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
69 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
70 after reception to before a subsequent reception. This should
71 mean slightly faster delivery, and also confirmation of reception
74 JH/02 Move from using the pcre library to pcre2. The former is no longer
75 being developed or supported (by the original developer).
77 JH/03 Constification work in the filters module required a major version
78 bump for the local-scan API. Specifically, the "headers_charset"
79 global which is visible via the API is now const and may therefore
80 not be modified by local-scan code.
82 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
83 sendfile() didi not account for the way the ClamAV driver code called it.
85 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
86 time check was being done for every character; replace that with one
89 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
90 sent was prefixed with a length byte.
92 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
93 RFC 5321. Previously Dovecot (at least) would log errors during
96 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
97 in 4.61 without a reason logged. Binaries will be bigger, which might
98 matter on diskspace-constrained systems, but debug is easier.
100 JH/09 Fix macro-definition during "-be" expansion testing. The move to
101 write-protected store for macros had not accounted for these runtime
102 additions; fix by removing this protection for "-be" mode.
104 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
105 handing out large-numbered file descriptors, violating the usual Unix
106 assumption (and required by Posix) that the lowest possible number will be
107 allocated by the kernel when a new one is needed. In the daemon, and any
108 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
109 are not useable for FD_SET() [and hence select()] and overwrite the stack.
110 Assorted crashes happen.
112 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
113 main-section options or in a connect ACL, the value from the first ever
114 connection was never replaced for subsequent connections. Found by
117 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
118 though only once PCRE2 was introduced: the memory accounting used under
119 debug offset allocations by an int, giving a hard trap in early startup.
120 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
122 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
123 with underbars is given. The write-protection of configuration introduced
124 in 4.95 trapped when normalisation was applied to an option not needing
127 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
129 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
130 failing to close the certificates directory, every hour or any time it
133 JH/16 Debugging initiated by an ACL control now continues through into routing
134 and transport processes. Previously debugging stopped any time Exim
135 re-execs, or for processing a queued message.
137 JH/17 The "expand" debug selector now gives more detail, specifically on the
138 result of expansion operators and items.
140 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
141 bad comparison between the option value and the name of the file to
142 be included was done, and a mismatch was wrongly identified.
143 4.88 to 4.95 are affected.
145 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
147 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
148 supplied as the db file-pair basename. Previously, if a directory
149 path was given, for example via the autoreply "once" option, the DB
150 file.pag and file.dir files would be created in that directory's
153 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
154 "taint" log_selector. These were previously deprecated.
156 JH/22 Fix static address-list lookups to properly return the matched item.
157 Previously only the domain part was returned.
159 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
160 the call into OpenSSL to send a TLS Close was being repeated; this
161 resulted in the library waiting for the peer's Close. If that was never
162 sent we waited forever. Fix by tracking send calls.
164 JH/24 The ${run} expansion item now expands its command string elements after
165 splitting. Previously it was before; the new ordering makes handling
166 zero-length arguments simpler. The old ordering can be obtained by
167 appending a new option "preexpand", after a comma, to the "run".
169 JH/25 Taint-check exec arguments for transport-initiated external processes.
170 Previously, tainted values could be used. This affects "pipe", "lmtp" and
171 "queryprogram" transport, transport-filter, and ETRN commands.
172 The ${run} expansion is also affected: in "preexpand" mode no part of
173 the command line may be tainted, in default mode the executable name
176 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
177 the the facility was not passed across execs, and only the first message
178 passed over a connection could use BDAT; any further ones using DATA.
180 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
181 uses $sending_ip_address and an interface is specified.
182 Previously any use of the local address in the EHLO name disabled
183 PIPECONNECT, the common case being to use the rDNS of it.
185 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
186 resumption. Previously verify failed because no certificate status is
187 passed on the wire for the restarted session. Fix by using the recorded
188 ocsp status of the stored session for the new connection.
190 JH/29 TLS resumption: the key for session lookup in the client now includes
191 more info that a server could potentially use in configuring a TLS
192 session, avoiding oferring mismatching sessions to such a server.
193 Previously only the server IP was used.
195 JH/30 Fix string_copyn() for limit greater than actual string length.
196 Previously the copied amount was the limit, which could result in a
197 overlapping memcpy for newly allocated destination soon after a
198 source string shorter than the limit. Found/investigated by KM.
200 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
201 close; it may be needed for a subsequent connection. This caused a
202 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
204 JH/32 Fix CHUNKING for a second message on a connection when the first was
205 rejected. Previously we did not reset the chunking-offered state, and
206 erroneously rejected the BDAT command. Investigation help from
209 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
210 an empty address. Previously the expansion returned an error.
212 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
213 proxy. Previously these were misparsed, leading to paniclog entries.
219 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
220 related applications. Previously an "H" was used where available info
221 says that "M" should be, so change to match.
223 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
224 as arguments, so an implementation trying to copy these into a local
225 buffer was taking a taint-enforcement trap. Fix by using dynamically
226 created buffers. Similar fix for radius expansion condition.
228 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
229 reasonable, eg. to count headers. Fix by using dynamically created
230 buffers rather than a local. Do similar fixes for ACL actions "dcc",
231 "log_reject_target", "malware" and "spam"; the arguments are expanded
232 so could be handling tainted values.
234 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
235 broken the (no-op) support for this sendmail command. Restore it
236 to doing nothing, silently, and returning good status.
238 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
239 record path was given (or the default used) without a leading directory
240 path, an error occurred on trying to open it. Use the transport's working
243 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
244 transport. Previously it was the name on the DNS A-record; use instead
245 the head of the CNAME chain leading there (if there is one). This seems
246 to align better with RFC 6125.
248 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
249 smtp_accept_max_per_host allocated resources which were not released
250 when the limit was exceeded. This eventually crashed the daemon. Fix
251 by adding a release action in that path.
253 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
254 expanded; previously using tainted values was rejected. Fix by using
255 dynamically-created buffers.
257 JH/09 Relax restrictions on ACL verify condition needing access to message
258 headers. Previously they were only permitted in data and non-smtp ACLs;
259 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
260 not_blind, header_sender and header_names_ascii verification.
262 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
263 Previously a macro used one argument twice; when called with the
264 argument as an expression having side-effects, incorrect operation
265 resulted. Use an inlineable function.
267 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
268 held open for a verify callout. Previously this wan not accounted for
269 and a corrupt onward SMTP conversation resulted.
271 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
272 Previously they were embedded naively in the constructed address; when
273 needed, strip the quoting and quote the entire local_part.
274 Also make the inbound_srs expansion condition handle quoting.
276 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
277 excluded, not matching the documentation.
279 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
280 was given for the sqlite_dbfile a trap resulted.
282 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
283 "name" argument resulted in a trap. There is no reason to disallow such;
284 this was a coding error.
286 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
287 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
288 spent suspended, ignoring the POSIX definition. Previously we assumed
289 it did and a constant offset from real time could be used as a correction.
290 Change to using the same clock source for the start-of-message and the
291 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
292 exists, just to get a clock slightly more aligned to reality.
294 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
295 RFC says it is optional some validators care. The missing char was not
296 intended but triggered by a line-wrap alignment. Discovery and fix by
297 Guillaume Outters, hacked on by JH.
299 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
300 name being quoted was tainted a trap would be taken. Fix by using
301 dynamically created buffers. The routine could have been called by a
302 rewrite with the "h" flag, by using the "-F" command-line option, or
303 by using a "name=" option on a control=submission ACL modifier.
305 JH/19 SPF: change the Authentication-Results expansion component to give
306 smtp.helo when the sender domain is empty. Previously it gave
309 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
310 not in 127.0.0.0/8 to help in spotting list domains taken over by a
311 domain-parking registrar.
313 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
314 Previously when a whitespace character was specified it was not inserted
315 after removing the newline.
317 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
318 the domain part of the recipient address. This overrides any tls_sni
319 option set, which was previously used.
321 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
324 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
325 is_tainted() had an off-by-one error in the overenthusiastic direction.
326 Find and fix by Gavan. Although NetBSD is not a supported platform for
327 4.94 this bug could affect other platforms.
329 PP/01 Fix default prime selection to be consistent.
330 One path used ike23 still, instead of exim.dev.20160529.3; now both
331 execution flows will use the same DH primes (currently
332 exim.dev.20160529.3).
334 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
335 option in smtp transport, to match the documentation. Previously
336 verification was not being done in some cases where it should have been.
338 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
339 than one server was defined and depending on the platform memory layout
340 details, an internal consistency trap could be hit while walking the list
343 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
344 files. The value is used by the authresults expansion item. Previously
345 if this was used in a router or transport, a crash could result.
347 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
348 have a sufficient incidence of EINTR returns from select that an
349 interaction with other operations done by the main daemon loop exposed
350 a bug in the error-handling. This was benign apart from the log
353 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
354 for consistency with delivered "=>" and failed "**" lines. While we're
355 there, handle PRX and TFO.
357 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
358 applied. This resulted, if any header-line rewrite rules were configured,
359 in a panic-log triggerable by sending a message with a long address in
360 a header. Fix by increasing the arbitrary limit to larger than a single
361 (dewrapped) 5322 header line maximum size.
363 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
364 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
365 RFC 6648 which deprecates X- options in protocols as a general practice.
366 Changeover between the implementations is handled by the mechanisms
369 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
370 a remote which uses callout/hold. Previously the local was queued.
372 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
375 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
376 log line. The asterisk is now consistently appended to the remote IP
377 (and port, if given), and will also be provided on defer and fail log
378 lines. Previously it could be placed on the local IP if that was being
379 logged, and was only provided on delivery lines.
381 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
383 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
384 authenticator client_send option. Previously the next char, after a pair
385 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
386 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
387 documentation. There is still no way to get a leading ^ immediately
388 after a NUL (ie. for the password of a PLAIN method authenticator.
390 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
391 files. For bad sizes read, delete the record and whine to paniclog.
393 JH/38 When logging an AUTH failure, as server, do not include sensitive
394 information. Previously, the credentials would be included if given
395 as part of the AUTH command line and an ACL denied authentication.
397 JH/39 Bug 2691: fix $local_part_data. When the matching list element
398 referred to a file, bad data was returned. This likely also affected
401 JH/40 The gsasl authenticator now supports caching of the salted password
402 generated by the client-side implementation. This required the addition
403 of a new variable: $auth4.
405 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
406 left undeleted; the attempt to re-create it then failed - resulting in
407 the usual "SIGHUP tp have daemon reload configuration" to not work.
408 This affected any platform not supporting "abstract" Unix-domain
409 sockets (i.e. not Linux).
411 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
412 recipients" response to RCPT in a later response, with a 250. The
413 previous coding assumed this would not happen, and under PIPELINING
414 would result in both lost and duplicate recipients for a message.
416 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
417 Previously the weighting was incorrectly applied. Similar fix for socks
418 proxies. Found and fixed by Heiko Schlichting.
420 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
421 not handle sub-lists included using the +namedlist syntax. While
422 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
423 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
426 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
427 Previously the permanent pool was used, so the sore could not be freed.
428 This meant a connection with many messages would use continually-growing
431 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
432 per-pool so as not to waste too much space. Previously a constant size
433 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
434 DOS attacks harder. The cost is wasted memory use in the larger blocks.
436 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
437 same space repeatedly, and a smaller process footprint.
439 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
440 for GnuTLS. Previously the name started "%s" which, while not a bug,
441 looked as if if might be one.
443 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
444 "continued connection") make the $proxy_* variables available. Previously
445 the information was not passed across the exec() call for subsequent
446 transport executions. This also mean that the log lines for the
447 messages can show the proxy information.
449 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
450 receive time. With modern systems the difference is significant.
451 The historical behaviour can be restored by disabling (a new) log_selector
452 "queue_time_exclusive".
454 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
455 content) were specifically tested for. Now, also cover expansions
456 resulting in ACL names and inline ACL content.
458 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
459 making it unusable in complex expressions.
461 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
462 fail with a taint-check on the spoolfile name, and leave the message
465 HS/01 Enforce absolute PID file path name.
467 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
469 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
471 PP/02 Bug 2643: Correct TLS DH constants.
472 A missing NUL termination in our code-generation tool had led to some
473 incorrect Diffie-Hellman constants in the Exim source.
474 Reported by kylon94, code-gen tool fix by Simon Arlott.
476 PP/03 Impose security length checks on various command-line options.
477 Fixes CVE-2020-SPRSS reported by Qualys.
479 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
480 better. Reported by Qualys.
482 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
483 providing a particularly obnoxious sender full name.
486 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
488 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
491 PP/08 Change default for recipients_max from unlimited to 50,000.
493 PP/09 Fix security issue with too many recipients on a message (to remove a
494 known security problem if someone does set recipients_max to unlimited,
495 or if local additions add to the recipient list).
496 Fixes CVE-2020-RCPTL reported by Qualys.
498 PP/10 Fix security issue in SMTP verb option parsing
499 Fixes CVE-2020-EXOPT reported by Qualys.
501 PP/11 Fix security issue in BDAT state confusion.
502 Ensure we reset known-good where we know we need to not be reading BDAT
503 data, as a general case fix, and move the places where we switch to BDAT
504 mode until after various protocol state checks.
505 Fixes CVE-2020-BDATA reported by Qualys.
507 HS/03 Die on "/../" in msglog file names
509 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
510 the Exim runtime user are allowed to create files.
512 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
515 QS/03 When reading the output from interpreted forward files we do not
516 pass the pipe between the parent and the interpreting process to
517 executed child processes (if any).
519 QS/04 Always die if requested from internal logging, even is logging is
522 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
523 the API; compilation noo longer completes with DMARC support included.
524 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
525 been reported on other platforms.
527 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
529 JH/56 Make the majority of info read from config files readonly, for defence-in-
530 depth against exploits. Suggestion by Qualys.
531 Not supported on Solaris 10.
533 JH/57 Fix control=fakreject for a custom message containing tainted data.
534 Previously this resulted in a log complaint, due to a re-expansion present
535 since fakereject was originally introduced.
537 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
538 modifier was given, a loop resulted.
540 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
541 pipelined SMTP command followed the BDAT LAST then it would be
542 incorrectly treated as part of the message body, causing a verification
545 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
546 require looser validation rules than those for 821-level addresses,
547 which only permit IP addresses.
553 JH/01 Avoid costly startup code when not strictly needed. This reduces time
554 for some exim process initialisations. It does mean that the logging
555 of TLS configuration problems is only done for the daemon startup.
557 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
559 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
560 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
563 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
564 DKIM signing is being done. Previously a transport_filter would always
565 disable CHUNKING, falling back to traditional DATA.
567 JH/05 Regard command-line recipients as tainted.
569 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
571 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
572 PAM library frees one of the arguments given to it, despite the
573 documentation. Therefore a plain malloc must be used.
575 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
576 on-stack buffers were used, resulting in a taint trap when DSN information
577 copied from a received message was written into the buffer.
579 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
580 the ordering of its ARC headers. This caused a crash.
582 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
583 a new record was being constructed with information from the peer, a trap
586 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
587 installation would get error messages from DMARC verify, when it hit the
588 nonexistent file indicated by the default. Distros wanting DMARC enabled
589 should both provide the file and set the option.
590 Also enforce no DMARC verification for command-line sourced messages.
592 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
593 could, depending on the platform, hang at the STARTTLS response.
595 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
596 message on a connection. Previously if one message had ARC headers and
597 the following one did not, a crash could result when adding an
598 Authentication-Results: header.
600 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
601 Exim main code and Exim-related utities. The introduction of taint
602 tracking also did many adjustments to string handling. Since then, eximon
603 frequently terminated with an assert failure.
605 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
606 check for 452 responses. This slightly helps the inefficieny of doing
607 a large alias-expansion into a recipient-limited target. The max_rcpt
608 transport option still applies (and at the current default, will override
609 the new feature). The check is done for either cause of synch, and forces
610 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
611 connection. The new facility is not tunable at this time.
613 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
614 library live data was being used, so the results became garbage. Make
615 copies while it is still usable.
617 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
618 on delivery deferred (==) and failed (**) lines (if a delivery was
619 attemtped). Previously it was only on completion (=>) lines.
621 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
622 for the expansion of the server_scram_iter and server_scram_salt options.
624 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
625 are now specifically given a NO_DATA response without hitting the system
626 resolver. The library goes on to do the now-standard TXT lookup.
627 Use of dnsdb lookups is not affected.
629 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
630 only retrieve the errormessage once. Previously two calls to dlerror()
631 were used, and the second one (for mainlog/paniclog) retrieved null
634 JH/20 Taint checking: disallow use of tainted data for
635 - the appendfile transport file and directory options
636 - the pipe transport command
637 - the autoreply transport file, log and once options
638 - file names used by the redirect router (including filter files)
640 - paths used by single-key lookups
641 Previously this was permitted.
643 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
644 adjusted the size of a major service buffer; this failed because the
645 buffer was in use at the time. Change to a compile-time increase in the
646 buffer size, when this authenticator is compiled into exim.
