in the appendfile transport was
6906 allowing one more file than it should have been.
6908 9. For compatibility with Sendmail, the command line option
6916 and sets the incoming protocol and host name (for trusted callers). The
6917 host name and its colon can be omitted when only the protocol is to be set.
6918 Note the Exim already has two private options, -pd and -ps, that refer to
6919 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6920 "s", but I don't think that's a major issue.
6922 10. A number of refactoring changes to the code, none of which should affect
6925 (a) The number of logging options was getting close to filling up the
6926 32-bit word that was used as a bit map. I have split them into two classes:
6927 those that are passed in the argument to log_write(), and those that are
6928 only ever tested independently outside of that function. These are now in
6929 separate 32-bit words, so there is plenty of room for expansion again.
6930 There is no change in the user interface or the logging behaviour.
6932 (b) When building, for example, log lines, the code previously used a
6933 macro that called string_cat() twice, in order to add two strings. This is
6934 not really sufficiently general. Furthermore, there was one instance where
6935 it was actually wrong because one of the argument was used twice, and in
6936 one call a function was used. (As it happened, calling the function twice
6937 did not affect the overall behaviour.) The macro has been replaced by a
6938 function that can join an arbitrary number of extra strings onto a growing
6941 (c) The code for expansion conditions now uses a table and a binary chop
6942 instead of a serial search (which was left over from when there were very
6943 few conditions). Also, it now recognizes conditions like "pam" even when
6944 the relevant support is not compiled in: a suitably worded error message is
6945 given if an attempt is made to use such a condition.
6947 11. Added ${time_interval:xxxxx}.
6949 12. A bug was causing one of the ddress fields not to be passed back correctly
6950 from remote delivery subprocesses. The field in question was not being
6951 subsequently used, so this caused to problems in practice.
6953 13. Added new log selectors queue_time and deliver_time.
6955 14. Might have fixed a bug in maildirsizefile handling that threw up
6956 "unexpected character" debug warnings, and recalculated the data
6957 unnecessarily. In any case, I expanded the warning message to give more
6960 15. Added the message "Restricted characters in address" to the statements in
6961 the default ACL that block characters like @ and % in local parts.
6963 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6964 Three changes have been made:
6966 (a) There was a serious bug; a negative response to MAIL caused the whole
6967 recipient domain to be cached as invalid, thereby blocking all messages
6968 to all local parts at the same domain, from all senders. This bug has
6969 been fixed. The domain is no longer cached after a negative response to
6970 MAIL if the sender used is not empty.
6972 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6975 (c) A new callout option, "use_sender" has been added for people who want
6976 the modified behaviour.
6982 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6985 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6986 indeed breaks things for older releases.
6988 3. Added additional logging to the case where there is a problem reading data
6989 from a filter that is running in a subprocess using a pipe, in order to
6990 try to track down a specific problem.
6992 4. Testing facility fudge: when running in the test harness and attempting
6993 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6994 getting "No route to host". Convert this to a timeout.
6996 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6999 6. Some OS don't have socklen_t but use size_t instead. This affects the
7000 fifth argument of getsockopt() amongst other things. This is now
7001 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7002 can be set for individual OS. I have set it for SunOS5, OSF1, and
7003 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7004 some earlier ones do not.
7006 7. Change 4.30/15 was not doing the test caselessly.
7008 8. The standard form for an IPv6 address literal was being rejected by address
7009 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7010 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7011 this, as well as the form without the "IPv6" on the front (but only when
7012 address literals are enabled, of course).
7014 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7016 10. Exim crashed if a message with an empty sender address specified by -f
7017 encountered a router with an errors_to setting. This could be provoked only
7018 by a command such as
7022 where an empty string was supplied; "<>" did not hit this bug.
7024 11. Installed PCRE release 4.5.
7026 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7027 remained set. It is now erased.
7029 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7030 times from message ids (which are base 36 rather than the normal 62).
7032 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7033 were being counted as actual protocol errors, and logged if the log
7034 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7035 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7036 and DATA following a set of rejected RCPTs do not count as protocol errors.
7037 In other words, Exim assumes they were pipelined, though this may not
7038 actually be the case. Of course, in all cases the client gets an
7039 appropriate error code.
7041 15. If a lookup fails in an ACL condition, a message about the failure may
7042 be available; it is used if testing the ACL cannot continue, because most
7043 such messages specify what the cause of the deferral is. However, some
7044 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7045 that caused an old message to be retained and used if a later statement
7046 caused a defer, replacing the real cause of the deferral.
7048 16. If an IP address had so many PTR records that the DNS lookup buffer
7049 was not large enough to hold them, Exim could crash while trying to process
7050 the truncated data. It now detects and logs this case.
7052 17. Further to 4.21/58, another change has been made: if (and only if) the
7053 first line of a message (the first header line) ends with CRLF, a bare LF
7054 in a subsequent header line has a space inserted after it, so as not to
7055 terminate the header.
7057 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7058 unnecessarily, used atoi() instead of strtol(), and didn't check the
7059 termination when getting file sizes from file names by regex.
7061 19. Completely re-implemented the support for maildirsize files, in the light
7062 of a number of problems with the previous contributed implementation
7063 (4.30/29). In particular:
7065 . If the quota is zero, the maildirsize file is maintained, but no quota is
7068 . If the maildir directory does not exist, it is created before any attempt
7069 to write a maildirsize file.
7071 . The quota value in the file is just a cache; if the quota is changed in
7072 the transport, the new value overrides.
7074 . A regular expression is available for excluding directories from the
7077 20. The autoreply transport checks the characters in options that define the
7078 message's headers; it allows continued headers, but it was checking with
7079 isspace() after an embedded newline instead of explicitly looking for a
7082 21. If all the "regular" hosts to which an address was routed had passed their
7083 expiry times, and had not reached their retry times, the address was
7084 bounced, even if fallback hosts were defined. Now Exim should go on to try
7087 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7088 equivalent code in the SMTP transport. Some hosts send humungous responses
7089 to HELO/EHLO, more than 1024 it seems.
7091 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7092 gives compiler warnings in some environments. I've now done it "properly",
7095 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7096 (because of problems with the built-in one) was declared to return uschar *
7097 instead of char *, causing compiler failure.
7099 25. Fixed a file descriptor leak when processing alias/forward files.
7101 26. Fixed a minor format string issue in dbfn.c.