648 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
649 previous fast-mode was untenable in the face of glibs using mmap to
650 support larger malloc requests.
652 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
653 New values supported, if defined on system where compiled:
654 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
655 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
657 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
658 running a limited number of proceses in parallel, a benefit is gained. The
659 amount varies with the platform hardware and load. The use of the option
660 queue_run_in_order means we cannot do this, as ordering becomes
663 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
664 had introduced a string-copy (for ensuring NUL-termination) which was not
665 appropriate for that case, which can include embedded NUL bytes in the
666 block of data. Investigation showed the copy to actually be needless, the
667 data being length-specified.
669 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
670 done during a receiving connection, and both used TLS, global info was
671 used rather than per-connection info for tracking the state of data
672 queued for transmission. This could result in a connection hang.
674 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
675 Previously, when delivering serveral messages down a single connection
676 only the first would provide a SIZE. This was due to the size information
677 not being properly tracked.
679 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
680 TAI (at 37 seconds currently), pretend to be in UTC for time-related
681 expansion and logging. Previously, spurious values such as a future
682 minute could be seen.
684 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
685 it could crash from a null-deref. This could also affect the
686 ${addresses: } operator and ${readsock } item.
688 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
689 message following a mime one, the variable was not reset.
691 JH/30 When an pipelined-connect fails at the first response, assume incorrect
692 cached capability (perhaps the peer reneged?) and immediately retry in
695 JH/31 Fix spurious detection of timeout while writing to transport filter.
697 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
698 an attempt to copy the string was made before checking it.
700 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
701 taint of the lookup key was maintained; we now regard the presence in the
702 filesystem as sufficient validation.
704 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
705 argument is supplied.
707 JH/35 The dsearch lookup now requires that the directory is an absolute path.
708 Previously this was not checked, and nonempty relative paths made an
709 access under Exim's current working directory.
711 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
712 Previously no event was raised.
714 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
715 parameter supplied by the sender MAIL FROM command. Previously it was
716 ignored, and only the check_spool_space option value for the required
719 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
720 the size of the signing public-key. Previously it was instead giving
721 the size of the signature hash.
723 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
724 the default. See the (new) dkim_verify_min_keysizes option.
726 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
727 and an ACL use a lookup for checking either the local_part or domain,
728 stale data could be accessed. Ensure that variable references are
729 dropped between messages.
731 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
732 by the client was not checked as pointing within response data before
733 being used. A malicious client could thus cause an out-of-bounds read and
734 possibly gain authentication. Fix by adding the check.
736 JH/42 Internationalisation: change the default for downconversion in the smtp
737 transport to be "if needed". Previously it was "as previously set" for
738 the message, which usually meant "if needed" for message-submission but
739 "no" for everything else. However, MTAs have been seen using SMTPUTF8
740 even when the envelope addresses did not need it, resulting in forwarding
741 failures to non-supporting MTAs. A downconvert in such cases will be
742 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
743 transport. The change does mean that addresses needing conversion will
744 be converted when previously a delivery failure would occur.
746 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
747 response was received it would be used unchecked in a fail-DSN, violating
748 standards on line-length limits. Truncate if needed.
750 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
751 doesn't work. (Additionally add a new main config option to configure the
758 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
759 side, to decode a TLS 1.3 packet capture.
761 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
762 Previously the default library behaviour applied, sending two, each in
765 JH/03 Debug output for ACL now gives the config file name and line number for
768 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
770 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
772 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
773 buffer overrun for (non-chunking) other transports.
775 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
776 TLS1.3, means that a server rejecting a client certificate is not visible
777 to the client until the first read of encrypted data (typically the
778 response to EHLO). Add detection for that case and treat it as a failed
779 TLS connection attempt, so that the normal retry-in-clear can work (if
780 suitably configured).
782 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
783 and/or domain. Found and fixed by Jason Betts.
785 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
786 configuration). If a CNAME target was not a wellformed name pattern, a
789 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
790 the OS reports them interleaved with other addresses.
792 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
793 used both for input and for a verify callout, both encrypted, SMTP
794 responses being sent by the server could be lost. This resulted in
795 dropped connections and sometimes bounces generated by a peer sending
798 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
799 string. Previously it was possible to cause undefined behaviour in a
800 library routine (usually a crash). Found by "zerons".
802 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
805 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
806 API was removed, so update to use the newer ones.
808 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
809 any timeout set, is taking a long time. Previously we would hang on to a
810 rotated logfile "forever" if the input was arriving with long gaps
811 (a previous attempt to fix addressed lack, for a long time, of initial
814 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
815 shared (NFS) environment. The length of the tempfile name is now
816 4 + 16 ("hdr.$message_exim_id") which might break on file
817 systems which restrict the file name length to lower values.
820 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
821 shared (NFS) environment.
823 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
824 did for all versions <4.90). Notably -M, -m, --invert, -I may be
827 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
828 on some platforms for bit 31.
830 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
831 to changes apparently associated with TLS1.3 handling some of the APIs
832 previously used were either nonfunctional or inappropriate. Strings
833 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
834 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
835 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
836 This affects log line X= elements, the $tls_{in,out}_cipher variables,
837 and the use of specific cipher names in the encrypted= ACL condition.
839 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
841 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
842 verification result was not updated unless hosts_require_ocsp applied.
844 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
845 queue_list_requires_admin set to false, non-admin users were denied the
848 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
849 directory-of-certs mode. Previously they were advertised despite the
852 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
853 A single TCP connection by a client will now hold a TLS connection open
854 for multiple message deliveries, by default. Previously the default was to
857 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
858 default. If built with the facility, DANE will be used. The facility
859 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
861 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
862 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
863 must be defined and you must still, unless you define DISABLE_TLS, manage
864 the the include-dir and library-file requirements that go with that
865 choice. Non-TLS builds are still supported.
867 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
870 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
871 default. If the platform supports and has the facility enabled, it will
872 be requested on all coneections.
874 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
875 controlled by the build-time option SUPPORT_PIPE_CONNECT.
877 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
879 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
880 success-DSN messages. Previously the From: header was always the default
881 one for these; the option was ignored.
883 JH/28 Fix the timeout on smtp response to apply to the whole response.
884 Previously it was reset for every read, so a teergrubing peer sending
885 single bytes within the time limit could extend the connection for a
886 long time. Credit to Qualsys Security Advisory Team for the discovery.
888 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
889 delivery address, which leaked information of the results of local
890 forwarding. Change to the original envelope recipient address, per
893 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
894 requested. Previously not bounce was generated and a log entry of
895 error ignored was made.
897 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
899 JH/32 Introduce a general tainting mechanism for values read from the input
900 channel, and values derived from them. Refuse to expand any tainted
901 values, to catch one form of exploit.
903 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
904 was unused and the unexpanded text used for the test. Found and
905 fixed by Ruben Jenster.
907 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
908 an attempt to use a TLS library read routine dereffed a nul pointer,
911 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
912 them in our smtp response.
914 JH/36 Have the general router option retry_use_local_part default to true when
915 any of the restrictive preconditions are set (to anything). Previously it
916 was only for check_local user. The change removes one item of manual
917 configuration which is required for proper retries when a remote router
918 handles a subset of addresses for a domain.
920 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
921 link count into consideration.
923 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
924 caused the extension of big_buffer, the following lines were ignored.
926 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
927 accordance with RFC 2308. Previously there was no expiry, so a longlived
928 receive process (eg. due to ACL delays) versus a short SOA value could
931 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
933 JH/39 Promote DMARC support to mainline.
935 JH/40 Bug 2452: Add a References: header to DSNs.
937 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
938 parameters. The relevant library call is documented as "Deprecated: This
939 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
940 3.6.0, DH parameters are negotiated following RFC7919."
942 HS/06 Change the default of dnssec_request_domains to "*"
944 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
945 carried on and emitted a BDAT command, even when PIPELINING was not
948 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
949 buffer was used for the filename, resulting in a trap when tainted
950 arguments (eg. $domain) were used.
952 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
953 recommended to avoid a possible server-load attack. The feature can be
954 re-enabled via the openssl_options main cofiguration option.
956 JH/45 local_scan API: documented the current smtp_printf() call. This changed
957 for version 4.90 - adding a "more data" boolean to the arguments.
958 Bumped the ABI version number also, this having been missed previously;
959 release versions 4.90 to 4.92.3 inclusive were effectively broken in
960 respect of usage of smtp_printf() by either local_scan code or libraries
961 accessed via the ${dlfunc } expansion item. Both will need coding
962 adjustment for any calls to smtp_printf() to match the new function
963 signature; a FALSE value for the new argument is always safe.
965 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
966 the file-offset (which the Linux syscall does, and exim expects); this
967 resulted in an indefinite loop.
969 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
970 to do ARC verification. The Authentication-Results: header line added
971 by the configuration then had no ARC item.
973 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
974 an incorrect mode was used for reading the result, resulting in it being
981 JH/01 Remove code calling the customisable local_scan function, unless a new
982 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
984 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
985 non-signal-safe functions being used.
987 JH/03 Bug 2269: When presented with a received message having a stupidly large
988 number of DKIM-Signature headers, disable DKIM verification to avoid
989 a resource-consumption attack. The limit is set at twenty.
991 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
992 report of oldest_pass in ${authres } in consequence, and separate out
993 some descriptions of reasons for verification fail.
995 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
996 files in the spool were present and unlocked. A queue-runner could spot
997 them, resulting in a duplicate delivery. Fix that by doing the unlock
998 after the unlink. Investigation by Tim Stewart. Take the opportunity to
999 add more error-checking on spoolfile handling while that code is being
1002 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1003 No known attacks, no CVE, this is defensive hardening.
1005 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1006 a queue-runner could start a delivery while other operations were ongoing.
1007 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1008 Found and investigated by Tim Stewart. Fix by using the open message data
1009 file handle rather than opening another, and not locally closing it (which
1010 releases a lock) for that case, while creating the temporary .eml format
1011 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1013 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1014 $sender_verify_failure/$recipient_verify_failure to "random".
1016 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1019 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1020 Previously this would segfault.
1022 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1025 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1026 like zero, since the resolver should be doing this for us, But we need one
1027 as a CNAME but no MX presence gets the CNAME returned; we need to check
1028 that doesn't point to an MX to declare it "no MX returned" rather than
1029 "error, loop". A new main option is added so the older capability of
1030 following some limited number of chain links is maintained.
1032 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1034 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1035 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1036 now-deprecated earlier definitions used only specified the range up to TLS
1037 1.2 (in the older-version library docs).
1039 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1041 JH/15 Rework TLS client-side context management. Stop using a global, and
1042 explicitly pass a context around. This enables future use of TLS for
1043 connections to service-daemons (eg. malware scanning) while a client smtp
1044 connection is using TLS; with cutthrough connections this is quite likely.
1046 JH/16 Fix ARC verification to do AS checks in reverse order.
1048 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1050 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1051 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1052 Previously the "utf8" would be re-prepended for every additional message.
1054 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1055 Previously thery were accepted, resulting in issues when attempting to
1056 forward messages to a non-supporting MTA.
1058 PP/02 Let -n work with printing macros too, not just options.
1060 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1061 one parent address was copied, and bogus data was used at delivery-logging
1062 time. Either a crash (after delivery) or bogus log data could result.
1063 Discovery and analysis by Tim Stewart.
1065 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1066 Previously if the string ended mid-character, we did not insert the
1067 promised '?' replacement.
1069 PP/04 Documentation: current string operators work on bytes, not codepoints.
1071 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1072 should pack well giving a smaller memory footprint so better caching and
1073 therefore performance. Group the declarations where this can't be done so
1074 that the byte-sized flag variables are not interspersed among pointer
1075 variables, giving a better chance of good packing by the compiler.
1077 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1078 non-null, to avoid issues with sites running BATV. Previously reports were
1079 sent with an empty envelope sender so looked like bounces.
1081 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1082 The ignore_error flag wasn't being returned from the filter subprocess so
1083 was not set for later routers. Investigation and fix by Matthias Kurz.
1085 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1086 and a msg:complete for the whole, when a message is manually removed using
1087 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1089 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1090 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1091 hope that is portable enough.
1093 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1094 requiring a known-CA anchor certificate; make it now rely entirely on the
1095 TLSA as an anchor. Checking the name on the leaf cert against the name
1096 on the A-record for the host is still done for TA (but not for EE mode).
1098 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1099 included in delivery lines for non-proxied connections, when compiled with
1100 SUPPORT_SOCKS and running with proxy logging enabled.
1102 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1103 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1104 move the existing event to fire before the normal logging of message
1105 failure so that custom logging is bracketed by normal logging.
1107 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1108 msg:fail:internal event. Developement by Matthias Kurz.
1110 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1111 far too small for todays use of crypto signatures stored there. Go all
1112 the way to the max DNS message size of 64kB, even though this might be
1113 overmuch for IOT constrained device use.
1115 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1116 copy a string over itself. The library routine is documented as not
1117 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1119 JH/32 For main options check_spool_space and check_inode_space, where the
1120 platform supports 64b integers, support more than the previous 2^31 kB
1121 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1122 the previous G, M, k.
1124 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1125 $authenticated_fail_id variable on authentication failure. Previously
1128 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1129 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1130 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1131 GNUTLS_SEC_PARAM_MEDIUM.
1133 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1134 side. Previously we would continue as if no SNI had been received.
1136 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1137 "<" character, which should be a list-separator specification, we walked
1138 off past the nul-terimation.
1140 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1141 causes) even when the retry time is not yet met. Previously they were
1142 not, meaning that when (say) an account was over-quota and temp-rejecting,
1143 and multiple senders' messages were queued, only one sender would get
1144 notified on each configured delay_warning cycle.
1146 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1148 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1149 cipher-suites, an error can be left on the stack even for a succeeding
1150 accept; this results in impossible error messages when a later operation
1153 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1154 return error codes indicating retry. Under TLS1.3 this becomes required.
1156 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1157 it only wrote the new authenticators, resulting in a lack of tracking of
1158 peer changes of ESMTP extensions until the next cache flush.
1160 JH/41 Fix the loop reading a message header line to check for integer overflow,
1161 and more-often against header_maxsize. Previously a crafted message could
1162 induce a crash of the recive process; now the message is cleanly rejected.
1164 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1165 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1171 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1172 When redis_servers is set to a list of > 1 element, and the Redis servers
1173 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1174 case of MOVED into a DEFER case instead, thus moving the query onto the
1175 next server in the list. For a cluster of N elements, all N servers must
1176 be defined in redis_servers.
1178 GF/02 Catch and remove uninitialized value warning in exiqsumm
1179 Check for existence of @ARGV before looking at $ARGV[0]
1181 JH/01 Replace the store_release() internal interface with store_newblock(),
1182 which internalises the check required to safely use the old one, plus
1183 the allocate and data copy operations duplicated in both (!) of the
1184 extant use locations.
1186 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1187 modifier. This matches the restriction on the commandline.
1189 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1190 Previously only the last row was returned.
1192 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1193 we assumed that tags in the header were well-formed, and parsed the
1194 element content after inspecting only the first char of the tag.
1195 Assumptions at that stage could crash the receive process on malformed
1198 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1199 While running the DKIM ACL we operate on the Permanent memory pool so that
1200 variables created with "set" persist to the DATA ACL. Also (at any time)
1201 DNS lookups that fail create cache records using the Permanent pool. But
1202 expansions release any allocations made on the current pool - so a dnsdb
1203 lookup expansion done in the DKIM ACL releases the memory used for the
1204 DNS negative-cache, and bad things result. Solution is to switch to the
1205 Main pool for expansions.
1206 While we're in that code, add checks on the DNS cache during store_reset,
1207 active in the testsuite.
1208 Problem spotted, and debugging aided, by Wolfgang Breyha.
1210 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1211 When none of the hosts presented to a transport match an already-open
1212 connection, close it and proceed with the list. Previously we would
1213 queue the message. Spotted by Lena with Yahoo, probably involving
1216 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1217 Previously a spurious "250 OK id=" response was appended to the proper
1220 JH/08 The "support for" informational output now, which built with Content
1221 Scanning support, has a line for the malware scanner interfaces compiled
1222 in. Interface can be individually included or not at build time.
1224 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1225 by the template makefile "src/EDITME". The "STREAM" support for an older
1226 ClamAV interface method is removed.
1228 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1229 rows affected is given instead).
1231 JH/11 The runtime Berkeley DB library version is now additionally output by
1232 "exim -d -bV". Previously only the compile-time version was shown.
1234 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1235 SMTP connection. Previously, when one had more recipients than the
1236 first, an abortive onward connection was made. Move to full support for
1237 multiple onward connections in sequence, handling cutthrough connection
1238 for all multi-message initiating connections.
1240 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1241 routers. Previously, a multi-recipient message would fail to match the
1242 onward-connection opened for the first recipient, and cause its closure.
1244 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1245 a timeout on read on a GnuTLS initiating connection, resulting in the
1246 initiating connection being dropped. This mattered most when the callout
1247 was marked defer_ok. Fix to keep the two timeout-detection methods
1250 JH/15 Relax results from ACL control request to enable cutthrough, in
1251 unsupported situations, from error to silently (except under debug)
1252 ignoring. This covers use with PRDR, frozen messages, queue-only and
1255 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1257 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1258 metadata, resulting in a crash in free().
1260 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1261 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1262 Broken also in d185889f4, with init system revamp.
1264 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1265 Previously we abruptly closed the connection after reading a malware-
1266 found indication; now we go on to read the "scan ok" response line,
1269 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1270 ACL. Previously, a crash would result.
1272 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1273 macro text after a replacement (previously it was only once per line) and
1274 by skipping builtin macros when searching for an uppercase lead character.
1276 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1277 for the build is renamed.
1279 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1280 was allocated for every new TLS startup, meaning one per message. Fix
1281 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1283 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1284 reported the original. Fix to report (as far as possible) the ACL
1285 result replacing the original.
1287 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1288 OpenSSL. Certificate information is loaded for every new TLS startup,
1289 and the resources needed to be freed.
1291 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1293 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1296 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1297 DATA response info to the (existing) per-recipient response info for
1298 the "C=" log element. It can have useful tracking info from the
1299 destination system. Patch from Simon Arlott.
1301 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1302 length value. Previously this would segfault.
1304 HS/02 Support Avast multiline protoocol, this allows passing flags to
1305 newer versions of the scanner.
1307 JH/28 Ensure that variables possibly set during message acceptance are marked
1308 dead before release of memory in the daemon loop. This stops complaints
1309 about them when the debug_store option is enabled. Discovered specifically
1310 for sender_rate_period, but applies to a whole set of variables.
1311 Do the same for the queue-runner and queue-list loops, for variables set
1312 from spool message files. Do the same for the SMTP per-message loop, for
1313 certain variables indirectly set in ACL operations.
1315 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1316 as a multi-recipient message from a mailinglist manager). The coding had
1317 an arbitrary cutoff number of characters while checking for more input;
1318 enforced by writing a NUL into the buffer. This corrupted long / fast
1319 input. The problem was exposed more widely when more pipelineing of SMTP
1320 responses was introduced, and one Exim system was feeding another.
1321 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1322 receiving system, and refused recipients seen by the sending system
1323 (propating to people being dropped from mailing lists).
1324 Discovered and pinpointed by David Carter.
1326 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1327 replaced by the ${authresults } expansion.
1329 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1331 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1332 allows proper process termination in container environments.
1334 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1335 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1337 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1338 and "err_perm", deprecated since 4.83 when the RFC-defined words
1339 "temperror" and "permerror" were introduced.
1341 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1342 transport-filters or DKIM-signing. The restriction was lost in the
1343 consolidation of verify-callout and delivery SMTP handling.
1344 Extend the restriction to also cover ARC-signing.
1346 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1347 in defer=pass mode supply a 450 to the initiator. Previously the message
1350 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1351 tls_require_ciphers is used as before.
1353 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1354 "pass_unscanned". Only tmpfails from the scanner are written to
1355 the paniclog, as they may require admin intervention (permission
1356 denied, license issues). Other scanner errors (like decompression
1357 bombs) do not cause a paniclog entry.
1359 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1360 Previously it was possible to log spurious information in receive log
1363 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1364 triggered odd behaviour from Outlook Express clients.
1366 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1369 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1370 since the IETF WG has not yet settled on that versus the original
1371 "bare" representation.
1373 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1374 Previously the millisecond value corrupted the output.
1375 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1376 corrupted the output.
1382 JH/01 Rework error string handling in TLS interface so that the caller in
1383 more cases is responsible for logging. This permits library-sourced
1384 string to be attached to addresses during delivery, and collapses
1385 pairs of long lines into single ones.
1387 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1388 during configuration. Wildcards are allowed and expanded.
1390 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1391 This permits better logging.
1393 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1394 do not close it down and have the child transport start it up again on
1395 the passed-on TCP connection. Instead, proxy the child (and any
1396 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1397 affected: the continued delivery log lines do not have any DNSSEC, TLS
1398 Certificate or OCSP information. TLS cipher information is still logged.
1400 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1401 identical IP addresses on different listening ports. Will also affect
1404 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1405 add noisy ifdef guards to special-case this sillyness.
1406 Patch from Bernd Kuhls.
1408 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1409 than 255 are no longer allowed.
1411 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1412 Disable the session-cache too, which might reduce our load. Since we
1413 currrectly use a new context for every connection, both as server and
1414 client, there is no benefit for these.
1415 GnuTLS appears to not support tickets server-side by default (we don't
1416 call gnutls_session_ticket_enable_server()) but client side is enabled
1417 by default on recent versions (3.1.3 +) unless the PFS priority string
1420 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1421 <https://reproducible-builds.org/specs/source-date-epoch/>.
1423 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1424 the check for any unsuccessful recipients did not notice the limit, and
1425 erroneously found still-pending ones.
1427 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1428 MSG_MORE. Only in-clear (not on TLS connections).
1430 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1431 transport-filter is involved we can buffer the headers in memory for
1432 creating the signature, and read the spool data file once for the
1433 signature and again for transmission.
1435 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1436 4.77 as the kernel support then wasn't solid, having issues in 64bit
1437 mode. Now, it's been long enough. Add support for FreeBSD also.
1439 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1440 case where the routing stage had gathered several addresses to send to
1441 a host before calling the transport for the first, we previously failed
1442 to close down TLS in the old transport process before passing the TCP
1443 connection to the new process. The new one sent a STARTTLS command
1444 which naturally failed, giving a failed delivery and bloating the retry
1445 database. Investigation and fix prototype from Wolfgang Breyha.
1447 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1448 false-negatives in the check that the sender had not preempted a response
1449 or prompt from Exim (running as a server), due to that code's lack of
1450 awareness of the SMTP input buffering.
1452 PP/04 Add commandline_checks_require_admin option.
1453 Exim drops privileges sanely, various checks such as -be aren't a
1454 security problem, as long as you trust local users with access to their
1455 own account. When invoked by services which pass untrusted data to
1456 Exim, this might be an issue. Set this option in main configuration
1457 AND make fixes to the calling application, such as using `--` to stop
1460 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1461 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1462 responses to those, into a single TLS record each way (this usually means
1463 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1466 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1467 affects you only if you're dancing at the edge of the param size limits.
1468 If you are, and this message makes sense to you, then: raise the
1469 configured limit or use OpenSSL 1.1. Nothing we can do for older
1472 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1473 cmdline element to get the documented default one. Previously it was
1476 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1477 are unsuable for non-message-specific reasons - eg. connection timeout,
1478 banner-time rejection.
1480 JH/16 Fix logging of delivery remote port, when specified by router, under
1483 PP/06 Repair manualroute's ability to take options in any order, even if one
1484 is the name of a transport.
1487 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1489 JH/17 Change the list-building routines interface to use the expanding-string
1490 triplet model, for better allocation and copying behaviour.
1492 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1493 Previously it was constructed the first time a possibly-matching string
1494 was met in the configuration file input during startup; now it is done
1497 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1498 compatible one, to avoid the (poorly documented) possibility of a config
1499 file in the working directory redirecting the DB files, possibly correpting
1500 some existing file. CVE-2017-10140 assigned for BDB.
1502 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1503 cache-hot. Previously, although the result was properly cached, the
1504 initial verify call returned a defer.
1506 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1507 the main verify for receipient in uncached-mode.
1509 JH/22 Retire historical build files to an "unsupported" subdir. These are
1510 defined as "ones for which we have no current evidence of testing".
1512 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1513 if present. Previously it was ignored.
1515 JH/24 Start using specified-initialisers in C structure init coding. This is
1516 a C99 feature (it's 2017, so now considered safe).
1518 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1519 if was a fixed-sized field and bitmask ops via macros; it is now more
1522 PP/07 GitHub PR 56: Apply MariaDB build fix.
1523 Patch provided by Jaroslav Škarvada.
1525 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1526 during Coverity cleanups [4.87 JH/47]
1527 Diagnosis and fix provided by Michael Fischer v. Mollard.
1529 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1530 the right size to place the terminating semicolon on its own folded
1531 line, the header hash was calculated to an incorrect value thanks to
1532 the (relaxed) space the fold became.
1534 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1535 and confused the parent.
1537 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1538 which could crash as a result. This could lead to undeliverable messages.
1540 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1543 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1544 matters on fast-turnover and PID-randomising systems, which were getting
1545 out-of-order delivery.
1547 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1548 a possibly-overlapping copy. The symptom was that "Remote host closed
1549 connection in response to HELO" was logged instead of the actual 4xx
1552 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1553 Previously only that bufferd was discarded, resulting in SYMTP command
1556 JH/32 DKIM: when a message has multiple signatures matching an identity given
1557 in dkim_verify_signers, run the dkim acl once for each. Previously only
1558 one run was done. Bug 2189.
1560 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1561 "panic the current process" to "deliberately defer". The panic log is
1562 still written with the problem list name; the mail and reject logs now
1563 get a temp-reject line for the message that was being handled, saying
1564 something like "domains check lookup or other defer". The SMTP 451
1565 message is still "Temporary local problem".
1567 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1568 A crafted sequence of BDAT commands could result in in-use memory beeing
1569 freed. CVE-2017-16943.
1571 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1572 from SMTP input. Previously it was always done; now only done for DATA
1573 and not BDAT commands. CVE-2017-16944.
1575 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1576 to the message (such as an overlong header line). Previously this was
1577 not done and we did not exit BDAT mode. Followon from the previous item
1578 though a different problem.
1584 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1585 than -2003 did; needs libidn2 in addition to libidn.
1587 JH/02 The path option on a pipe transport is now expanded before use.
1589 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1590 Patch provided by "Björn", documentation fix added too.
1592 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1593 missing a wire-to-host endian conversion.
1595 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1596 close after a BDAT command line could be taken as a following command,
1597 giving a synch failure. Fix by only checking for synch immediately
1598 before acknowledging the chunk.
1600 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1601 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1602 macro. Patches provided by Josh Soref.
1604 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1605 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1606 by IANA as a service extension. However, John Klensin suggests that we
1609 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1610 the dkim code may be unix-mode line endings rather than smtp wire-format
1611 CRLF, so prepend a CR to any bare LF.
1613 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1614 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1616 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1617 the first were themselves being wrongly included in the feed into dkim
1618 processing; with most chunk sizes in use this resulted in an incorrect
1619 body hash calculated value.
1621 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1622 DKIM signature block, for verification. Although advised against by
1623 standards it is specifically not ruled illegal.
1625 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1627 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1628 missing a body hash (the bh= tag).
1630 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1631 It seems that HAProxy sends the Proxy Protocol information in clear and
1632 only then does a TLS startup, so do the same.
1634 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1635 TCP connections (such as for Spamd) unless the daemon successfully set
1636 Fast Open mode on its listening sockets. This fixes breakage seen on
1637 too-old kernels or those not configured for Fast Open, at the cost of
1638 requiring both directions being enabled for TFO, and TFO never being used
1639 by non-daemon-related Exim processes.
1641 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1642 endings, at least on the first header line. Try to canonify any that get
1643 past that check, despite the cost.
1645 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1646 now limited to an arbitrary five deep, while parsing addresses with the
1647 strip_excess_angle_brackets option enabled.
1649 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1650 instead leave the unprompted TLS handshake in socket buffer for the
1651 TLS library to consume.
1653 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1655 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1657 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1658 region in one operation, for speed, and this leaves assigned pointers
1659 dangling. Add checks run only under the testsuite which checks all
1660 variables at a store-reset and panics on a dangling pointer; add code
1661 explicitly nulling out all the variables discovered. Fixes one known
1662 bug: a transport crash, where a dangling pointer for $sending_ip_address
1663 originally assigned in a verify callout, is re-used.
1665 PP/06 Drop '.' from @INC in various Perl scripts.
1667 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1669 PP/08 Reduce a number of compilation warnings under clang; building with
1670 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1671 should be warning-free.
1673 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1675 HS/01 Fix portability problems introduced by PP/08 for platforms where
1676 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1678 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1679 chunk. This allows us to accept broken chunked messages. We need a more
1680 general solution here.
1682 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1683 already-broken messages in the queue.
1685 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1687 JH/19 Fix reference counting bug in routing-generated-address tracking.
1693 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1694 supports it and a size is available (ie. the sending peer gave us one).
1696 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1697 years of being deprecated). The replacements are the ACLs
1698 acl_smtp_mime and acl_not_smtp_mime.
1700 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1701 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1702 or even an in-clear connection were permitted. Now, if the host lookup
1703 was dnssec and dane was requested then the host is only used if the
1704 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1705 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1706 if one fails this test.
1707 This means that a poorly-configured remote DNS will make it incommunicado;
1708 but it protects against a DNS-interception attack on it.
1710 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1711 non-noisy when a race steals the message being considered.
1713 JH/05 If main configuration option tls_certificate is unset, generate a
1714 self-signed certificate for inbound TLS connections.
1716 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1717 in rewrites and routers.
1719 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1720 and logged a warning sing 4.83; now they are a configuration file error.
1722 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1723 (lacking @domain). Apply the same qualification processing as RCPT.
1725 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1727 JH/10 Support ${sha256:} applied to a string (as well as the previous
1730 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1731 a cutthrough deliver is pending, as we always want to make a connection.
1732 This also avoids re-routing the message when later placing the cutthrough
1733 connection after a verify cache hit.
1734 Do not update it with the verify result either.
1736 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1737 when routing results in more than one destination address.
1739 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1740 signing (which inhibits the cutthrough capability). Previously only
1741 the presence of an option was tested; now an expansion evaluating as
1742 empty is permissible (obviously it should depend only on data available
1743 when the cutthrough connection is made).
1745 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1746 the relevant preceding SMTP command did not note the pipelining mode.
1748 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1749 Previously they were not counted.
1751 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1752 as one having no matching records. Previously we deferred the message
1753 that needed the lookup.
1755 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1756 distinguished as "(=".
1758 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1759 for missing MX records. Previously it only worked for missing A records.
1761 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1763 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1764 after the data-go-ahead and data-ack. Patch from Jason Betts.
1766 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1767 even for a "none" policy. Patch from Tony Meyer.
1769 JH/22 Fix continued use of a connection for further deliveries. If a port was
1770 specified by a router, it must also match for the delivery to be
1773 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1774 When one of the recipients of a message was unsuitable for the connection
1775 (has no matching addresses), we lost track of needing to mark it
1776 deferred. As a result mail would be lost.
1778 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1780 JH/25 Decoding ACL controls is now done using a binary search; the source code
1781 takes up less space and should be simpler to maintain. Merge the ACL
1782 condition decode tables also, with similar effect.
1784 JH/26 Fix problem with one_time used on a redirect router which returned the
1785 parent address unchanged. A retry would see the parent address marked as
1786 delivered, so not attempt the (identical) child. As a result mail would
1789 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1790 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1791 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1794 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1795 Main config options check_{log,spool}_{inodes,space} are now
1796 100 inodes, 10MB unless set otherwise in the configuration.
1798 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1799 Previously it only applied to the main-section connection policy
1802 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1804 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1805 by me. Added RFC7919 DH primes as an alternative.
1807 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1808 are not in the system include path.
1810 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1811 GnuTLS, when a session startup failed (eg because the client disconnected)
1812 Exim did stdio operations after fclose. This was exposed by a recent
1813 change which nulled out the file handle after the fclose.
1815 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1816 signed directly by the cert-signing cert, rather than an intermediate
1817 OCSP-signing cert. This is the model used by LetsEncrypt.
1819 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1821 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1822 an incoming connection.
1824 HS/02 Bug 1802: Do not half-close the connection after sending a request
1827 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1828 fallback to "prime256v1".
1830 JH/34 SECURITY: Use proper copy of DATA command in error message.
1831 Could leak key material. Remotely exploitable. CVE-2016-9963.
1837 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1838 and 3.4.4 - once the server is enabled to respond to an OCSP request
1839 it does even when not requested, resulting in a stapling non-aware
1840 client dropping the TLS connection.
1842 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1843 support variable-length bit vectors. No functional change.
1845 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1846 The I= interface field on outgoing lines is now after the H= remote
1847 host field, same as incoming lines. There is a separate
1848 outgoing_interface log selector which allows you to disable the
1851 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1852 If not running log_selector +smtp_connection the mainlog would be held
1853 open indefinitely after a "too many connections" event, including to a
1854 deleted file after a log rotate. Leave the per net connection logging
1855 leaving it open for efficiency as that will be quickly detected by the
1856 check on the next write.
1858 HS/01 Bug 1671: Fix post transport crash.
1859 Processing the wait-<transport> messages could crash the delivery
1860 process if the message IDs didn't exist for some reason. When
1861 using 'split_spool_directory=yes' the construction of the spool
1862 file name failed already, exposing the same netto behaviour.
1864 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1865 mime_regex ACL conditions.