7103 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7105 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7106 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7109 29. When a sender address is verified, it is cached, to save repeating the test
7110 when there is more than one recipient in a message. However, when the
7111 verification involves a callout, it is possible for different callout
7112 options to be set for different recipients. It is too complicated to keep
7113 track of this in the cache, so now Exim always runs a verification when a
7114 callout is required, relying on the callout cache for the optimization.
7115 The overhead is duplication of the address routing, but this should not be
7118 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7119 to be verified with callout=postmaster, and the main callout worked but the
7120 postmaster check failed, the verification correctly failed. However, if a
7121 subsequent RCPT command asked for sender verification *without* the
7122 postmaster check, incorrect caching caused this verification also to fail,
7125 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7126 it was not caching the DNS options (qualify_single, search_parents) that
7127 were used when the lookup failed. A subsequent lookup with different
7128 options therefore always gave the same answer, though there were cases
7129 where it should not have. (Example: a "domains = !$mx_any" option on a
7130 dnslookup router: the "domains" option is always processed without any
7131 widening, but the router might have qualify_single set.) Now Exim uses the
7132 cached value only when the same options are set.
7134 32. Added John Jetmore's "exipick" utility to the distribution.
7136 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7137 than a timeout (e.g. a certificate is required, and is not provided), an
7138 Exim server now closes the connection immediately. Previously it waited for
7139 the client to close - but if the client is SSL, it seems that they each
7140 wait for each other, leading to a delay before one of them times out.
7142 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7143 maintained 0.8.x compatibility because I don't think many are using it, and
7144 it is clearly obsolete.
7146 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7149 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7150 client certificate was expired. A simple patch fixes this, though I don't
7151 understand the full logic of why the verify callback is called multiple
7154 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7155 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7156 which causes problems with some clients (such as the Certicom SSL Plus
7157 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7158 disables the coutermeasure allowing Eudora to connect."
7160 38. Exim was not checking that a write() to a log file succeeded. This could
7161 lead to Bad Things if a log got too big, in particular if it hit a file
7162 size limit. Exim now panics and dies if it cannot write to a log file, just
7163 as it does if it cannot open a log file.
7165 39. Modified OS/Makefile-Linux so that it now contains
7167 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7169 The two -D definitions ensure that Exim is compiled with large file
7170 support, which makes it possible to handle log files that are bigger than
7173 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7174 instance) a domain was checked against a named list that involved a lookup,
7175 causing $domain_data to be set, then another domain was checked against the
7176 same list, then the first domain was re-checked, the value of $domain_data
7177 after the final check could be wrong. In particular, if the second check
7178 failed, it could be set empty. This bug probably also applied to
7181 41. The strip_trailing_dot option was not being applied to the address given
7182 with the -f command-line option.
7184 42. The code for reading a message's header from the spool was incrementing
7185 $received_count, but never initializing it. This meant that the value was
7186 incorrect (doubled) while delivering a message in the same process in which
7187 it was received. In the most common configuration of Exim, this never
7188 happens - a fresh exec is done - but it can happen when
7189 deliver_drop_privilege is set.
7191 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7192 it now includes up to 150 characters of the unexpected data in the log
7195 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7196 and building data strings. The size of both of these buffers was 10 000
7197 bytes - far larger than anybody would *ever* want, thought I. Needless to
7198 say, somebody hit the limit. I have increased the maximum line length to
7199 20 000 and the maximum data length of concatenated lines to 100 000. I have
7200 also fixed two bugs, because there was no checking on these buffers. Tsk,
7201 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7202 buffer is too small.
7204 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7205 lsearch lookups. Now it does.
7207 46. When parsing a route_list item in a manualroute router, a fixed-length
7208 buffer was used for the list of hosts. I made this 1024 bytes long,
7209 thinking that nobody would ever have a list of hosts that long. Wrong.
7210 Somebody had a whole pile of complicated expansion conditions, and the
7211 string was silently truncated, leading to an expansion error. It turns out
7212 that it is easier to change to an unlimited length (owing to other changes
7213 that have happened since this code was originally written) than to build
7214 structure for giving a limitation error. The length of the item that
7215 expands into the list of hosts is now unlimited.
7217 47. The lsearch lookup could not handle data where the length of text line was
7218 more than 4095 characters. Such lines were truncated, leading to shortened
7219 data being returned. It should now handle lines of any length.
7221 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7222 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7225 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7226 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7227 no longer happens in comment lines. A list of replacements is now placed
7228 at the head of all of the source files, except those whose only change is
7229 to replace PERL_COMMAND in the very first #! line.
7231 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7232 messages on the queue, with a bottom-up merge sort, using code contributed
7233 by Michael Haardt. This should make operations like -bp somewhat faster on
7234 large queues. It won't affect queue runners, except when queue_run_in_order
7237 51. Installed eximstats 1.31 in the distribution.
7239 52. Added support for SRV lookups to the dnslookup router.
7241 53. If an ACL referred to $message_body or $message_body_end, the value was not
7242 reset for any messages that followed in the same SMTP session.
7244 54. The store-handling optimization for building very long strings was not
7245 differentiating between the different store pools. I don't think this
7246 actually made any difference in practice, but I've tidied it.
7248 55. While running the routers to verify a sender address, $sender_address
7249 was still set to the sender address. This is wrong, because when routing to
7250 send a bounce to the sender, it would be empty. Therefore, I have changed
7251 it so that, while verifying a sender address, $sender_address is set to <>.
7252 (There is no change to what happens when verifying a recipient address.)
7254 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7255 target A or AAAA records (if not already returned) without resetting the
7256 qualify_single or search_parents options of the DNS resolver. These are
7257 inappropriate in this case because the targets of MX and SRV records must
7258 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7259 target that, when qualified, matched something in the local domain. These
7260 two options are now turned off when doing these lookups.
7262 57. It seems that at least some releases of Reiserfs (which does not have the
7263 concept of a fixed number of inodes) returns zero and not -1 for the
7264 number of available inodes. This interacted badly with check_spool_inodes,
7265 which assumed that -1 was the "no such thing" setting. What I have done is
7266 to check that the total number of inodes is greater than zero before doing
7267 the test of how many are available.
7269 58. When a "warn" ACL statement has a log_message modifier, the message is
7270 remembered, and not repeated. This is to avoid a lot of repetition when a
7271 message has many recipients that cause the same warning to be written.
7272 However, Exim was preserving the list of already written lines for an
7273 entire SMTP session, which doesn't seem right. The memory is now reset if a
7274 new message is started.
7276 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7277 debugging output unless log_rewrite was also set.
7279 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7280 of (int)(handle) when we know that handle contains (void *)(-1).
7282 61. The Exim daemon panic-logs an error return when it closes the incoming
7283 connection. However "connection reset by peer" seems to be common, and
7284 isn't really an error worthy of noting specially, so that particular error
7287 62. When Exim is trying to find all the local interfaces, it used to panic and
7288 die if the ioctl to get the interface flags failed. However, it seems that
7289 on at least one OS (Solaris 9) it is possible to have an interface that is
7290 included in the list of interfaces, but for which you get a failure error
7291 for this call. This happens when the interface is not "plumbed" into a
7292 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7293 failure of the "get flags" call assumes that the interface is down.
7295 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7296 makes life easier for people who are doing arithmetic on fields extracted
7297 from dates, where you often get leading zeros that should not be
7298 interpreted as octal.
7300 64. Added qualify_domain to the redirect router, to override the global
7303 65. If a pathologically long header line contained very many addresses (the
7304 report of this problem mentioned 10 000) and each of them was rewritten,
7305 Exim could use up a very large amount of memory. (It kept on making new
7306 copies of the header line as it rewrote, and never released the old ones.)
7307 At the expense of a bit more processing, the header rewriting function has
7308 been changed so that it no longer eats memory in this way.
7310 66. The generation of the Received: header has been moved from the time that a
7311 message starts to be received, to the time that it finishes. The timestamp
7312 in the Received: header should now be very close to that of the <= log
7313 line. There are two side-effects of this change:
7315 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7316 logged header lines no longer include the local Received: line, because
7317 it has not yet been created. The same applies to a copy of the message
7318 that is returned to a non-SMTP sender when a message is rejected.
7320 (b) When a filter file is tested using -bf, no additional Received: header
7321 is added to the test message. After some thought, I decided that this
7324 This change does not affect the value of $received_for. It is still set
7325 after address rewriting, but before local_scan() is called.
7327 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7329 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7330 gave an unhelpful panic error message, and a defer error. I have managed to
7331 change this behaviour so that it now rejects any supplied certificate,
7332 which seems right, as the list of acceptable certificates is empty.
7334 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7335 gave an unhelpful defer error. I have not managed to make this reject any
7336 supplied certificates, but the error message it gives is "no certificate
7337 supplied", which is not helpful.
7339 70. exigrep's output now also includes lines that are not associated with any
7340 message, but which match the given pattern. Implemented by a patch from
7341 Martin Sluka, which also tidied up the Perl a bit.
7343 71. Recipient callout verification, like sender verification, was using <> in
7344 the MAIL FROM command. This isn't really the right thing, since the actual
7345 sender may affect whether the remote host accepts the recipient or not. I
7346 have changed it to use the actual sender in the callout; this means that
7347 the cache record is now keyed on a recipient/sender pair, not just the
7348 recipient address. There doesn't seem to be a real danger of callout loops,
7349 since a callout by the remote host to check the sender would use <>.
7350 [SEE ABOVE: changed after hitting problems.]
7352 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7353 temporary errors. However, in the case of such a code being given after
7354 the end of a data transmission (i.e. after ".") Exim was failing to write
7355 a retry record for the message. (Yes, there was some broken host that was
7356 actually sending 8xx at this point.)
7358 73. An unknown lookup type in a host list could cause Exim to panic-die when
7359 the list was checked. (An example that provoked this was putting <; in the
7360 middle of a list instead of at the start.) If this happened during a DATA
7361 ACL check, a -D file could be left lying around. This kind of configuration
7362 error no longer causes Exim to die; instead it causes a defer error. The
7363 incident is still logged to the main and panic logs.
7365 74. Buglet left over from Exim 3 conversion. The message "too many messages
7366 in one connection" was written to the rejectlog but not the mainlog, except
7367 when address rewriting (yes!) was being logged.
7369 75. Added write_rejectlog option.
7371 76. When a system filter was run not as root (that is, when system_filter_user
7372 was set), the values of the $n variables were not being returned to the
7373 main process; thus, they were not subsequently available in the $sn
7376 77. Added +return_path_on_delivery log selector.
7378 78. A connection timeout was being treated differently from recipients deferred
7379 when testing hosts_max_try with a message that was older than the host's
7380 retry timeout. (The host should not be counted, thus allowing all hosts to
7381 be tried at least once before bouncing.) This may have been the cause of an
7382 occasionally reported bug whereby a message would remain on the queue
7383 longer than the retry timeout, but would be bounced if a delivery was
7384 forced. I say "may" because I never totally pinned down the problem;
7385 setting up timeout/retry tests is difficult. See also the next item.
7387 79. The ultimate address timeout was not being applied to errors that involved
7388 a combination of host plus message (for example, a timeout on a MAIL
7389 command). When an address resolved to a number of possible hosts, and they
7390 were not all tried for each delivery (e.g. because of hosts_max_try), a
7391 message could remain on the queue longer than the retry timeout.
7393 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7396 81. Fixed an obscure SMTP outgoing bug which required at least the following
7397 conditions: (a) there was another message waiting for the same server;
7398 (b) the server returned 5xx to all RCPT commands in the first message so
7399 that the message was not completed; (c) the server dropped the connection
7400 or gave a negative response to the RSET that Exim sends to abort the
7401 transaction. The observed case was a dropped connection after DATA that had
7402 been sent in pipelining mode. That is, the server had advertised PIPELINING
7403 but was not implementing it correctly. The effect of the bug was incorrect
7404 behaviour, such as trying another host, and this could lead to a crash.
7410 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7411 and daemon.c were passed as pointers to ints; they should have been
7412 pointers to socklen_t variables (which are typically unsigned ints).
7414 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7417 3. Fixed a really odd bug that affected only the testing scheme; patching a
7418 certain fixed string in the binary changed the value of another string that
7419 happened to be identical to the end of the original first string.
7421 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7422 name", it returns that address as the IP address. On some operating
7423 systems (e.g. Solaris), it also passes back the IP address string as the
7424 "host name". However, on others (e.g. Linux), it passes back an empty
7425 string. Exim wasn't checking for this, and was changing the host name to an
7426 empty string, assuming it had been canonicalized.