1867 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1868 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1869 to HELO, local diagnostic string.
1871 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1872 log line to debug. Even when configured with a tls authenticator many
1873 client connections are expected to not authenticate in this way, so
1874 an authenticate fail is not an error.
1876 HS/02 Add the Exim version string to the process info. This way exiwhat
1877 gives some more detail about the running daemon.
1879 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1880 matter for fast-change records such as DNSBLs.
1882 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1883 retry record, even if constant. There may be multiple transports with
1884 different interface settings and the retry behaviour needs to be kept
1887 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1889 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1891 JH/10 Bug 840: fix log_defer_output option of pipe transport
1893 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1894 hosts_randomize. This matters a lot when combined with mua_wrapper.
1896 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1897 ${quote_pgsql:<string>} operator.
1899 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1901 JH/14 Change header file name for freeradius-client. Relevant if compiling
1902 with Radius support; from the Gentoo tree and checked under Fedora.
1904 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1906 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1907 option transport_filter, meaning no filtering.
1909 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1911 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1912 defaults to "*" (all hosts). The variable is now available when not built
1913 with TLS, default unset, mainly to enable keeping the testsuite sane.
1914 If a server certificate is not supplied (via tls_certificate) an error is
1915 logged, and clients will find TLS connections fail on startup. Presumably
1916 they will retry in-clear.
1917 Packagers of Exim are strongly encouraged to create a server certificate
1918 at installation time.
1920 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1921 with the $config_file variable.
1923 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1924 in transport context, after the attempt, and per-recipient. The latter type
1925 is per host attempted. The event data is the error message, and the errno
1926 information encodes the lookup type (A vs. MX) used for the (first) host,
1927 and the trailing two digits of the smtp 4xx response.
1929 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1930 to write to mainlog (or rejectlog, paniclog) in the window between file
1931 creation and permissions/ownership being changed. Particularly affects
1932 installations where exicyclog is run as root, rather than exim user;
1933 result is that the running daemon panics and dies.
1935 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1937 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1938 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1939 "pri" and "weight". Note that the previous implicit priority given by the
1940 list order is no longer honoured.
1942 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1943 for DKIM processing.
1945 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1946 by defining SUPPORT_SOCKS.
1948 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1949 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1950 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1951 variables are renamed to proxy_{local,external}_{address,port}.
1953 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1954 for a build by defining SUPPORT_I18N
1956 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1957 of the query string, and make ${quote_redis:} do that quoting.
1959 JH/29 Move Events support from Experimental to mainline, enabled by default
1960 and removable for a build by defining DISABLE_EVENT.
1962 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1964 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1965 cached by the daemon.
1967 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1968 by defining LOOKUP_REDIS. The libhiredis library is required.
1970 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1971 keys are given for lookup.
1973 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1974 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1975 only supported when built with TLS support. The PolarSSL SHA routines
1976 are still used when the TLS library is too old for convenient support.
1978 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1979 openssl_options), for security. OpenSSL forces this from version 1.1.0
1980 server-side so match that on older versions.
1982 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1983 allocation for $value could be released as the expansion processing
1984 concluded, but leaving the global pointer active for it.
1986 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1987 and to use the domains and local_parts ACL conditions.
1989 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1990 incorrectly not doubled on cutthrough transmission, hence seen as a
1991 body-termination at the receiving system - resulting in truncated mails.
1992 Commonly the sender saw a TCP-level error, and retransmitted the message
1993 via the normal store-and-forward channel. This could result in duplicates
1994 received - but deduplicating mailstores were liable to retain only the
1995 initial truncated version.
1997 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1999 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2001 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2002 we're in there, support oversigning also; bug 1309.
2004 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2006 HS/04 Add support for keep_environment and add_environment options.
2008 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2009 either intentional arithmetic overflow during PRNG, or testing config-
2012 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2013 delivery resulted in actual delivery. Cancel cutthrough before DATA
2016 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2017 rejects a recipient: pass the reject to the originator.
2019 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2020 Many were false-positives and ignorable, but it's worth fixing the
2023 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2024 for the new environment-manipulation done at startup. Move the routines
2025 from being local to tls.c to being global via the os.c file.
2027 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2028 an extract embedded as result-arg for a map, the first arg for extract
2029 is unavailable so we cannot tell if this is a numbered or keyed
2030 extraction. Accept either.
2036 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2039 JH/02 The smtp transport option "multi_domain" is now expanded.
2041 JH/03 The smtp transport now requests PRDR by default, if the server offers
2044 JH/04 Certificate name checking on server certificates, when exim is a client,
2045 is now done by default. The transport option tls_verify_cert_hostnames
2046 can be used to disable this per-host. The build option
2047 EXPERIMENTAL_CERTNAMES is withdrawn.
2049 JH/05 The value of the tls_verify_certificates smtp transport and main options
2050 default to the word "system" to access the system default CA bundle.
2051 For GnuTLS, only version 3.0.20 or later.
2053 JH/06 Verification of the server certificate for a TLS connection is now tried
2054 (but not required) by default. The verification status is now logged by
2055 default, for both outbound TLS and client-certificate supplying inbound
2058 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2061 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2062 Status Notification (bounce) messages are now MIME format per RFC 3464.
2063 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2064 under the control of the dsn_advertise_hosts option, and routers may
2065 have a dsn_lasthop option.
2067 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2068 default, modifiable by a malware= option. The list separator for
2069 the options can now be changed in the usual way. Bug 68.
2071 JH/10 The smtp_receive_timeout main option is now expanded before use.
2073 JH/11 The incoming_interface log option now also enables logging of the
2074 local interface on delivery outgoing connections.
2076 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2077 if the interface and destination host and port all match.
2079 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2082 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2083 Patch from Andrew Lewis.
2085 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2086 now supports optional time-restrictions, weighting, and priority
2087 modifiers per server. Patch originally by <rommer@active.by>.
2089 JH/16 The spamd_address main option now supports a mixed list of local
2090 and remote servers. Remote servers can be IPv6 addresses, and
2091 specify a port-range.
2093 JH/17 Bug 68: The spamd_address main option now supports an optional
2094 timeout value per server.
2096 JH/18 Bug 1581: Router and transport options headers_add/remove can
2097 now have the list separator specified.
2099 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2102 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2105 JH/21 Support for the A6 type of dns record is withdrawn.
2107 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2108 rather than the verbs used.
2110 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2111 from 255 to 1024 chars.
2113 JH/24 Verification callouts now attempt to use TLS by default.
2115 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2116 are generic router options now. The defaults didn't change.
2118 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2119 Original patch from Alexander Shikoff, worked over by JH.
2121 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2122 files if ZCAT_COMMAND is not executable.
2124 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2126 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2128 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2129 Normally benign, it bites when the pair was led to by a CNAME;
2130 modern usage is to not canonicalize the domain to a CNAME target
2131 (and we were inconsistent anyway for A-only vs AAAA+A).
2133 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2135 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2136 when evaluating $sender_host_dnssec.
2138 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2139 $sender_helo_dnssec variable.
2141 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2143 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2145 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2147 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2148 documented as working, but never had. Support all but $spam_report.
2150 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2151 added for tls authenticator.
2153 HS/03 Add perl_taintmode main config option
2159 TL/01 When running the test suite, the README says that variables such as
2160 no_msglog_check are global and can be placed anywhere in a specific
2161 test's script, however it was observed that placement needed to be near
2162 the beginning for it to behave that way. Changed the runtest perl
2163 script to read through the entire script once to detect and set these
2164 variables, reset to the beginning of the script, and then run through
2165 the script parsing/test process like normal.
2167 TL/02 The BSD's have an arc4random API. One of the functions to induce
2168 adding randomness was arc4random_stir(), but it has been removed in
2169 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2170 function when detected.
2172 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2173 cause callback expansion.
2175 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2176 syntax errors in an expansion can be treated as a string instead of
2177 logging or causing an error, due to the internal use of bool_lax
2178 instead of bool when processing it.
2180 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2181 server certificates when making smtp deliveries.
2183 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2185 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2187 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2189 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2190 Merged patch from Sebastian Wiedenroth.
2192 JH/05 Fix results-pipe from transport process. Several recipients, combined
2193 with certificate use, exposed issues where response data items split
2194 over buffer boundaries were not parsed properly. This eventually
2195 resulted in duplicates being sent. This issue only became common enough
2196 to notice due to the introduction of connection certificate information,
2197 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2199 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2200 size buffer was used, resulting in syntax errors when an expansion
2203 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2204 version 3.3.6 or later.
2206 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2207 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2208 both become "event_action", the variables become $event_name, $event_data
2209 and $event_defer_errno. There is a new variable $verify_mode, usable in
2210 routers, transports and related events. The tls:cert event is now also
2211 raised for inbound connections, if the main configuration event_action
2214 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2215 early OCSP support, but no stapling (appears to be less than 1.0.0).
2217 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2218 server certificate names available under the smtp transport option
2219 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2222 JH/10 Time-related extraction expansions from certificates now use the main
2223 option "timezone" setting for output formatting, and are consistent
2224 between OpenSSL and GnuTLS compilations. Bug 1541.
2226 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2227 encoded parameter in the incoming message. Bug 1558.
2229 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2230 include certificate info, eximon was claiming there were spoolfile
2233 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2235 JH/14 Log delivery-related information more consistently, using the sequence
2236 "H=<name> [<ip>]" wherever possible.
2238 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2239 are problematic for Debian distribution, omit them from the release
2242 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2244 JH/16 Fix string representation of time values on 64bit time_t architectures.
2247 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2248 output list separator was used.
2253 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2254 checkers that were complaining about end of non-void function with no
2257 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2258 This was a regression introduced in 4.83 by another bugfix.
2260 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2262 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2263 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2269 TF/01 Correctly close the server side of TLS when forking for delivery.
2271 When a message was received over SMTP with TLS, Exim failed to clear up
2272 the incoming connection properly after forking off the child process to
2273 deliver the message. In some situations the subsequent outgoing
2274 delivery connection happened to have the same fd number as the incoming
2275 connection previously had. Exim would try to use TLS and fail, logging
2276 a "Bad file descriptor" error.
2278 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2279 utilities have not been installed.
2281 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2282 temporary space as the ACL may create new global variables.
2284 TL/01 LDAP support uses per connection or global context settings, depending
2285 upon the detected version of the libraries at build time.
2287 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2288 to extract and use the src ip:port in logging and expansions as if it
2289 were a direct connection from the outside internet. PPv2 support was
2290 updated based on HAProxy spec change in May 2014.
2292 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2294 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2295 Properly escape header and check for NULL return.
2297 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2300 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2302 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2303 characters in header names, implemented as a verify condition.
2304 Contributed by Michael Fischer v. Mollard.
2306 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2307 results permerror and temperror. Previous values are deprecated but
2308 still accepted. In a future release, err_perm and err_temp will be
2309 completely removed, which will be a backward incompatibility if the
2310 ACL tests for either of these two old results. Patch contributed by
2311 user bes-internal on the mailing list.
2313 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2315 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2316 selectors, in both main and reject logs.
2318 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2321 JH/07 Add malware type "sock" for talking to simple daemon.
2323 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2325 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2326 routers/transports under cutthrough routing.
2328 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2329 numbers. Touch up "bool" conditional to keep the same definition.
2331 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2333 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2335 JH/12 Expand items in router/transport headers_add or headers_remove lists
2336 individually rather than the list as a whole. Bug 1452.
2338 Required for reasonable handling of multiple headers_ options when
2339 they may be empty; requires that headers_remove items with embedded
2340 colons must have them doubled (or the list-separator changed).
2342 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2343 view the policy declared in the DMARC record. Currently, $dmarc_status
2344 is a combined value of both the record presence and the result of the
2347 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2349 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2350 dnslookup router and the smtp transport (applying to the forward
2353 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2354 of ldap servers used for a specific lookup. Patch provided by Heiko
2357 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2358 New variable $lookup_dnssec_authenticated for observability.
2360 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2361 Patch submitted by Lars Timman.
2363 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2365 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2366 Requires trusted mode and valid format message id, aborts otherwise.
2367 Patch contributed by Heiko Schlichting.
2369 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2370 certextract with support for various fields. Bug 1358.
2372 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2373 is requested by default, modifiable by smtp transport option
2376 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2377 operate on certificate variables to give certificate fingerprints
2378 Also new ${sha256:cert_variable}.
2380 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2382 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2383 Christian Aistleitner.
2385 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2387 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2388 file. Patch from Wolfgang Breyha.
2390 JH/25 Expand the coverage of the delivery $host and $host_address to
2391 client authenticators run in verify callout. Bug 1476.
2393 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2394 align with daemon_smtp_ports. Bug 72.
2396 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2397 support and error reporting did not work properly.
2399 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2400 and is readable. Patch from Andrew Colin Kissa.
2402 TL/14 Enhance documentation of ${run expansion and how it parses the
2403 commandline after expansion, particularly in the case when an
2404 unquoted variable expansion results in an empty value.
2406 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2408 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2409 by renaming away from it. Addresses GH issue 15, reported by
2412 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2414 TL/15 SECURITY: prevent double expansion in math comparison functions
2415 (can expand unsanitized data). Not remotely exploitable.
2422 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2424 PP/02 Make -n do something, by making it not do something.
2425 When combined with -bP, the name of an option is not output.
2427 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2430 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2431 $sender_host_name and config options to manage this, and basic check
2434 PP/05 DSCP support for outbound connections and control modifier for inbound.
2436 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2437 (Only plugin which currently uses this is kerberos4, which nobody should
2438 be using, but we should make it available and other future plugins might
2439 conceivably use it, even though it would break NAT; stuff *should* be
2440 using channel bindings instead).
2442 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2443 name; added for Sendmail compatibility; requires admin caller.
2444 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2445 just ignore it); requires trusted caller.
2446 Also parse but ignore: -Ac -Am -X<logfile>
2449 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2451 TL/02 Add +smtp_confirmation as a default logging option.
2453 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2454 Patch by Magnus Holmgren from 2007-02-20.
2456 TL/04 Bugzilla 1281 - Spec typo.
2457 Bugzilla 1283 - Spec typo.
2458 Bugzilla 1290 - Spec grammar fixes.
2460 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2462 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2464 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2465 mailing list by by Dmitry Isaikin.
2467 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2469 JH/02 Support "G" suffix to numbers in ${if comparisons.
2471 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2473 NM/01 Bugzilla 1197 - Spec typo
2474 Bugzilla 1196 - Spec examples corrections
2476 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2478 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2479 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2482 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2483 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2485 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2486 "acl {{name}{arg}...}", and optional args on acl condition
2489 JH/05 Permit multiple router/transport headers_add/remove lines.
2491 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2493 JH/07 Avoid using a waiting database for a single-message-only transport.
2494 Performance patch from Paul Fisher. Bugzilla 1262.
2496 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2499 JH/09 Add $headers_added variable, with content from use of ACL modifier
2500 add_header (but not yet added to the message). Bugzilla 199.
2502 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2503 Pulled from Bugzilla 817 by Wolfgang Breyha.
2505 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2507 (nb: this is the same fix as in Exim 4.80.1)
2509 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2512 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2515 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2517 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2518 advertise SMTP AUTH mechanism to us, instead of a generic
2519 protocol violation error. Also, make Exim more robust to bad
2520 data from the Dovecot auth socket.
2522 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2524 When a queue runner is handling a message, Exim first routes the
2525 recipient addresses, during which it prunes them based on the retry
2526 hints database. After that it attempts to deliver the message to
2527 any remaining recipients. It then updates the hints database using
2530 So if a recipient address works intermittently, it can get repeatedly
2531 deferred at routing time. The retry hints record remains fresh so the
2532 address never reaches the final cutoff time.
2534 This is a fairly common occurrence when a user is bumping up against
2535 their storage quota. Exim had some logic in its local delivery code
2536 to deal with this. However it did not apply to per-recipient defers
2537 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2539 This change adds a proper retry rule check during routing so that the
2540 final cutoff time is checked against the message's age. We only do
2541 this check if there is an address retry record and there is not a
2542 domain retry record; this implies that previous attempts to handle
2543 the address had the retry_use_local_parts option turned on. We use
2544 this as an approximation for the destination being like a local
2545 delivery, as in LMTP.
2547 I suspect this new check makes the old local delivery cutoff check
2548 redundant, but I have not verified this so I left the code in place.
2550 TF/02 Correct gecos expansion when From: is a prefix of the username.
2552 Test 0254 submits a message to Exim with the header
2556 When I ran the test suite under the user fanf2, Exim expanded
2557 the header to contain my full name, whereas it should have added
2558 a Resent-Sender: header. It erroneously treats any prefix of the
2559 username as equal to the username.
2561 This change corrects that bug.
2563 GF/01 DCC debug and logging tidyup
2564 Error conditions log to paniclog rather than rejectlog.
2565 Debug lines prefixed by "DCC: " to remove any ambiguity.
2567 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2569 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2570 Bug spotted by Jeremy Harris; was flawed since initial commit.