7428 5. Although rare, it is permitted to have more than one PTR record for a given
7429 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7430 all the names associated with an address, because they do in Solaris.
7431 However, it seems that they do not in Linux for data that comes from the
7432 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7433 I found this out when I moved to a new Linux workstation and tried to run
7434 the Exim test suite.
7436 To get round this problem I have changed the code so that it now does its
7437 own call to the DNS to look up PTR records when searching for a host name.
7438 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7439 addresses that are only in /etc/hosts are still found.
7441 This behaviour is, however, controlled by an option called host_lookup_
7442 order, which defaults to "bydns:byaddr". If people want to use the other
7443 order, or indeed, just use one or the other means of lookup, they can
7444 specify it in this variable.
7446 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7447 some operating systems, this comes back from gethostbyaddr() as an empty
7448 string, and this is what Exim used to test for. However, it seems that in
7449 other systems, "." is yielded. Exim now tests for this case too.
7451 7. The values of check_spool_space and check_log_space are now held internally
7452 as a number of kilobytes instead of an absolute number of bytes. If a
7453 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7454 kilobyte. This means that much larger values can be stored.
7456 8. Exim monitor: an attempt to get the action menu when not actually pointing
7457 at a message produces an empty menu entitled "No message selected". This
7458 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7459 no entries in it ("Shell widget menu has zero width and/or height"). So I
7460 have added a single, blank menu entry in this case.
7462 9. Added ${quote_local_part.
7464 10. MIME decoding is now applied to the contents of Subject: header lines when
7467 11. Now that a reference to $sender_host_address automatically causes a reverse
7468 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7469 host lookup before query-style lookups in lists that might use this
7470 variable. This has therefore been abolished, and the "net-" prefix is no
7471 longer necessary for query-style lookups.
7473 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7474 have been a typo for LFLAGS, so it has been changed.
7476 13. The install script calls Exim with "-C /dev/null" in order to find the
7477 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7478 to be output. However, since Exim outputs its version number before the
7479 error, it didn't break the script. It just looked ugly. I fixed this by
7480 always allowing "-C /dev/null" if the caller is root.
7482 14. Ignore overlarge ACL variable number when reading spool file - insurance
7483 against a later release with more variables having written the file.
7485 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7486 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7487 this, as well as the form without the "IPv6" on the front.
7489 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7490 OS/Makefile-Darwin file.
7492 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7493 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7495 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7496 variable. However, some OS use ints for inodes. I've added cast to long int
7497 to get rid of the compiler warning.
7499 19. I had forgotten to lock out "/../" in configuration file names when
7500 ALT_CONFIG_PREFIX was set.
7502 20. Routers used for verification do not need to specify transports. However,
7503 if such a router generated a host list, and callout was configured, Exim
7504 crashed, because it could not find a port number from the (non-existent)
7505 transport. It now assumes port 25 in this circumstance.
7507 21. Added the -t option to exigrep.
7509 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7510 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7511 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7512 them is set, LOOKUP_LSEARCH is forced.
7514 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7516 24. Added sender and host information to the "rejected by local_scan()" log
7517 line; previously there was no indication of these.
7519 25. Added .include_if_exists.
7521 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7522 Linux. It turns out that not all file systems support this. Apparently some
7523 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7524 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7525 not supported on the file descriptor, is now ignored when Exim is trying to
7526 sync a directory. This applies only to Linux.
7528 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7530 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7531 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7534 29. Added support for maildirsize files from supplied patch (modified a bit).
7536 30. The use of :fail: followed by an empty string could lead Exim to respond to
7537 sender verification failures with (e.g.):
7539 550 Verification failed for <xxx>
7540 550 Sender verify failed
7542 where the first response line was missing the '-' that indicates it is not
7543 the final line of the response.
7545 31. The loop for finding the name of the user that called Exim had a hardwired
7546 limit of 10; it now uses the value of finduser_retries, which is used for
7547 all other user lookups.
7549 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7552 33. Exim was neglecting to zero errno before one call of strtol() when
7553 expanding a string and expecting an integer value. On some systems this
7554 resulted in spurious "integer overflow" errors. Also, it was casting the
7555 result into an int without checking.
7557 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7558 did not work. The code looks as if it has *never* worked, though it appears
7559 to have been documented since at least release 1.62. I have made it work.
7561 35. The "timeout_DNS" error in retry rules, also documented since at least
7562 1.62, also never worked. As it isn't clear exactly what this means, and
7563 clearly it isn't a major issue, I have abolished the feature by treating it
7564 as "timeout", and writing a warning to the main and panic logs.
7566 36. The display of retry rules for -brt wasn't always showing the error code
7569 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7570 timeout_connect_A, timeout_connect_MX.
7572 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7573 to the empty sender.
7575 39. The daemon was not analysing the content of -oX till after it had closed
7576 stderr and disconnected from the controlling terminal. This meant that any
7577 syntax errors were only noted on the panic log, and the return code from
7578 the command was 0. By re-arranging the code a little, I've made the
7579 decoding happen first, so such errors now appear on stderr, and the return
7580 code is 1. However, the actual setting up of the sockets still happens in
7581 the disconnected process, so errors there are still only recorded on the
7584 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7585 connections (as happens on some IP stacks) was logged at start up time as
7586 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7587 it from "IPv6 and IPv4", which means that two separate sockets are being
7590 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7591 says whether AF_INET or AF_INET6 was passed as an argument.
7593 42. Exiwhat output was messed up when time zones were included in log
7596 43. Exiwhat now gives more information about the daemon's listening ports,
7597 and whether -tls-on-connect was used.
7599 44. The "port" option of the smtp transport is now expanded.
7601 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7602 silently ignored. Now an error message is written to the main and panic
7605 46. There's a new ACL modifier called "logwrite" which writes to a log file
7606 as soon as it is encountered.
7608 47. Added $local_user_uid and $local_user_gid at routing time.
7610 48. Exim crashed when trying to verify a sender address that was being
7613 49. Exim was recognizing only a space character after ".include". It now also
7614 recognizes a tab character.
7616 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7617 extracting the relevant information from the specification. The man page no
7618 longer contains scrambled data for the -d option, and I've added a section
7619 at the front about calling Exim under different names.
7621 51. Added "extra_headers" argument to the "mail" command in filter files.
7623 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7626 53. Installed eximstats 1.29.
7628 54. Added transport_filter_timeout as a generic transport option.
7630 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7631 Cc: header lines. This was required by RFC 822, but it not required by RFC
7634 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7635 incoming messages that did not have them. Now it does so only if the
7636 message originates locally, that is, if there is no associated remote host
7637 address. When Resent- header lines are present, this applies to the Resent-
7638 lines rather than the non-Resent- lines.
7640 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7641 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7643 58. Messages for configuration errors now include the name of the main
7644 configuration file - useful now that there may be more than one file in a
7645 list (.included file names were always shown).
7647 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7648 for those rare installations that do not start the daemon as root or run it
7649 setuid root. I've cut out the call to initgroups() if the daemon is not
7652 60. The Exim user and group can now be bound into the binary as text strings
7653 that are looked up at the start of Exim's processing.
7655 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7657 62. Added $mailstore_basename variable.
7659 63. Installed patch to sieve.c from Michael Haardt.
7661 64. When Exim failed to open the panic log after failing to open the main log,
7662 the original message it was trying to log was written to stderr and debug
7663 output, but if they were not available (the usual case in production), it
7664 was lost. Now it is written to syslog before the two lines that record the
7665 failures to open the logs.
7667 65. Users' Exim filters run in subprocesses under the user's uid. It is
7668 possible for a "deliver" command or an alias in a "personal" command to
7669 provoke an address rewrite. If logging of address rewriting is configured,
7670 this fails because the process is not running as root or exim. There may be
7671 a better way of dealing with this, but for the moment (because 4.30 needs
7672 to be released), I have disabled address rewrite logging when running a
7673 filter in a non-root, non-exim process.
7679 1. The buildconfig auxiliary program wasn't quoting the value set for
7680 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7681 not defined. This bug was masked in 4.22 by the effect that was fixed in
7684 2. Some messages that were rejected after a message id was allocated were
7685 shown as "incomplete" by exigrep. It no longer does this for messages that
7686 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7688 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7689 have allow_domain_literals set, the ID did not get logged in the <= line.
7690 Domain literals are now always recognized in Message-ID: header lines.
7692 4. The first argument for a ${extract expansion item is the key name or field
7693 number. Leading and trailing spaces in this item were not being ignored,
7694 causing some misleading effects.
7696 5. When deliver_drop_privilege was set, single queue runner processes started
7697 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7698 same command in the process it spins off) were not dropping privilege.
7700 6. When the daemon running as "exim" started a queue runner, it always
7701 re-executed Exim in the spun-off process. This is a waste of effort when
7702 deliver_drop_privilege is set. The new process now just calls the
7703 queue-runner function directly.
7709 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7712 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7713 ignored. Though the use of -oP was forcing the writing of a pid file, it
7714 was always written to the default place.
7716 3. If the message "no IP address found for host xxxx" is generated during
7717 incoming verification, it is now followed by identification of the incoming
7718 connection (so you can more easily find what provoked it).
7720 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7722 5. Added some features to "harden" Exim a bit more against certain attacks:
7724 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7725 be put in Local/Makefile. This is like the never_users runtime option,
7726 but it cannot be overridden. The default setting is "root".
7728 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7729 prefix string with which any file named in a -C command line option
7732 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7733 is retained for -C and -D only if the caller of Exim is root. Without
7734 it, the exim user may also use -C and -D and retain privilege.
7736 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7737 command line option is disabled.
7739 6. Macro names set by the -D option must start with an upper case letter, just
7740 like macro names defined in the configuration file.
7742 7. Added "dereference=" facility to LDAP.
7744 8. Two instances of the typo "uknown" in the source files are fixed.
7746 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7747 the Configure-Makefile script screwed up while processing it.
7749 10. Incorporated PCRE 4.4.
7751 11. The SMTP synchronization check was not operating right at the start of an
7752 SMTP session. For example, it could not catch a HELO sent before the client
7753 waited for the greeting. There is now a check for outstanding input at the
7754 point when the greeting is written. Because of the duplex, asynchronous
7755 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7756 way, but not yet received, when the check is performed.
7758 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7759 on TCP/IP sockets, because this apparently causes some broken clients to
7762 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7763 unchanged) from the Cygwin maintainer.
7765 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7766 was supported instead of testing for mbx. Effectively a typo.
7768 15. The spa authenticator server code was not checking that the input it
7769 received was valid base64.
7771 16. The debug output line for the "set" modifier in ACLs was not showing the
7772 name of the variable that was being set.
7774 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7776 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7777 lookup on its own. Something else was needed to trigger the lookup. For
7778 example, a match in host_lookup or the need for a host name in a host list.
7779 Now, if $sender_host_name is referenced and the host name has not yet been
7780 looked up, a lookup is performed. If the lookup fails, the variable remains
7781 empty, and $host_lookup_failed is set to "1".
7783 19. Added "eqi" as a case-independent comparison operator.
7785 20. The saslauthd authentication condition could segfault if neither service
7786 nor realm was specified.
7788 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7789 error message that was logged was misleading, and incoming SMTP
7790 connections were dropped. The message is now more accurate, and temporary
7791 errors are given to SMTP connections.
7793 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7794 (except RSET) with a 421 error, until QUIT is received. However, it was
7795 failing to send a response to QUIT.
7797 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7798 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7799 ACL is now run after the helo_try_verify_hosts code.
7801 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7802 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7803 case-independent, so other case variants are also recognized). Apparently
7804 some systems use these upper case variants.
7806 25. If more than two messages were waiting for the same host, and a transport
7807 filter was specified for the transport, Exim sent two messages over the
7808 same TCP/IP connection, and then failed with "socket operation on non-
7809 socket" when it tried to send the third.
7811 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7813 27. The extern definition of crypt16() in expand.c was not being excluded when
7814 the OS had its own crypt16() function.
7816 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7817 as a preferred synonym for return_size_limit, both as an option and as an
7820 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7822 30. Changed the default configuration ACL to relax the local part checking rule
7823 for addresses that are not in any local domains. For these addresses,
7824 slashes and pipe symbols are allowed within local parts, but the sequence
7825 /../ is explicitly forbidden.
7827 31. SPA server authentication was not clearing the challenge buffer before
7830 32. log_message in a "warn" ACL statement was writing to the reject log as
7831 well as to the main log, which contradicts the documentation and doesn't
7832 seem right (because no rejection is happening). So I have stopped it.