2571 Would have resulted in OCSP responses post-SNI triggering an Exim
2572 NULL dereference and crash.
2574 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2576 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2577 Bug detection, analysis and fix by Samuel Thibault.
2578 Bugzilla 1331, Debian bug #698092.
2580 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2582 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2583 Server implementation by Todd Lyons, client by JH.
2584 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2585 config variable "prdr_enable" controls whether the server
2586 advertises the facility. If the client requests PRDR a new
2587 acl_data_smtp_prdr ACL is called once for each recipient, after
2588 the body content is received and before the acl_smtp_data ACL.
2589 The client is controlled by both of: a hosts_try_prdr option
2590 on the smtp transport, and the server advertisement.
2591 Default client logging of deliveries and rejections involving
2592 PRDR are flagged with the string "PRDR".
2594 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2595 fclose(). Diagnosis by Todd Lyons.
2597 PP/17 Update configure.default to handle IPv6 localhost better.
2598 Patch by Alain Williams (plus minor tweaks).
2601 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2602 This is now consistent with GnuTLS, and is now documented: the
2603 previous undocumented portable approach to treating the option as
2604 unset was to force an expansion failure. That still works, and
2605 an empty string is now equivalent.
2607 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2608 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2609 not performing validation itself.
2611 PP/20 Added force_command boolean option to pipe transport.
2612 Patch from Nick Koston, of cPanel Inc.
2614 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2617 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2619 PP/21 Fix eximon continuous updating with timestamped log-files.
2620 Broken in a format-string cleanup in 4.80, missed when I repaired the
2621 other false fix of the same issue.
2622 Report and fix from Heiko Schlichting.
2625 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2626 Report from Prashanth Katuri.
2628 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2629 It's SecureTransport, so affects any MacOS clients which use the
2630 system-integrated TLS libraries, including email clients.
2632 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2633 using a MIME ACL for non-SMTP local injection.
2634 Report and assistance in diagnosis by Warren Baker.
2636 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2638 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2640 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2641 last id that failed so it may be referenced in subsequent ACL's.
2643 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2646 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2647 ldap library initialization, allowing self-signed CA's to be
2648 used. Also properly sets require_cert option later in code by
2649 using NULL (global ldap config) instead of ldap handle (per
2650 session). Bug diagnosis and testing by alxgomz.
2652 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2653 the src/util/ subdirectory.
2655 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2656 renamed to Transport Post Delivery Action by Jeremy Harris, as
2659 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2660 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2661 redis_servers = needs to be configured which will be used by the redis
2662 lookup. Patch from Warren Baker, of The Packet Hub.
2664 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2666 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2667 hostname or reverse DNS when processing a host list. Used suggestions
2668 from multiple comments on this bug.
2670 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2672 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2673 Missed a few lines, added it to make the runtest require no keyboard
2676 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2677 contains upper case chars. Make router use caseful_local_part.
2679 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2680 support when GnuTLS has been built with p11-kit.
2686 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2688 This, or similar/improved, will also be change PP/11 of 4.82.
2694 PP/01 Handle short writes when writing local log-files.
2695 In practice, only affects FreeBSD (8 onwards).
2696 Bugzilla 1053, with thanks to Dmitry Isaikin.
2698 NM/01 Bugzilla 949 - Documentation tweak
2700 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2703 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2705 PP/02 Implemented gsasl authenticator.
2707 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2709 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2710 `pkg-config foo` for cflags/libs.
2712 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2713 with rest of GSASL and with heimdal_gssapi.
2715 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2716 `pkg-config foo` for cflags/libs for the TLS implementation.
2718 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2719 properties get this fed in as external SSF. A number of robustness
2720 and debugging improvements to the cyrus_sasl authenticator.
2722 PP/08 cyrus_sasl server now expands the server_realm option.
2724 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2725 Patch by Jeremy Harris.
2727 PP/10 Added dbmjz lookup type.
2729 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2731 PP/12 MAIL args handles TAB as well as SP, for better interop with
2732 non-compliant senders.
2733 Analysis and variant patch by Todd Lyons.
2735 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2736 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2737 Patch from Dirk Mueller <dmueller@suse.com>
2739 PP/13 tls_peerdn now print-escaped for spool files.
2740 Observed some $tls_peerdn in wild which contained \n, which resulted
2741 in spool file corruption.
2743 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2744 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2745 or write after TLS renegotiation, which otherwise led to messages
2748 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2749 as a tracking header (ie: a signed header comes before the signature).
2750 Patch from Wolfgang Breyha.
2752 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2753 comma-sep list; embedded commas doubled.
2755 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2757 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2759 Report and patch from Dmitry Banschikov.
2761 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2762 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2763 failures. We appear to now support TLS1.1+ with Exim.
2765 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2766 lets Exim select keys and certificates based upon TLS SNI from client.
2767 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2768 before an outbound SMTP session. New log_selector, +tls_sni.
2770 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2771 NULL dereference. Report and patch from Alun Jones.
2773 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2774 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2775 Not seeing resolver debug output on NetBSD, but suspect this is a
2776 resolver implementation change.
2778 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2779 Left warnings. Added "eximon gdb" invocation mode.
2781 PP/21 Defaulting "accept_8bitmime" to true, not false.
2783 PP/22 Added -bw for inetd wait mode support.
2785 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2786 locate the relevant includes and libraries. Made this the default.
2788 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2789 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2791 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2792 This may cause build issues on older platforms.
2794 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2795 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2796 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2797 Added SNI support via GnuTLS too.
2798 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2800 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2802 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2803 Applied second patch from Janne, implementing suggestion to default
2804 multiple-strings-in-record handling to match SPF spec.
2806 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2808 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2809 read-only, out of scope).
2810 Patch from Wolfgang Breyha, report from Stuart Northfield.
2812 PP/29 Fix three issues highlighted by clang analyser static analysis.
2813 Only crash-plausible issue would require the Cambridge-specific
2814 iplookup router and a misconfiguration.
2815 Report from Marcin Mirosław.
2817 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2819 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2820 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2821 As part of this, removing so much warning spew let me fix some minor
2822 real issues in debug logging.
2824 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2825 assignment on my part. Fixed.
2827 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2828 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2829 Janne Snabb (who went above and beyond: thank you).
2831 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2832 string otherwise requires a connection and a bunch more work and it's
2833 relatively easy to get wrong. Should also expose TLS library linkage
2836 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2837 64-bit ${eval} (JH/03).
2839 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2840 GNU libc to support some of the 64-bit stuff, should not lead to
2841 conflicts. Defined before os.h is pulled in, so if a given platform
2842 needs to override this, it can.
2844 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2845 protection layer was required, which is not implemented.
2846 Bugzilla 1254, patch from Wolfgang Breyha.
2848 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2849 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2850 tls_dhparam take prime identifiers. Also unbreak combination of
2851 OpenSSL+DH_params+TLSSNI.
2853 PP/39 Disable SSLv2 by default in OpenSSL support.
2859 PP/01 Solaris build fix for Oracle's LDAP libraries.
2860 Bugzilla 1109, patch from Stephen Usher.
2862 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2864 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2867 TF/02 Fix a couple more cases where we did not log the error message
2868 when unlink() failed. See also change 4.74-TF/03.
2870 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2871 lock up or crash if it happened to be inside a call to libc when it
2872 got a SIGUSR1 from exiwhat.
2874 The SIGUSR1 handler appends the current process status to the process
2875 log which is later printed by exiwhat. It used to use the general
2876 purpose logging code to do this, but several functions it calls are
2877 not safe for signals.
2879 The new output code in the SIGUSR1 handler is specific to the process
2880 log, and simple enough that it's easy to inspect for signal safety.
2881 Removing some special cases also simplifies the general logging code.
2882 Removing the spurious timestamps from the process log simplifies
2885 TF/04 Improved ratelimit ACL condition.
2887 The /noupdate option has been deprecated in favour of /readonly which
2888 has clearer semantics. The /leaky, /strict, and /readonly update modes
2889 are mutually exclusive. The update mode is no longer included in the
2890 database key; it just determines when the database is updated. (This
2891 means that when you upgrade Exim will forget old rate measurements.)
2893 Exim now checks that the per_* options are used with an update mode that
2894 makes sense for the current ACL. For example, when Exim is processing a
2895 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2896 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2897 must specify per_mail/readonly. If you omit the update mode it defaults to
2898 /leaky where that makes sense (as before) or /readonly where required.
2900 The /noupdate option is now undocumented but still supported for
2901 backwards compatibility. It is equivalent to /readonly except that in
2902 ACLs where /readonly is required you may specify /leaky/noupdate or
2903 /strict/noupdate which are treated the same as /readonly.
2905 A useful new feature is the /count= option. This is a generalization
2906 of the per_byte option, so that you can measure the throughput of other
2907 aggregate values. For example, the per_byte option is now equivalent
2908 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2910 The per_rcpt option has been generalized using the /count= mechanism
2911 (though it's more complicated than the per_byte equivalence). When it is
2912 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2913 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2914 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2915 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2916 Note that using per_rcpt with a non-readonly update mode in more than
2917 one ACL will cause the recipients to be double-counted. (The per_mail
2918 and per_byte options don't have this problem.)
2920 The handling of very low rates has changed slightly. If the computed rate
2921 is less than the event's count (usually one) then this event is the first
2922 after a long gap. In this case the rate is set to the same as this event's
2923 count, so that the first message of a spam run is counted properly.
2925 The major new feature is a mechanism for counting the rate of unique
2926 events. The new per_addr option counts the number of different
2927 recipients that someone has sent messages to in the last time period. It
2928 behaves like per_rcpt if all the recipient addresses are different, but
2929 duplicate recipient addresses do not increase the measured rate. Like
2930 the /count= option this is a general mechanism, so the per_addr option
2931 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2932 example, measure the rate that a client uses different sender addresses
2933 with the options per_mail/unique=$sender_address. There are further
2934 details in the main documentation.
2936 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2938 TF/06 Removed a few PCRE remnants.
2940 TF/07 Automatically extract Exim's version number from tags in the git
2941 repository when doing development or release builds.
2943 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2944 Bugzilla 879. Patch from Paul Fisher.
2946 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2947 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2950 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2952 PP/05 Variable $av_failed, true if the AV scanner deferred.
2953 Bugzilla 1078. Patch from John Horne.
2955 PP/06 Stop make process more reliably on build failure.
2956 Bugzilla 1087. Patch from Heiko Schlittermann.
2958 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2959 Bugzilla 1089. Patch from Heiko Schlittermann.
2961 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2962 Bugzilla 1131. Patch from Holger Weiß.
2964 PP/09 Handle IPv6 addresses with SPF.
2965 Bugzilla 860. Patch from Wolfgang Breyha.
2967 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2969 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2972 PP/11 match_* no longer expand right-hand-side by default.
2973 New compile-time build option, EXPAND_LISTMATCH_RHS.
2974 New expansion conditions, "inlist", "inlisti".
2976 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2978 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2980 PP/14 fix log_write() format string regression from TF/03.
2981 Bugzilla 1152. Patch from Dmitry Isaikin.
2987 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2989 PP/02 Harmonised TLS library version reporting; only show if debugging.
2990 Layout now matches that introduced for other libraries in 4.74 PP/03.
2992 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2994 PP/04 New "dns_use_edns0" global option.
2996 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2999 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3000 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3002 TK/01 Updated PolarSSL code to 0.14.2.
3003 Bugzilla 1097. Patch from Andreas Metzler.
3005 PP/07 Catch divide-by-zero in ${eval:...}.
3006 Fixes bugzilla 1102.
3008 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3011 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3012 format-string attack -- SECURITY: remote arbitrary code execution.
3014 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3015 time unintentionally subject to list matching rules, letting the header
3016 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3017 arbitrary string expansion). This allowed for information disclosure.
3019 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3020 INT_MIN/-1 -- value coerced to INT_MAX.
3026 NM/01 Workaround for PCRE version dependency in version reporting
3029 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3030 This fixes portability to compilers other than gcc, notably
3031 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3033 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3034 makefiles for portability to HP-UX and POSIX correctness.
3036 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3037 Also via indented variable definition in the Makefile.
3038 (Debugging by Oliver Heesakkers).
3040 PP/02 Restore caching of spamd results with expanded spamd_address.
3041 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3043 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3044 Improves build reliability. Fix from: Frank Elsner
3046 NM/02 Fix wide character breakage in the rfc2047 coding
3047 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3049 NM/03 Allow underscore in dnslist lookups
3050 Fixes bug 1026. Patch from Graeme Fowler
3052 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3053 Code patches from Adam Ciarcinski of NetBSD.
3055 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3058 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3059 is logged, to avoid truncation. Patch from John Horne.
3061 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3062 Patch from Jakob Hirsch.
3064 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3065 SQL string expansion failure details.
3066 Patch from Andrey Oktyabrski.
3068 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3069 Patch from Simon Arlott.
3071 PP/09 New lookups functionality failed to compile on old gcc which rejects
3072 extern declarations in function scope.
3073 Patch from Oliver Fleischmann
3075 PP/10 Use sig_atomic_t for flags set from signal handlers.
3076 Check getgroups() return and improve debugging.
3077 Fixed developed for diagnosis in bug 927 (which turned out to be
3080 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3081 Patch from Mark Zealey.
3083 PP/12 Bugzilla 1056: Improved spamd server selection.
3084 Patch from Mark Zealey.
3086 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3087 Based on patch from Heiko Schlittermann.
3089 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3090 Patch from Uwe Doering, sign-off by Michael Haardt.
3092 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3093 variable declaration deep within a block. Bug and patch from
3096 PP/15 lookups-Makefile IRIX compatibility coercion.
3098 PP/16 Make DISABLE_DKIM build knob functional.
3100 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3101 Patch by Simon Arlott
3103 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3104 variable argument macros. Our copy now differs from upstream.
3110 TF/01 Failure to get a lock on a hints database can have serious
3111 consequences so log it to the panic log.
3113 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3114 controlled using the smtp_confirmation log selector.
3116 TF/03 Include the error message when we fail to unlink a spool file.
3118 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3119 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3120 for maintaining out-of-tree patches for some time.
3122 PP/01 Bugzilla 139: Documentation and portability issues.
3123 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3124 Handle per-OS dynamic-module compilation flags.
3126 PP/02 Let /dev/null have normal permissions.
3127 The 4.73 fixes were a little too stringent and complained about the
3128 permissions on /dev/null. Exempt it from some checks.
3129 Reported by Andreas M. Kirchwitz.
3131 PP/03 Report version information for many libraries, including
3132 Exim version information for dynamically loaded libraries. Created
3133 version.h, now support a version extension string for distributors
3134 who patch heavily. Dynamic module ABI change.
3136 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3137 privilege escalation vulnerability whereby the Exim run-time user
3138 can cause root to append content of the attacker's choosing to
3141 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3144 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3145 If dropping privileges for untrusted macros, we disabled normal logging
3146 on the basis that it would fail; for the Exim run-time user, this is not
3147 the case, and it resulted in successful deliveries going unlogged.
3148 Fixed. Reported by Andreas Metzler.
3154 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3155 only prepend for the Resent-* case. Fixes regression introduced in
3156 Exim 4.70 by NM/22 for Bugzilla 607.
3158 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3159 increasing numbers of administrators be bitten by this.
3161 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3163 PP/03 Bugzilla 994: added openssl_options main configuration option.
3165 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3167 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3169 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3171 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3172 without a peer certificate, leading to a segfault because of an
3173 assumption that peers always have certificates. Be a little more
3174 paranoid. Problem reported by Martin Tscholak.
3176 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3177 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3178 NB: ClamAV planning to remove STREAM in "middle of 2010".
3179 CL also introduces -bmalware, various -d+acl logging additions and
3180 more caution in buffer sizes.
3182 PP/09 Implemented reverse_ip expansion operator.
3184 PP/10 Bugzilla 937: provide a "debug" ACL control.
3186 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3188 PP/12 Bugzilla 973: Implement --version.
3190 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3192 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3194 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3196 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3197 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3198 ignore trailing whitespace.
3200 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3202 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3205 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3206 Notification from Dr Andrew Aitchison.
3208 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3209 ExtendedDetectionInfo response format.
3210 Notification from John Horne.
3212 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3215 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3216 XSL and documented dependency on system catalogs, with examples of how
3219 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3222 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3223 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3224 configuration file which is writeable by the Exim user or group.
3226 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3227 of configuration files to cover files specified with the -C option if
3228 they are going to be used with root privileges, not just the default
3231 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3232 option (effectively making it always true).
3234 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3235 files to be used while preserving root privileges.
3237 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3238 that rogue child processes cannot use them.
3240 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3241 run-time user, instead of root.
3243 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3244 Exim run-time user without dropping privileges.
3246 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3247 result string, instead of calling string_vformat() twice with the same
3250 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3251 for other users. Others should always drop root privileges if they use
3252 -C on the command line, even for a whitelisted configure file.
3254 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3256 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3262 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3263 $data_path, and $header_path variables; fixed documentation bugs and
3266 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3267 exipick to access non-standard spools, including the "frozen" queue
3270 NM/01 Bugzilla 965: Support mysql stored procedures.
3271 Patch from Alain Williams
3273 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3275 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3276 Patch from Andreas Metzler
3278 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3279 Patch from Kirill Miazine
3281 NM/05 Bugzilla 671: Added umask to procmail example.
3283 JJ/03 installed exipick 20100323.0, fixing doc bug
3285 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3286 directory. Notification and patch from Dan Rosenberg.
3288 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3290 TK/02 Improve log output when DKIM signing operation fails.
3292 MH/01 Treat the transport option dkim_domain as a colon separated
3293 list, not as a single string, and sign the message with each element,
3294 omitting multiple occurences of the same signer.
3296 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3297 Bugzilla 985, 986. Patch by Simon Arlott
3299 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3300 Patch by Simon Arlott
3302 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3303 MBX locking. Notification from Dan Rosenberg.
3309 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3311 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3313 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3315 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3317 NM/04 Bugzilla 915: Flag broken perl installation during build.
3323 TK/01 Added patch by Johannes Berg that expands the main option
3324 "spamd_address" if it starts with a dollar sign.
3326 TK/02 Write list of recipients to X-Envelope-Sender header when building
3327 the mbox-format spool file for content scanning (suggested by Jakob
3330 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3331 (http://www.dcc-servers.net/) support via dccifd. Activated by
3332 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3334 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3335 by Mark Daniel Reidel <mr@df.eu>.
3337 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3338 When building exim an external PCRE library is now needed -
3339 PCRE is a system library on the majority of modern systems.
3340 See entry on PCRE_LIBS in EDITME file.
3342 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3343 conversation. Added nologin parameter to request.
3344 Patch contributed by Kirill Miazine.
3346 TF/01 Do not log submission mode rewrites if they do not change the address.
3348 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3350 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3351 log files in place. Contributed by Roberto Lima.
3353 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3355 TF/03 Bugzilla 615: When checking the local_parts router precondition
3356 after a local_part_suffix or local_part_prefix option, Exim now
3357 does not use the address's named list lookup cache, since this
3358 contains cached lookups for the whole local part.
3360 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3361 Robert Millan. Documentation is in experimental-spec.txt.
3363 TF/04 Bugzilla 668: Fix parallel build (make -j).
3365 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3367 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3368 Patch provided by Jan Srzednicki.
3370 TF/05 Leading white space used to be stripped from $spam_report which
3371 wrecked the formatting. Now it is preserved.
3373 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3374 that they are available at delivery time.
3376 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3378 TF/08 TLS error reporting now respects the incoming_interface and
3379 incoming_port log selectors.
3381 TF/09 Produce a more useful error message if an SMTP transport's hosts
3382 setting expands to an empty string.
3384 NM/06 Bugzilla 744: EXPN did not work under TLS.
3385 Patch provided by Phil Pennock.
3387 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3388 Patch provided by Richard Godbee.
3390 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3391 acl_smtp_notquit, added index entry.
3393 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3394 Patch provided by Eugene Bujak.
3396 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3397 accept(). Patch provided by Maxim Dounin.
3399 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3400 Patch provided by Phil Pennock.
3402 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3404 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3405 Patch provided by Brad "anomie" Jorsch.
3407 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3408 Patch provided by Dean Brooks.
3410 TK/05 Add native DKIM support (does not depend on external libraries).
3412 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3413 Patch provided by Graeme Fowler.
3415 NM/16 Bugzilla 851: Documentation example syntax fix.
3417 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3419 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3422 NM/19 Bugzilla 745: TLS version reporting.
3423 Patch provided by Phil Pennock.
3425 NM/20 Bugzilla 167: bool: condition support.
3426 Patch provided by Phil Pennock.
3428 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3429 clients. Patch provided by Phil Pennock.
3431 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3432 Patch provided by Brad "anomie" Jorsch.
3434 NM/23 Bugzilla 687: Fix misparses in eximstats.
3435 Patch provided by Heiko Schlittermann.
3437 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3438 Patch provided by Heiko Schlittermann.
3440 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3441 plus update to original patch.
3443 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3445 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3446 Patch provided by David Brownlee.
3448 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3450 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3452 NM/30 Bugzilla 888: TLS documentation bugfixes.
3454 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3456 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3457 Unlike the original bugzilla I have changed all shell scripts in src tree.
3459 NM/33 Bugzilla 898: Transport filter timeout fix.
3460 Patch by Todd Rinaldo.
3462 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3463 Patch by Serge Demonchaux.
3465 NM/35 Bugzilla 39: Base64 decode bug fixes.
3466 Patch by Jakob Hirsch.
3468 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3470 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3472 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3474 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3480 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3481 ALT-N's libdkim that I have put here:
3482 http://duncanthrax.net/exim-experimental/
3484 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3485 were called 'true' and it seems that C99 defines that as a reserved
3486 keyword to be used with 'bool' variable types. That means you could
3487 not include C99-style headers which use bools without triggering
3488 build errors in sieve.c.
3490 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3491 as mailq or other aliases. Changed the --help handling significantly
3492 to do whats expected. exim_usage() emits usage/help information.
3494 SC/01 Added the -bylocaldomain option to eximstats.
3496 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3498 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3500 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3506 PH/01 Another patch from the Sieve maintainer.
3508 PH/02 When an IPv6 address is converted to a string for single-key lookup
3509 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3510 dots are used instead of colons so that keys in lsearch files need not
3511 contain colons. This was done some time before quoting was made available
3512 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3513 (notated using the quote facility) so as to distinguish them from IPv4
3514 keys. This meant that lookups for IP addresses in host lists did not work
3515 for iplsearch lookups.
3517 This has been fixed by arranging for IPv6 addresses to be expressed with
3518 colons if the lookup type is iplsearch. This is not incompatible, because
3519 previously such lookups could never work.
3521 The situation is now rather anomalous, since one *can* have colons in
3522 ordinary lsearch keys. However, making the change in all cases is
3523 incompatible and would probably break a number of configurations.
3525 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3528 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3529 conversion specification without a maximum field width, thereby enabling
3530 a rogue spamd server to cause a buffer overflow. While nobody in their
3531 right mind would setup Exim to query an untrusted spamd server, an
3532 attacker that gains access to a server running spamd could potentially
3533 exploit this vulnerability to run arbitrary code as the Exim user.
3535 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3536 $primary_hostname instead of what libspf2 thinks the hosts name is.
3538 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3539 a directory entry by the name of the lookup key. Previously, if a
3540 symlink pointed to a non-existing file or a file in a directory that
3541 Exim lacked permissions to read, a lookup for a key matching that
3542 symlink would fail. Now it is enough that a matching directory entry
3543 exists, symlink or not. (Bugzilla 503.)
3545 PH/03 The body_linecount and body_zerocount variables are now exported in the
3548 PH/04 Added the $dnslist_matched variable.
3550 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3551 This means they are set thereafter only if the connection becomes
3554 PH/06 Added the client_condition to authenticators so that some can be skipped
3555 by clients under certain conditions.
3557 PH/07 The error message for a badly-placed control=no_multiline_responses left
3558 "_responses" off the end of the name.
3560 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3562 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3563 (without spaces) instead of just copying the configuration text.
3565 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3567 PH/11 Added $max_received_linelength.
3569 PH/12 Added +ignore_defer and +include_defer to host lists.
3571 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3572 way in which PCRE > 7.0 is built.
3574 PH/14 Implemented queue_only_load_latch.
3576 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3577 MAIL command. The effect was to mangle the value on 64-bit systems.
3579 PH/16 Another patch from the Sieve maintainer.
3581 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3583 PH/18 If a system quota error occurred while trying to create the file for
3584 a maildir delivery, the message "Mailbox is full" was not appended to the
3585 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3586 only to a quota excession during the actual writing of the file.
3588 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3589 characters?) which causes problems in log lines. The DN values are now
3590 passed through string_printing() before being added to log lines.
3592 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3593 and InterBase are left for another time.)
3595 PH/21 Added message_body_newlines option.
3597 PH/22 Guard against possible overflow in moan_check_errorcopy().
3599 PH/23 POSIX allows open() to be a macro; guard against that.
3601 PH/24 If the recipient of an error message contained an @ in the local part
3602 (suitably quoted, of course), incorrect values were put in $domain and
3603 $local_part during the evaluation of errors_copy.
3609 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3610 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3613 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3614 issue a MAIL command.
3616 PH/02 In an ACL statement such as
3618 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3620 if a client was not listed at all, or was listed with a value other than
3621 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3622 the condition was not true (as it should be), so access was not denied.
3623 The bug was that the ! inversion was incorrectly passed on to the second
3624 item. This has been fixed.
3626 PH/03 Added additional dnslists conditions == and =& which are different from
3627 = and & when the dns lookup returns more than one IP address.
3629 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3630 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3632 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3633 FSYNC, which compiles an option called disable_fsync that allows for
3634 bypassing fsync(). The documentation is heavily laced with warnings.
3636 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3638 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3639 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3640 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3641 including adding "make clean"; (3) Added -fPIC when compiling the test
3642 dynamically loaded module, to get rid of a warning.
3644 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3645 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3646 = 0s. The bug is otherwise harmless.
3648 PH/07 There was a bug in the dovecot authenticator such that the value of
3649 $auth1 could be overwritten, and so not correctly preserved, after a
3650 successful authentication. This usually meant that the value preserved by
3651 the server_setid option was incorrect.
3653 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3655 PH/09 Installed PCRE release 7.0.
3657 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3658 run for batched SMTP input. It is now run at the start of every message
3659 in the batch. While fixing this I discovered that the process information
3660 (output by running exiwhat) was not always getting set for -bs and -bS
3661 input. This is fixed, and it now also says "batched" for BSMTP.
3663 PH/11 Added control=no_pipelining.
3665 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3666 patch, slightly modified), and move the expansion of helo_data till after
3667 the connection is made in the smtp transport (so it can use these
3670 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3672 PH/14 Added log_selector = +pid.
3674 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3676 PH/16 Add ${if forany and ${if forall.
3678 PH/17 Added dsn_from option to vary the From: line in DSNs.
3680 PH/18 Flush SMTP output before performing a callout, unless control =
3681 no_callout_flush is set.
3683 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3684 was true (the default) a successful delivery failed to delete the retry
3685 item, thus causing premature timeout of the address. The bug is now
3688 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3690 PH/21 Long custom messages for fakedefer and fakereject are now split up
3691 into multiline responses in the same way that messages for "deny" and
3692 other ACL rejections are.
3694 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3695 with slight modification.
3697 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3698 draft, changing the syntax and factoring some duplicate code".
3700 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3701 for deliveries of the second and subsequent messages over the same SMTP
3704 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3705 ${reduce, with only minor "tidies".
3707 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3709 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3710 expansion side effects.
3712 PH/27 When a message times out after an over-quota error from an Exim-imposed
3713 quota, the bounce message says "mailbox is full". This message was not
3714 being given when it was a system quota that was exceeded. It now should
3717 MH/03 Made $recipients available in local_scan(). local_scan() already has
3718 better access to the recipient list through recipients_list[], but
3719 $recipients can be useful in postmaster-provided expansion strings.
3721 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3722 in the case of a MAIL command with additional options following the
3723 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3724 were accidentally chopped off.
3726 PH/29 SMTP synchronization checks are implemented when a command is read -
3727 there is a check that no more input is waiting when there shouldn't be
3728 any. However, for some commands, a delay in an ACL can mean that it is
3729 some time before the response is written. In this time, more input might
3730 arrive, invalidly. So now there are extra checks after an ACL has run for
3731 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3732 pipelining has not been advertised.
3734 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3736 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3737 returned in the SMTP response when smtp_return_error_details was false.
3738 This has been fixed.
3740 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3741 instead of the C I/O that was originally supplied, because problems were
3742 reported on Solaris.
3744 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3745 Exim which did not show up earlier: it was assuming that a call to
3746 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3747 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3748 was a macro that became an assignment, so it seemed to work. This has
3749 changed to a proper function call with a void return, hence the compile
3750 error. Exim's code has been fixed.
3752 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3755 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3757 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3759 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3760 because it turns out to be too big or there is a timeout) there is no
3761 "Completed" line in the log. When some messages of this type were
3762 selected by exigrep, they were listed as "not completed". Others were
3763 picked up by some special patterns. I have improved the selection
3764 criteria to be more general.
3766 PH/38 The host_find_failed option in the manualroute router can now be set
3767 to "ignore", to completely ignore a host whose IP address cannot be
3768 found. If all hosts are ignored, the behaviour is controlled by the new
3769 host_all_ignored option.
3771 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3772 homing or because of multiple MX records with /mx) generated more than
3773 one IP address, and the following item turned out to be the local host,
3774 all the secondary addresses of the first item were incorrectly removed
3775 from the list, along with the local host and any following hosts (which
3776 is what is supposed to happen).
3778 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3779 whoever called Exim into the -H file. In the case of the daemon it was
3780 behaving confusingly. When first started, it used values for whoever
3781 started the daemon, but after a SIGHUP it used the Exim user (because it
3782 calls itself on a restart). I have changed the code so that it now always
3785 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3786 message are rejected with the same error (e.g. no authentication or bad
3787 sender address), and a DATA command is nevertheless sent (as can happen
3788 with PIPELINING or a stupid MUA), the error message that was given to the
3789 RCPT commands is included in the rejection of the DATA command. This is
3790 intended to be helpful for MUAs that show only the final error to their
3793 PH/42 Another patch from the Sieve maintainer.
3795 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3796 Eximstats - Fixed some broken HTML links and added missing column headers
3798 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3799 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3801 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3803 PH/43 Yet another patch from the Sieve maintainer.
3805 PH/44 I found a way to check for a TCP/IP connection going away before sending
3806 the response to the final '.' that terminates a message, but only in the
3807 case where the client has not sent further data following the '.'
3808 (unfortunately, this is allowed). However, in many cases there won't be
3809 any further data because there won't be any more messages to send. A call
3810 to select() can be used: if it shows that the input is "ready", there is
3811 either input waiting, or the socket has been closed. An attempt to read
3812 the next input character can distinguish the two cases. Previously, Exim
3813 would have sent an OK response which the client would never have see.
3814 This could lead to message repetition. This fix should cure that, at
3815 least in a lot of common cases.
3817 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3818 advertised in response to EHLO.
3824 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3825 fixed by 4.65/MH/01 (is this a record?) are fixed:
3827 (i) An empty string was always treated as zero by the numeric comparison
3828 operators. This behaviour has been restored.
3830 (ii) It is documented that the numeric comparison operators always treat
3831 their arguments as decimal numbers. This was broken in that numbers
3832 starting with 0 were being interpreted as octal.
3834 While fixing these problems I realized that there was another issue that
3835 hadn't been noticed. Values of message_size_limit (both the global option
3836 and the transport option) were treated as octal if they started with 0.
3837 The documentation was vague. These values are now always treated as
3838 decimal, and I will make that clear in the documentation.
3844 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3845 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3848 MH/01 Don't check that the operands of numeric comparison operators are
3849 integers when their expansion is in "skipping" mode (fixes bug
3850 introduced by 4.64-PH/07).
3852 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3853 child addresses, Exim now panics and dies. Previously, because the count
3854 is held in a short int, deliveries were likely to be lost. As such a
3855 large number of recipients for a single message is ridiculous
3856 (performance will be very, very poor), I have chosen to impose a limit
3857 rather than extend the field.
3863 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3864 leftover -K file (the existence of which was triggered by #402).
3865 While we were at it, introduced process PID as part of the -K
3866 filename. This should rule out race conditions when creating
3869 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3870 processing considerably. Previous code took too long for large mails,
3871 triggering a timeout which in turn triggers #401.
3873 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3874 in the DK code in transports.c. sendfile() is not really portable,
3875 hence the _LINUX specificness.
3877 TF/01 In the add_headers option to the mail command in an Exim filter,
3878 there was a bug that Exim would claim a syntax error in any
3879 header after the first one which had an odd number of characters
3882 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3883 callout verification, Exim cached a "reject" for the entire domain. This
3884 is correct for most verifications, but it is not correct for a recipient
3885 verification with use_sender or use_postmaster set, because in that case
3886 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3887 case of MAIL FROM:<> rejection from other early rejections (e.g.
3888 rejection of HELO). When verifying a recipient using a non-null MAIL
3889 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3890 Whatever the result of the callout, the value of the domain cache is
3891 left unchanged (for any other kind of callout, getting as far as trying
3892 RCPT means that the domain itself is ok).
3894 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3897 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3898 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3899 session, the daemon ignores ECONNECT errors and logs others; it now
3900 ignores EPIPE as well.
3902 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3903 (quoted-printable decoding).
3905 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3906 later the small subsequent patch to fix an introduced bug.
3908 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3910 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3912 PH/08 An error is now given if message_size_limit is specified negative.