7834 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7835 However, I am unable to do any testing of this.
7837 34. Fixed an infelicity in the appendfile transport. When checking directories
7838 for a mailbox, to see if any needed to be created, it was accidentally
7839 using path names with one or more superfluous leading slashes; tracing
7840 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7842 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7843 contents are added to the log line that is written for every discarded
7844 recipient. (Previously a log_message setting was ignored.)
7846 36. The ${quote: operator now quotes the string if it is empty.
7848 37. The install script runs exim in order to find its version number. If for
7849 some reason other than non-existence or emptiness, which it checks, it
7850 could not run './exim', it was installing it with an empty version number,
7851 i.e. as "exim-". This error state is now caught, and the installation is
7854 38. An argument was missing from the function that creates an error message
7855 when Exim fails to connect to the socket for saslauthd authentication.
7856 This could cause Exim to crash, or give a corrupted message.
7858 39. Added isip, isip4, and isip6 to ${if conditions.
7860 40. The ACL variables $acl_xx are now saved with the message, and can be
7861 accessed later in routers, transports, and filters.
7863 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7864 strings in the file are not string-expanded.
7866 42. If a MAIL command specified a SIZE value that was too large to fit into an
7867 int variable, the check against message_size_limit failed. Such values are
7868 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7869 day this will have to be increased, but I don't think I want to be around
7870 when emails are that large.
7877 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7878 iconv() is not standard in FreeBSD.
7880 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7881 IPv6 enabled. The observed symptom was a segmentation fault on return from
7882 the function os_common_find_running_interfaces() in src/os.c.
7884 3. In the check_special_case() function in daemon.c I had used "errno" as an
7885 argument name, which causes warnings on some systems. This was basically a
7886 typo, since it was named "eno" in the comments!
7888 4. The code that waits for the clock to tick (at a resolution of some fraction
7889 of a second) so as to ensure message-id uniqueness was always waiting for
7890 at least one whole tick, when it could have waited for less. [This is
7891 almost certainly not relevant at current processor speeds, where it is
7892 unlikely to ever wait at all. But we try to future-proof.]
7894 5. The function that sleeps for a time interval that includes fractions of a
7895 second contained a race. It did not block SIGALRM between setting the
7896 timer, and suspending (a couple of lines later). If the interval was short
7897 and the sigsuspend() was delayed until after it had expired, the suspension
7898 never ended. On busy systems this could lead to processes getting stuck for
7901 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7902 process, before it forks any delivery processes. The open lookup caching
7903 mechanism meant that the open file or database connection was passed into
7904 the delivery process. The problem was that delivery processes always tidy
7905 up cached lookup data. This could cause a problem for the next delivery
7906 process started by the queue runner, because the external queue runner
7907 process does not know about the closure. So the next delivery process
7908 still has data in the lookup cache. In the case of a file lookup, there was
7909 no problem because closing a file descriptor in a subprocess doesn't affect
7910 the parent. However, if the lookup was caching a connection to a database,
7911 the connection was closed, and the second delivery process was likely to
7912 see errors such as "PGSQL: query failed: server closed the connection
7913 unexpectedly". The problem has been fixed by closing all cached lookups
7914 in a queue runner before running a delivery process.
7916 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7917 seem to have the "const" qualifier which it has on other OS. I've
7920 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7921 *of the same type* (client or server) with the same public name that an
7922 error should be diagnosed.
7924 9. When Exim looked up a host name for an IP address, but failed to find the
7925 original IP address when looking up the host name (a safety check), it
7926 output the message "<ip address> does not match any IP for NULL", which was
7927 confusing, to say the least. The bug was that the host name should have
7928 appeared instead of "NULL".
7930 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7931 user that is built into the binary, and the -C or -D options is used, root
7932 privilege is dropped before the configuration file is read. In addition,
7933 logging is switched to stderr instead of the normal log files. If the
7934 configuration then re-defines the Exim user, the unprivileged environment
7935 is probably not what is expected, so Exim logs a panic warning message (but
7938 However, if deliver_drop_privilege is set, the unprivileged state may well
7939 be exactly what is intended, so the warning has been cut out in that case,
7940 and Exim is allowed to try to write to its normal log files.
7946 1. smtp_return_error_details was not giving details for temporary sender
7947 or receiver verification errors.
7949 2. Diagnose a configuration error if two authenticators have the same public
7952 3. Exim used not to create the message log file for a message until the first
7953 delivery attempt. This could be confusing when incoming messages were held
7954 for policy or load reasons. The message log file is now created at the time
7955 the message is received, and an initial "Received" line is written to it.
7957 4. The automatically generated man page for command line options had a minor
7958 bug that caused no ill effects; however, a more serious problem was that
7959 the procedure for building the man page automatically didn't always
7960 operate. Consequently, release 4.20 contains an out-of-date version. This
7961 shouldn't happen again.
7963 5. When building Exim with embedded Perl support, the script that builds the
7964 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7965 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7967 6. The freeze_tell option was not being used for messages that were frozen on
7968 arrival, either by an ACL or by local_scan().
7970 7. Added the smtp_incomplete_transaction log selector.
7972 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7973 was accepting AUTH without a new EHLO.
7975 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7976 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7979 10. Logging of TCP/IP connections (when configured) now happens in the main
7980 daemon process instead of the child process, so that the TCP/IP connection
7981 count is more accurate (but it can never be perfect).
7983 11. The use of "drop" in a nested ACL was not being handled correctly in the
7984 outer ACL. Now, if condition failure induced by the nested "drop" causes
7985 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7986 or "require"), the connection is dropped.
7988 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7989 that yield "discard" can now be used with an "accept" or a "discard" verb,
7990 but an error is generated for any others (because I can't see a useful way
7991 to define what should happen).
7993 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7994 are now processed in the same way as lines in the Exim configuration file.
7995 In particular, continuation lines are supported.
7997 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7999 15. Added -ti meaning -t -i.
8001 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8002 domains, and warn by logging if others are found.
8004 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8005 returned from GIFCONF when Exim is trying to find the list of interfaces on
8006 a host. The code in os.c has been modified to copy each ifreq to an aligned
8007 structure in all cases.
8009 Also, in some cases, the returned ifreq's were being copied to a 'struct
8010 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8011 means the last couple of bytes of an IPv6 address could be chopped if the
8012 ifreq contained only a normal sockaddr (14 bytes storage).