3914 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3915 to be given (somewhat) arbitrary names.
3917 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3920 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3921 miscellaneous code fixes
3923 PH/10 Added the log_reject_target ACL modifier to specify where to log
3926 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3927 hostname. This is wrong, because it relates to the incoming message (and
3928 probably the interface on which it is arriving) and not to the outgoing
3929 callout (which could be using a different interface). This has been
3930 changed to use the value of the helo_data option from the smtp transport
3931 instead - this is what is used when a message is actually being sent. If
3932 there is no remote transport (possible with a router that sets up host
3933 addresses), $smtp_active_hostname is used.
3935 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3936 tweaks were necessary in order to get it to work (see also 21 below):
3937 (a) The code assumed that strncpy() returns a negative number on buffer
3938 overflow, which isn't the case. Replaced with Exim's string_format()
3940 (b) There were several signed/unsigned issues. I just did the minimum
3941 hacking in of casts. There is scope for a larger refactoring.
3942 (c) The code used strcasecmp() which is not a standard C function.
3943 Replaced with Exim's strcmpic() function.
3944 (d) The code set only $1; it now sets $auth1 as well.
3945 (e) A simple test gave the error "authentication client didn't specify
3946 service in request". It would seem that Dovecot has changed its
3947 interface. Fortunately there's a specification; I followed it and
3948 changed what the client sends and it appears to be working now.
3950 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3953 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3954 address A is aliased to B and C, where B exists and C does not. Without
3955 -v the output is "A verified" because verification stops after a
3956 successful redirection if more than one address is generated. However,
3957 with -v the child addresses are also verified. Exim was outputting "A
3958 failed to verify" and then showing the successful verification for C,
3959 with its parentage. It now outputs "B failed to verify", showing B's
3960 parentage before showing the successful verification of C.
3962 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3963 look up a TXT record in a specific list after matching in a combined
3966 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3967 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3968 they consult the DNS. I had assumed they would set it the way they
3969 wanted; and indeed my experiments on Linux seem to show that in some
3970 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3971 To be on the safe side, however, I have now made the interface to
3972 host_find_byname() similar to host_find_bydns(), with an argument
3973 containing the DNS resolver options. The host_find_byname() function now
3974 sets these options at its start, just as host_find_bydns() does. The smtp
3975 transport options dns_qualify_single and dns_search_parents are passed to
3976 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3977 of host_find_byname() use the default settings of RES_DEFNAMES
3978 (qualify_single) but not RES_DNSRCH (search_parents).
3980 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3981 spool_read_header() do less string testing, by means of a preliminary
3982 switch on the second character of optional "-foo" lines. (This is
3983 overdue, caused by the large number of possibilities that now exist.
3984 Originally there were few.) While I was there, I also converted the
3985 str(n)cmp tests so they don't re-test the leading "-" and the first
3986 character, in the hope this might squeeze out yet more improvement.
3988 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3989 flag allowing group syntax was set by the header_syntax check but not
3990 turned off, possible causing trouble later; (2) The flag was not being
3991 set at all for the header_verify test, causing "group"-style headers to
3992 be rejected. I have now set it in this case, and also caused header_
3993 verify to ignore an empty address taken from a group. While doing this, I
3994 came across some other cases where the code for allowing group syntax
3995 while scanning a header line wasn't quite right (mostly, not resetting
3996 the flag correctly in the right place). These bugs could have caused
3997 trouble for malformed header lines. I hope it is now all correct.
3999 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4000 with the "reply" argument non-NULL. The code, however (which originally
4001 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4002 but it didn't always do it. This confused somebody who was copying the
4003 code for some other use. I have removed all the tests.
4005 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4006 feature that was used to support insecure browsers during the U.S. crypto
4007 embargo. It requires special client support, and Exim is probably the
4008 only MTA that supported it -- and would never use it because real RSA is
4009 always available. This code has been removed, because it had the bad
4010 effect of slowing Exim down by computing (never used) parameters for the
4011 RSA_EXPORT functionality.
4013 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4014 authenticator to fail if there's a tab character in the incoming data
4015 (there should never be unless someone is messing about, as it's supposed
4016 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4017 if the connection is using TLS or if the remote IP is the same as the
4018 local IP, and the "valid-client-cert option" if a client certificate has
4021 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4022 authenticators. This can be used for authorization after authentication
4023 succeeds. (In the case of plaintext, it servers for both authentication
4026 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4027 if any retry times were supplied.
4029 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4030 connection, where there is no client IP address to check. In this
4031 situation, the verify now always succeeds.
4033 PH/25 Applied John Jetmore's -Mset patch.
4035 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4037 PH/27 In a string expansion for a processed (not raw) header when multiple
4038 headers of the same name were present, leading whitespace was being
4039 removed from all of them, but trailing whitespace was being removed only
4040 from the last one. Now trailing whitespace is removed from each header
4041 before concatenation. Completely empty headers in a concatenation (as
4042 before) are ignored.
4044 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4045 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4047 PH/29 [Removed. This was a change that I later backed out, and forgot to
4048 correct the ChangeLog entry (that I had efficiently created) before
4049 committing the later change.]
4051 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4052 address errors (4xx response to RCPT) over the same connection as other
4053 messages routed to the same hosts. Such deliveries are always "forced",
4054 so retry times are not inspected. This resulted in far too many retries
4055 for the affected addresses. The effect occurred only when there were more
4056 hosts than the hosts_max_try setting in the smtp transport when it had
4057 the 4xx errors. Those hosts that it had tried were not added to the list
4058 of hosts for which the message was waiting, so if all were tried, there
4059 was no problem. Two fixes have been applied:
4061 (i) If there are any address or message errors in an SMTP delivery, none
4062 of the hosts (tried or untried) are now added to the list of hosts
4063 for which the message is waiting, so the message should not be a
4064 candidate for sending over the same connection that was used for a
4065 successful delivery of some other message. This seems entirely
4066 reasonable: after all the message is NOT "waiting for some host".
4067 This is so "obvious" that I'm not sure why it wasn't done
4068 previously. Hope I haven't missed anything, but it can't do any
4069 harm, as the worst effect is to miss an optimization.
4071 (ii) If, despite (i), such a delivery is accidentally attempted, the
4072 routing retry time is respected, so at least it doesn't keep
4073 hammering the server.
4075 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4076 in ${readsocket because some servers need this prod.
4078 PH/32 Added some extra debug output when updating a wait-xxx database.
4080 PH/33 The hint "could be header name not terminated by colon", which has been
4081 given for certain expansion errors for a long time, was not being given
4082 for the ${if def:h_colon_omitted{... case.
4084 PH/34 The spec says: "With one important exception, whenever a domain list is
4085 being scanned, $domain contains the subject domain." There was at least
4086 one case where this was not true.
4088 PH/35 The error "getsockname() failed: connection reset by peer" was being
4089 written to the panic log as well as the main log, but it isn't really
4090 panic-worthy as it just means the connection died rather early on. I have
4091 removed the panic log writing for the ECONNRESET error when getsockname()
4094 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4095 runs only) independently of the message's sender address. This meant
4096 that, if the 4xx error was in fact related to the sender, a different
4097 message to the same recipient with a different sender could confuse
4098 things. In particular, this can happen when sending to a greylisting
4099 server, but other circumstances could also provoke similar problems.
4100 I have changed the default so that the retry time for these errors is now
4101 based a combination of the sender and recipient addresses. This change
4102 can be overridden by setting address_retry_include_sender=false in the
4105 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4106 remote server are returned as part of bounce messages. This was not
4107 happening for LMTP over a pipe (the lmtp transport), but now it is the
4108 same for both kinds of LMTP.
4110 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4111 in header lines that were in fact CNAMEs. This is no longer the case.
4113 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4114 and queue runs started by the daemon processed all messages. This has
4115 been fixed so that -R and -S can now usefully be given with -q<time>.
4117 PH/40 Import PCRE release 6.7 (fixes some bugs).
4119 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4121 PH/42 Give an error if -q is specified more than once.
4123 PH/43 Renamed the variables $interface_address and $interface_port as
4124 $received_ip_address and $received_port, to make it clear that these
4125 values apply to message reception, and not to the outgoing interface when
4126 a message is delivered. (The old names remain recognized, of course.)
4128 PH/44 There was no timeout on the connect() call when using a Unix domain
4129 socket in the ${readsocket expansion. There now is.
4131 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4132 be meaningful with "accept".
4134 SC/01 Eximstats V1.43
4135 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4137 SC/02 Eximstats V1.44
4138 Use a glob alias rather than an array ref in the generated
4139 parser. This improves both readability and performance.
4141 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4142 Collect SpamAssassin and rejection statistics.
4143 Don't display local sender or destination tables unless
4144 there is data to show.
4145 Added average volumes into the top table text output.
4147 SC/04 Eximstats V1.46
4148 Collect data on the number of addresses (recipients)
4149 as well as the number of messages.
4151 SC/05 Eximstats V1.47
4152 Added 'Message too big' to the list of mail rejection
4153 reasons (thanks to Marco Gaiarin).
4155 SC/06 Eximstats V1.48
4156 Mainlog lines which have GMT offsets and are too short to
4157 have a flag are now skipped.
4159 SC/07 Eximstats V1.49 (Alain Williams)
4160 Added the -emptyok flag.
4162 SC/08 Eximstats V1.50
4163 Fixes for obtaining the IP address from reject messages.
4165 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4166 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4167 whitespace changes from 4.64-PH/27
4169 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4172 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4173 are found, allow negative numbers in numeric criteria)
4175 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4177 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4178 to match changes made in 4.64-PH/43
4180 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4182 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4183 "domain" is actually an IP address, and force a failure. This locks out
4184 those revolvers/nameservers that support "A-for-A" lookups, in
4185 contravention of the specifications.
4187 PH/48 When a host name was looked up from an IP address, and the subsequent
4188 forward lookup of the name timed out, the host name was left in
4189 $sender_host_name, contrary to the specification.
4191 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4192 restricted to single-key lookups, Exim was not diagnosing an error if
4193 * or *@ was used with a query-style lookup.
4195 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4197 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4198 long ago, but noone interested enough thought of it. Let's just say that
4199 the "1.1" means that there are some new functions that weren't there at
4200 some point in the past.
4202 PH/51 Error processing for expansion failure of helo_data from an smtp
4203 transport during callout processing was broken.
4205 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4206 tested/used via the -bh/-bhc/-bs options.
4208 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4209 bug, fixed in subsequent PCRE releases).
4211 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4212 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4214 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4220 SC/01 Use a glob alias rather than an array ref in eximstats generated
4221 parser. This improves both readability and performance.
4223 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4224 Don't display local sender or destination tables in eximstats unless
4225 there is data to show.
4226 Added average volumes into the eximstats top table text output.
4228 SC/03 Collect data on the number of addresses (recipients) as well
4229 as the number of messages in eximstats.
4231 TF/01 Correct an error in the documentation for the redirect router. Exim
4232 does (usually) call initgroups() when daemonizing.
4234 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4235 with consistent privilege compared to when running as a daemon.
4237 TF/03 Note in the spec that $authenticated_id is not set for local
4238 submissions from trusted users.
4240 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4241 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4243 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4244 by adding some example configuration directives to the default
4245 configuration file. A little bit of work is required to uncomment the
4246 directives and define how usernames and passwords are checked, but
4247 there is now a framework to start from.
4249 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4250 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4251 without this. I don't know how relevant this is to other LDAP libraries.
4253 PH/02 Add the verb name to the "unknown ACL verb" error.
4255 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4257 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4259 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4260 directory not expanded when it should be if an expanded home directory
4261 was set for the address (which is overridden by the transport).
4263 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4266 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4267 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4268 because it is too late at that time, and has no effect.
4270 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4271 security issue with \' (bugzilla #107). I could not use the
4272 PQescapeStringConn() function, because it needs a PGconn value as one of
4275 PH/08 When testing addresses using -bt, indicate those final addresses that
4276 are duplicates that would not cause an additional delivery. At least one
4277 person was confused, thinking that -bt output corresponded to deliveries.
4278 (Suppressing duplicates isn't a good idea as you lose the information
4279 about possibly different redirections that led to the duplicates.)
4281 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4282 systems where poll() doesn't work, in particular OS X.
4284 PH/10 Added more information to debugging output for retry time not reached.
4286 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4287 operations in malware.c.
4289 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4292 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4293 syslog_duplication set false, log lines that would normally be written
4294 both the the main log and to the reject log were not written to syslog at
4297 PH/14 In the default configuration, change the use of "message" in ACL warn
4298 statements to "add_header".
4300 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4301 not followed by a command (e.g. "seen endif").
4303 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4304 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4307 PH/17 Added extra conditions to the default value of delay_warning_condition
4311 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4312 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4313 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4316 The Auto-Submitted: and various List- headers are standardised, whereas I
4317 don't think Precedence: ever was.
4319 PH/18 Refactored debugging code in route_finduser() to show more information,
4320 in particular, the error code if getpwnam() issues one.
4322 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4323 This is apparently needed in addition to the PH/07 change above to avoid
4324 any possible encoding problems.
4326 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4327 but not after initializing Perl.
4329 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4330 output them only if debugging. By default they are written stderr,
4331 apparently, which is not desirable.
4333 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4336 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4339 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4341 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4342 authenticated or an ident call has been made. Suppress the default
4343 values for $authenticated_id and $authenticated_sender (but permit -oMai
4344 and -oMas) when testing with -bh.
4346 PH/24 Re-jigged the order of the tests in the default configuration so that the
4347 tests for valid domains and recipients precede the DNS black list and CSA
4348 tests, on the grounds that those ones are more expensive.
4350 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4351 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4352 This bug exists in every version of Exim that I still have, right back to
4355 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4356 However, an attempt to turn on case-sensitivity in a regex key by
4357 including (?-i) didn't work because the subject string was already
4358 lowercased, and the effects were non-intuitive. It turns out that a
4359 one-line patch can be used to allow (?-i) to work as expected.
4365 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4366 other effects) broke the use of negated acl sub-conditions.
4368 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4371 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4372 "Deny" causes Exim to reject the incoming connection with a 554 error.
4373 Unfortunately, if there is a major crisis, such as a disk failure,
4374 tcp-wrappers gives "deny", whereas what one would like would be some
4375 kind of temporary error. A kludge has been added to help with this.
4376 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4377 554 error is used if errno is still zero or contains ENOENT (which occurs
4378 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4381 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4383 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4384 errors. Otherwise a message that provokes a temporary error (when other
4385 messages do not) can cause a whole host to time out.
4387 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4388 addresses were routed directly to files or pipes from a redirect router.
4389 File deliveries just didn't batch; pipe deliveries might have suffered
4392 PH/06 A failure to get a lock for a hints database would erroneously always say
4393 "Failed to get write lock", even when it was really a read lock.
4395 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4396 of 0600. This has been changed to use the value of the lockfile_mode
4397 option (which defaults to 0600).
4399 PH/08 Applied small patch from the Sieve maintainer.
4401 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4402 folder from quota calculations, a direct delivery into this folder messed
4403 up the contents of the maildirsize file. This was because the regex was
4404 used only to exclude .Trash (or whatever) when the size of the mailbox
4405 was calculated. There was no check that a delivery was happening into an
4406 excluded directory. This bug has been fixed by ignoring all quota
4407 processing for deliveries into excluded directories.
4409 PH/10 Added the maildirfolder_create_regex option to appendfile.
4415 PH/01 The code for finding all the local interface addresses on a FreeBSD
4416 system running IPv6 was broken. This may well have applied to all BSD
4417 systems, as well as to others that have similar system calls. The broken
4418 code found IPv4 interfaces correctly, but gave incorrect values for the
4419 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4420 that it would not match correctly against @[] and not recognize the IPv6
4423 PH/02 The ipliteral router was not recognizing addresses of the form user@
4424 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4426 PH/03 Added disable_ipv6.
4428 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4429 decoded form, because it is most often used to construct To: headers
4430 lines in autoreplies, and the decoded form may well be syntactically
4431 invalid. However, $reply_address has leading white space removed, and all
4432 newlines turned into spaces so that the autoreply transport does not
4435 PH/05 If group was specified without a user on a router, and no group or user
4436 was specified on a transport, the group from the router was ignored.
4438 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4439 for visible compile-time settings that can be used to change these
4440 numbers, for those that want even more. Backwards compatibility with old
4441 spool files has been maintained. However, going back to a previous Exim
4442 release will lost any variables that are in spool files.
4444 PH/07 Two small changes when running in the test harness: increase delay when
4445 passing a TCP/IP connection to a new process, in case the original
4446 process has to generate a bounce, and remove special handling of
4447 127.0.0.2 (sic), which is no longer necessary.
4449 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4450 be the same on different OS.