8014 18. Named domain lists were not supported in the hosts_treat_as_local option.
8015 An entry such as +xxxx was not recognized, and was treated as a literal
8018 19. Ensure that header lines added by a DATA ACL are included in the reject log
8019 if the ACL subsequently rejects the message.
8021 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8022 MD5 (which is deprecated).
8024 21. When testing a filter file using -bf, Exim was writing a message when it
8025 took the sender from a "From " line in the message, but it was not doing so
8026 when it took $return_path from a Return-Path: header line. It now does.
8028 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8029 with a valid header line field name (a series of printing characters
8030 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8032 23. Changed "disc" in the source to "disk" to conform to the documentation and
8033 the book and for uniformity.
8035 24. Ignore Sendmail's -Ooption=value command line item.
8037 25. When execve() failed while trying to run a command in a pipe transport,
8038 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8039 could be confused with a return value of 69 from the command itself. This
8040 has been changed to 127, the value the shell returns if it is asked to run
8041 a non-existent command. The wording for the related log line suggests a
8042 non-existent command as the problem.
8044 26. If received_header_text expands to an empty string, do not add a Received:
8045 header line to the message. (Well, it adds a token one on the spool, but
8046 marks it "old" so that it doesn't get used or transmitted.)
8048 27. Installed eximstats 1.28 (addition of -nt option).
8050 28. There was no check for failure on the call to getsockname() in the daemon
8051 code. This can fail if there is a shortage of resources on the system, with
8052 ENOMEM, for example. A temporary error is now given on failure.
8054 29. Contrary to the C standard, it seems that in some environments, the
8055 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8056 program. Exim now does this explicitly; it affects the formatting of
8057 timestamps using strftime().
8059 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8060 complaints. I've now initialized all the variables, to avoid this.
8062 32. Header lines added by a system filter were not being "seen" during
8063 transport-time rewrites.
8065 33. The info_callback() function passed to OpenSSL is set up with type void
8066 (*)(SSL *, int, int), as described somewhere. However, when calling the
8067 function (actually a macro) that sets it up, the type void(*)() is
8068 expected. I've put in a cast to prevent warnings from picky compilers.
8070 34. If a DNS black list lookup found a CNAME record, but there were no A
8071 records associated with the domain it pointed at, Exim crashed.
8073 35. If a DNS black list lookup returned more than one A record, Exim ignored
8074 all but the first. It now scans all returned addresses if a particular IP
8075 value is being sought. In this situation, the contents of the
8076 $dnslist_value variable are a list of all the addresses, separated by a
8079 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8080 to accept a host name and all its aliases if the forward lookup for any of
8081 them yielded the IP address of the incoming connection. Now it accepts only
8082 those names whose forward lookup yields the correct IP address. Any other
8083 names are discarded. This closes a loophole whereby a rogue DNS
8084 administrator could create reverse DNS records to break through a
8085 wildcarded host restriction in an ACL.
8087 37. If a user filter or a system filter that ran in a subprocess used any of
8088 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8089 the wrong values were passed to the pipe command ($thisaddress had the
8090 value of $0, $0 had the value of $1, etc). This bug was introduced by
8091 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8093 38. Improved the line breaking for long SMTP error messages from ACLs.
8094 Previously, if there was no break point between 40 and 75 characters, Exim
8095 left the rest of the message alone. Two changes have been made: (a) I've
8096 reduced the minimum length to 35 characters; (b) if it can't find a break
8097 point between 35 and 75 characters, it looks ahead and uses the first one
8098 that it finds. This may give the occasional overlong line, but at least the
8099 remaining text gets split now.
8101 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8102 file descriptors might be low, and that setting 1000 would always raise it.
8103 It turns out that in some environments, the limit is already over 1000 and
8104 that lowering it causes trouble. So now Exim takes care not to decrease it.
8106 40. When delivering a message, the value of $return_path is set to $sender_
8107 address at the start of routing (routers may change the value). By an
8108 oversight, this default was not being set up when an address was tested by
8109 -bt or -bv, which affected the outcome if any router or filter referred to
8112 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8113 log only when "message" or "log_message" are set. However, if one of the
8114 conditions was an address verification, or a call to a nested ACL, the
8115 messages generated by the underlying test were being passed through. This
8116 no longer happens. The underlying message is available in $acl_verify_
8117 message for both "message" and "log_message" expansions, so it can be
8118 passed through if needed.
8120 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8121 new expansion $bh_ to give the encoded byte string without charset
8122 translation. Translation happens only if iconv() is available; HAVE_ICONV
8123 indicates this at build time. HEADERS_CHARSET gives the charset to
8124 translate to; headers_charset can change it in the configuration, and
8125 "headers charset" can change it in an individual filter file.
8127 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8128 that creates RFC 2047 encoded "words" labels them as that charset instead
8129 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8130 expansion operator; (ii) when Exim creates a From: line for a local
8131 message; (iii) when a header line is rewritten to include a "phrase" part.
8133 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8134 buggy, causing it to skip the first lines of messages whose message ID
8135 ended in 'D'. This would not have bitten before Exim release 4.14, because
8136 message IDs were unlikely to end in 'D' before then. The effect was to have
8137 incorrect size information for certain domains.
8139 45. #include "config.h" was missing at the start of the crypt16.c module. This
8140 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8143 46. If there was a timeout during a "random" callout check, Exim treated it as
8144 a failure of the random address, and carried on sending RSET and the real
8145 address. If the delay was just some slowness somewhere, the response to the
8146 original RCPT would be taken as a response to RSET and so on, causing
8147 mayhem of various kinds.
8149 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8150 when I implemented it. It didn't allow for the fact that some option values
8151 may legitimately be negative (e.g. size_addition), and it didn't even do
8152 the right test for positive values.
8154 48. Domain names in DNS records are case-independent. Exim always looks them up
8155 in lower case. Some resolvers return domain names in exactly the case they
8156 appear in the zone file, that is, they may contain uppercase letters. Not
8157 all resolvers do this - some return always lower case. Exim was treating a
8158 change of case by a resolver as a change of domain, similar to a widening
8159 of a domain abbreviation. This triggered its re-routing code and so it was
8160 trying to route what was effectively the same domain again. This normally
8161 caused routing to fail (because the router wouldn't handle the domain
8162 twice). Now Exim checks for this case specially, and just changes the
8163 casing of the domain that it ultimately uses when it transmits the message
8166 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8169 50. If a filter generated a file delivery with a non-absolute name (possible if
8170 no home directory exists for the router), the forbid_file option was not
8173 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8174 the existing equality matching.