4452 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4455 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4456 whether --show-vars was specified or not
4458 JJ/02 exipick: Added support for new ACL variable spool format introduced
4461 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4462 syntactically invalid From: or Reply-to: line, and a filter used this to
4463 generate an autoreply, and therefore failed to obtain an address for the
4464 autoreply, Exim could try to deliver to a non-existent relative file
4465 name, causing unrelated and misleading errors. What now happens is that
4466 it logs this as a hard delivery error, but does not attempt to create a
4469 PH/11 The exinext utility has a -C option for testing purposes, but although
4470 the given file was scanned by exinext itself; it wasn't being passed on
4471 when Exim was called.
4473 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4474 an end-of-file indication when reading a command response.
4476 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4477 compiled. In many other places in Exim, IPv6 addresses are always
4478 recognized, so I have changed this. It also means that IPv4 domain
4479 literals of the form [IPV4:n.n.n.n] are now always recognized.
4481 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4482 used if the router is not running as root, for example, when verifying at
4483 ACL time, or when using -bh. The debugging output from this situation was
4484 non-existent - all you got was a failure to exec. I have made two
4487 (a) Failures to set uid/gid, the current directory, or a process leader
4488 in a subprocess such as that created by queryprogram now generate
4489 suitable debugging output when -d is set.
4491 (b) The queryprogram router detects when it is not running as root,
4492 outputs suitable debugging information if -d is set, and then runs
4493 the subprocess without attempting to change uid/gid.
4495 PH/15 Minor change to Makefile for building test_host (undocumented testing
4498 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4499 additional section of a DNS packet that returns MX or SRV records.
4500 Instead, it always explicitly searches for A/AAAA records. This avoids
4501 major problems that occur when a DNS server includes only records of one
4502 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4503 fixed another bug: if SRV records were looked up and the corresponding
4504 address records were *not* found in the additional section, the port
4505 values from the SRV records were lost.
4507 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4508 using the correct key (the original address) when searching the retry
4509 rules in order to find which one to use for generating the retry hint.
4511 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4512 adding the default one. Similarly, if it contains a Reply-To: header, the
4513 errors_reply_to option, if set, is not used.
4515 PH/19 When calculating a retry time, Exim used to measure the "time since
4516 failure" by looking at the "first failed" field in the retry record. Now
4517 it does not use this if it is later than than the arrival time of the
4518 message. Instead it uses the arrival time. This makes for better
4519 behaviour in cases where some deliveries succeed, thus re-setting the
4520 "first failed" field. An example is a quota failure for a huge message
4521 when small messages continue to be delivered. Without this change, the
4522 "time since failure" will always be short, possible causing more frequent
4523 delivery attempts for the huge message than are intended.
4524 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4526 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4527 $1, $2, $3) because the numerical variables can be reset during some
4528 expansion items (e.g. "match"), thereby losing the authentication data.
4530 PH/21 Make -bV show the size of off_t variables so that the test suite can
4531 decide whether to run tests for quotas > 2G.
4533 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4534 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4535 filecount value is greater than 2G or if a quota value is greater than 2G
4536 on a system where the size of off_t is not greater than 4, a panic error
4539 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4540 never match. The debug and -bh output now contains an explicit error
4541 message indicating a malformed IPv4 address or mask.
4543 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4544 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4545 PH/23 above applies.
4547 PH/25 Do not write to syslog when running in the test harness. The only
4548 occasion when this arises is a failure to open the main or panic logs
4549 (for which there is an explicit test).
4551 PH/26 Added the /no_tell option to "control=freeze".
4553 PH/27 If a host name lookup failed very early in a connection, for example, if
4554 the IP address matched host_lookup and the reverse lookup yielded a name
4555 that did not have a forward lookup, an error message of the form "no IP
4556 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4557 could be logged. Now it outputs the IP address instead of "NULL".
4559 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4560 allows the sender and the authenticated sender to be set when
4561 submitting a message from within Exim. Since child_open_exim() is
4562 documented for local_scan(), the new function should be too.
4564 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4565 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4566 results in an empty string is now treated as unset.
4568 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4570 PH/31 Added sender_verify_fail logging option.
4572 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4573 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4574 tidied the source and removed it altogether.
4576 PH/33 When a queue run was abandoned because the load average was too high, a
4577 log line was always written; now it is written only if the queue_run log
4578 selector is set. In addition, the log line for abandonment now contains
4579 information about the queue run such as the pid. This is always present
4580 in "start" and "stop" lines but was omitted from the "abandon" line.
4582 PH/34 Omit spaces between a header name and the colon in the error message that
4583 is given when verify = headers_syntax fails (if there are lots of them,
4584 the message gets confusing).
4586 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4587 names, as there are now some PTR records that contain slashes. This check
4588 is only to protect against broken name servers that fall over on strange
4589 characters, so the fact that it applies to all lookups doesn't matter.
4591 PH/36 Now that the new test suite is complete, we can remove some of the
4592 special code in Exim that was needed for the old test suite. For example,
4593 sorting DNS records because real resolvers return them in an arbitrary
4594 order. The new test suite's fake resolver always returns records in the
4597 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4598 bounces) except when queue_only is set, to avoid logging races between
4599 the different processes.
4601 PH/38 Panic-die if .include specifies a non-absolute path.
4603 PH/39 A tweak to the "H" retry rule from its user.
4605 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4606 a label. They prevented compilation on older perls.
4608 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4609 a warning to be raised on newish perls.
4611 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4612 on queue. Changes to match documented behaviour of showing count of
4613 messages matching specified criteria.
4615 PH/40 Changed the default ident timeout from 30s to 5s.
4617 PH/41 Added support for the use of login_cap features, on those BSD systems
4618 that have them, for controlling the resources used by pipe deliveries.
4620 PH/42 The content-scanning code uses fopen() to create files in which to put
4621 message data. Previously it was not paying any attention to the mode of
4622 the files. Exim runs with umask(0) because the rest of the code creates
4623 files with open(), and sets the required mode explicitly. Thus, these
4624 files were ending up world-writeable. This was not a big issue, because,
4625 being within the spool directory, they were not world-accessible. I have
4626 created a function called modefopen, which takes an additional mode
4627 argument. It sets umask(777), creates the file, chmods it to the required
4628 mode, then resets the umask. All the relevant calls to fopen() in the
4629 content scanning code have been changed to use this function.
4631 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4632 to 24 hours. This avoids potential overflow problems when processing G
4633 and H retry rules. I suspect nobody ever tinkers with this value.
4635 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4637 PH/45 When the plaintext authenticator is running as a client, the server's
4638 challenges are checked to ensure they are valid base64 strings. By
4639 default, the authentication attempt is cancelled if an invalid string is
4640 received. Setting client_ignore_invalid_base64 true ignores these errors.
4641 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4642 they are received. Thus, the responses can be made to depend on the
4643 challenges. If an invalid string is ignored, an empty string is placed in
4646 PH/46 Messages that are created by the autoreply transport now contains a
4647 References: header, in accordance with RFCs 2822 and 3834.
4649 PH/47 Added authenticated_sender_force to the smtp transport.
4651 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4653 PH/49 Installed latest patch from the Sieve maintainer.
4655 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4656 was also set, the appendfile transport was unnecessarily scanning a
4657 directory of message files (e.g. for maildir delivery) to find the count
4658 of files (along with the size), even though it did not need this
4659 information. It now does the scan only if it needs to find either the
4660 size of the count of files.
4662 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4664 PH/52 Two bugs concerned with error handling when the smtp transport is
4667 (i) Exim was not creating retry information for temporary errors given
4668 for individual recipients after the DATA command when the smtp transport
4669 was used in LMTP mode. This meant that they could be retried too
4670 frequently, and not timed out correctly.
4672 (ii) Exim was setting the flag that allows error details to be returned
4673 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4674 recipients that were returned after the DATA command.
4676 PH/53 This is related to PH/52, but is more general: for any failing address,
4677 when detailed error information was permitted to be returned to the
4678 sender, but the error was temporary, then after the final timeout, only
4679 "retry timeout exceeded" was returned. Now it returns the full error as
4680 well as "retry timeout exceeded".
4682 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4683 do this, and (what is worse) MTAs that accept it.
4685 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4686 will now be deprecated.
4688 PH/56 New os.c-cygwin from the Cygwin maintainer.
4690 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4691 formats (previously only available in exim formats via -bpr, -bpru,
4692 and -bpra. Now also available in native and exiqgrep formats)
4694 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4695 with very large, slow to parse queues
4697 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4699 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4701 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4702 responses to authentication challenges, though it was showing the
4703 challenges; (ii) I've removed the CR characters from the debug output for
4706 PH/58 Allow for the insertion of a newline as well as a space when a string
4707 is turned into more than one encoded-word during RFC 2047 encoding. The
4708 Sieve code now uses this.
4710 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4711 data_4xx, lost_connection, tls_required.
4713 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4714 message was being sent as an SMTP response.
4716 PH/61 Add -l and -k options to exicyclog.
4718 PH/62 When verifying, if an address was redirected to one new address, so that
4719 verification continued, and the new address failed or deferred after
4720 having set something in $address_data, the value of $address_data was not
4721 passed back to the ACL. This was different to the case when no
4722 redirection occurred. The value is now passed back in both cases.
4724 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4725 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4726 use login_cap.h, so on its own it isn't the distinguishing feature. The
4727 new name refers directly to the setclassresources() function.
4729 PH/65 Added configuration files for NetBSD3.
4731 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4733 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4734 is preferred over IPv4.
4736 PH/68 The bounce_return_message and bounce_return_body options were not being
4737 honoured for bounces generated during the reception of non-SMTP messages.
4738 In particular, this applied to messages rejected by the ACL. This bug has
4739 been fixed. However, if bounce_return_message is true and bounce_return_
4740 body is false, the headers that are returned for a non-SMTP message
4741 include only those that have been read before the error was detected.
4742 (In the case of an ACL rejection, they have all been read.)
4744 PH/69 The HTML version of the specification is now built in a directory called
4745 spec_html instead of spec.html, because the latter looks like a path with
4746 a MIME-type, and this confuses some software.
4748 PH/70 Catch two compiler warnings in sieve.c.
4750 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4751 function verify_get_ident() calls ip_connect() to connect a socket, but
4752 if the "connect()" function timed out, ip_connect() used to close the
4753 socket. However, verify_get_ident() also closes the socket later, and in
4754 between Exim writes to the log, which may get opened at this point. When
4755 the socket was closed in ip_connect(), the log could get the same file
4756 descriptor number as the socket. This naturally causes chaos. The fix is
4757 not to close the socket in ip_connect(); the socket should be closed by
4758 the function that creates it. There was only one place in the code where
4759 this was missing, in the iplookup router, which I don't think anybody now
4760 uses, but I've fixed it anyway.
4762 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4763 well as to direct DNS lookups. Otherwise the handling of names in host
4764 lists is inconsistent and therefore confusing.
4770 PH/01 Two changes to the default runtime configuration:
4772 (1) Move the checks for relay_from_hosts and authenticated clients from
4773 after to before the (commented out) DNS black list checks.
4775 (2) Add control=submission to the relay_from_hosts and authenticated
4776 clients checks, on the grounds that messages accepted by these
4777 statements are most likely to be submissions.
4779 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4781 (1) Generate an error if the third argument for the ${prvs expansion is
4784 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4787 (3) Reset the variables that are obtained from the first argument of
4788 ${prvscheck and used in the second argument before leaving the code,
4789 because their memory is reclaimed, so using them afterwards may do
4792 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4793 one (it's much easier than Tom thought :-).
4795 (5) Because of (4), we can now allow for the use of $prvscheck_result
4796 inside the third argument.
4798 PH/03 For some reason, the default setting of PATH when running a command from
4799 a pipe transport was just "/usr/bin". I have changed it to
4802 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4803 anything to be listed in the output from -bV.
4805 PH/05 When a filter generated an autoreply, the entire To: header line was
4806 quoted in the delivery log line, like this:
4808 => >A.N.Other <ano@some.domain> <original@ddress> ...
4810 This has been changed so that it extracts the operative address. There
4811 may be more than one such address. If so, they are comma-separated, like
4814 => >ano@some.domain,ona@other.domain <original@ddress> ...
4816 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4817 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4818 not being looked up in the rDNS to get a host name, Exim was showing the
4819 IP address twice in Received: lines, even though the IP addresses were
4820 identical. For example:
4822 Received: from [1.2.3.4] (helo=[1.2.3.4])
4824 However, if the real host name was known, it was omitting the HELO data
4825 if it matched the actual IP address. This has been tidied up so that it
4826 doesn't show the same IP address twice.
4828 PH/07 When both +timestamp and +memory debugging was on, the value given by
4829 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4830 called by the debug printing, thereby overwriting the timestamp buffer.
4831 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4833 PH/08 When the original message was included in an autoreply transport, it
4834 always said "this is a copy of the message, including all the headers",
4835 even if body_only or headers_only was set. It now gives an appropriate
4838 PH/09 Applied a patch from the Sieve maintainer which:
4840 o fixes some comments
4841 o adds the (disabled) notify extension core
4842 o adds some debug output for the result of if/elsif tests
4843 o points to the current vacation draft in the documentation
4844 and documents the missing references header update
4848 o fixes a bug in processing the envelope test (when testing
4849 multiple envelope elements, the last element determined the
4852 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4853 Electronic Mail") by including:
4855 Auto-submitted: auto-generated
4857 in the messages that it generates (bounce messages and others, such as
4858 warnings). In the case of bounce messages for non-SMTP messages, there was
4859 also a typo: it was using "Auto_submitted" (underscore instead of
4860 hyphen). Since every message generated by Exim is necessarily in response
4861 to another message, thes have all been changed to:
4863 Auto-Submitted: auto-replied
4865 in accordance with these statements in the RFC:
4867 The auto-replied keyword:
4869 - SHOULD be used on messages sent in direct response to another
4870 message by an automatic process,
4872 - MUST NOT be used on manually-generated messages,
4874 - MAY be used on Delivery Status Notifications (DSNs) and Message
4875 Disposition Notifications (MDNs),
4877 - MUST NOT be used on messages generated by automatic or periodic
4878 processes, except for messages which are automatic responses to
4881 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4882 to the default Received: header definition.
4884 PH/12 Added log selector acl_warn_skipped (default on).
4886 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4887 variables because (a) they are in the wrong storage pool and (b) even if
4888 they were copied, it wouldn't work properly because of the caching.
4890 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4891 checking when decoding. Apparently there are clients that generate
4892 overlong encoded strings. Why am I not surprised?
4894 PH/15 If the first argument of "${if match_address" was not empty, but did not
4895 contain an "@" character, Exim crashed. Now it writes a panic log message
4896 and treats the condition as false.
4898 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4900 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4901 extension "envelope-auth". The code is finished and in agreement with
4902 other implementations, but there is no documentation so far and in fact,
4903 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4904 not changing the active code.
4906 Print executed "if" and "elsif" statements when debugging is used. This
4907 helps a great deal to understand what a filter does.
4909 Document more things not specified clearly in RFC3028. I had all this
4910 sorted out, when out of a sudden new issues came to my mind. Oops."
4912 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4915 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4916 canonical form (as documented). However, after a host name lookup from
4917 the IP address, check_host() was doing a simple string comparison with
4918 addresses acquired from the DNS when checking that the found name did
4919 have the original IP as one of its addresses. Since any found IPv6
4920 addresses are likely to be in abbreviated form, the comparison could
4921 fail. Luckily, there already exists a function for doing the comparison
4922 by converting both addresses to binary, so now that is used instead of
4923 the text comparison.
4925 PH/20 There was another similar case to PH/19, when a complete host name was
4926 given in a host list; looking up its IP address could give an abbreviated
4927 form, whereas the current host's name might or might not be abbreviated.
4928 The same fix has been applied.
4934 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4935 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4938 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4939 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4941 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4943 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4944 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4945 Also, for "Auto-Submitted", treat anything other than "no" as
4946 non-personal, in accordance with RFC 3834. (Previously it treated
4947 anything starting "auto-" as non-personal.)
4949 TF/01 The control=submission/name=... option had a problem with syntax
4950 errors if the name included a slash character. The /name= option
4951 now slurps the rest of the string, so it can include any characters
4952 but it must come last in the list of options (after /sender_retain
4955 PH/05 Some modifications to the interface to the fake nameserver for the new
4963 TK/01 Added the "success_on_redirect" address verification option. See
4964 NewStuff for rationale and an example.
4966 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4968 PH/02 Patch to exigrep to allow it to work on syslog lines.
4970 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4971 fread() to skip over the body file's header line, because in Cygwin the
4972 header line is locked and is inaccessible.
4974 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4975 co-exist for some time) to make it clear that it is the Exim ID that is
4976 referenced, not the Message-ID: header line.
4978 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4979 string_format() function, because snprintf() does not exist on all
4982 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4983 ${stat: expansion item.
4985 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4986 protocol synchronization error", to keep the pedants happy.
4988 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4989 well as for IRIX systems, when gcc is being used. See the host.c source
4992 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4994 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4997 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4998 transport and to the smtp transport in LMTP mode.
5000 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5002 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5003 run a filter in a subprocess. This could lead to confusion in subsequent
5004 