8176 52. Exim was using ints instead of ino_t variables in some places where it was
8177 dealing with inode numbers.
8179 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8180 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8181 called TMPDIR, and if it finds it is different, it changes its value.
8183 54. The smtp_printf() function is now made available to local_scan() so
8184 additional output lines can be written before returning. There is also an
8185 smtp_fflush() function to enable the detection of a dropped connection.
8186 The variables smtp_input and smtp_batched_input are exported to
8189 55. Changed the default runtime configuration: the message "Unknown user"
8190 has been removed from the ACL, and instead placed on the localuser router,
8191 using the cannot_route_message feature. This means that any verification
8192 failures that generate their own messages won't get overridden. Similarly,
8193 the "Unrouteable address" message that was in the ACL for unverifiable
8194 relay addresses has also been removed.
8196 56. Added hosts_avoid_esmtp to the smtp transport.
8198 57. The exicyclog script was not checking for the esoteric option
8199 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8200 will work only if exicyclog is run under the appropriate euid.
8202 58. Following a discussion on the list, the rules by which Exim recognises line
8203 endings on incoming messages have been changed. The -dropcr and drop_cr
8204 options are now no-ops, retained only for backwards compatibility. The
8205 following line terminators are recognized: LF CRLF CR. However, special
8206 processing applies to CR:
8208 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8209 nor a local message in the state where . is a terminator.
8211 (ii) If a bare CR is encountered in a header line, an extra space is added
8212 after the line terminator so as not to end the header. The reasoning
8213 behind this is that bare CRs in header lines are most likely either
8214 to be mistakes, or people trying to play silly games.
8216 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8217 was being incorrectly given as 18 bytes larger than it should have been.
8218 This is a VOB (very old bug).
8220 60. This may never have affected anything current, but just in case it has:
8221 When the local host is found other than at the start of a list of hosts,
8222 the local host, those with the same MX, and any that follow, are discarded.
8223 When the list in question was part of a longer list of hosts, the following
8224 hosts (not currently being processed) were also being discarded. This no
8225 longer happens. I'm not sure if this situation could ever has previously
8228 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8231 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8232 header. This is something that is recommended in a new Internet Draft, and
8233 is something that is documented as being done by Sendmail. There are two
8234 possible values. For messages generated by the autoreply transport, Exim
8237 Auto-Submitted: auto-replied
8239 whereas for all other generated messages (e.g. bounces) it adds
8241 Auto-Submitted: auto-generated
8243 63. The "personal" condition in filters now includes a test for the
8244 Auto-Submitted: header. If it contains the string "auto-" the message it
8245 not considered personal.
8247 64. Added rcpt_include_affixes as a generic transport option.
8249 65. Added queue_only_override (default true).
8251 66. Added the syslog_duplication option.
8253 67. If what should have been the first header line of a message consisted of
8254 a space followed by a colon, Exim was mis-interpreting it as a header line.
8255 It isn't of course - it is syntactically invalid and should therefore be
8256 treated as the start of the message body. The misbehaviour could have
8257 caused a number of strange effects, including loss of data in subsequent
8258 header lines, and spool format errors.
8260 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8261 client host had authenticated. This control can now be exercised by an ACL
8262 for more flexibility.
8264 69. By default, callouts do not happen when testing with -bh. There is now a
8265 variant, -bhc, which does actually run the callout code, including
8266 consulting and updating the callout cache.
8268 70. Added support for saslauthd authentication, courtesy of Alexander
8271 71. If statvfs() failed on the spool or log directories while checking their
8272 size for availability, Exim confusingly gave the error "space shortage".
8273 Furthermore, in debugging mode it crashed with a floating point exception.
8274 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8275 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8276 serious problem, Exim now writes to the main and panic logs when this
8277 happens, with details of the failure. It then refuses to accept the
8278 incoming message, giving the message "spool directory problem" or "log
8279 directory problem" with a 421 code for SMTP messages.
8281 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8282 0, 1, and 2 exist, because some OS complain for execs without them (see
8283 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8284 descriptors. However, the code omitted to check that the open succeeded,
8285 causing mysterious errors if for some reason the permissions on /dev/null
8286 got screwed. Now Exim writes a message to the main and panic logs, and
8287 bombs out if it can't open /dev/null.
8289 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8290 interact so that it is all more flexible. It is supposed to remain
8291 backwards compatible. Also added extra_local_interfaces.
8293 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8294 to bomb out with an assertion failure - to the client this appears as a
8295 connection drop. This problem occurs in the part of the code that was taken
8296 from the Samba project. Fortunately, the assertion is in a very simple
8297 function, so I have fixed this by reproducing the function inline in the
8298 one place where it is called, and arranging for authentication to fail
8299 instead of killing the process with assert().
8301 75. The SPA client code was not working when the server requested OEM rather
8302 than Unicode encoding.
8304 76. Added code to make require_files with a specific uid setting more usable in
8305 the case where statting the file as root fails - usually a non-root-mounted
8306 NFS file system. When this happens and the failure is EACCES, Exim now
8307 forks a subprocess and does the per-uid checking as the relevant uid.
8309 77. Added process_log_path.
8311 78. If log_file_path was not explicitly set, a setting of check_log_space or
8312 check_log_inodes was ignored.
8314 79. If a space check for the spool or log partitions fails, the incident is now
8315 logged. Of course, in the latter case the data may get lost...
8317 80. Added the %p formatting code to string_format() so that it can be used to
8318 print addresses in debug_print(). Adjusted all the address printing in the
8319 debugging in store.c to use %p rather than %d.
8321 81. There was a concern that a line of code in smtp_in.c could overflow a
8322 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8323 initially expressed, the concern was not well-founded, because trailing
8324 spaces are removed early. However, if the trailing spaces were followed by
8325 a NULL, they did not get removed, so the overflow was possible. Two fixes
8328 (a) I re-wrote the offending code in a cleaner fashion.
8329 (b) If an incoming SMTP command contains a NULL character, it is rejected
8332 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8333 runs initgroups(), so that if the Exim user is in any additional groups,
8334 they will be used during message reception.
8340 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob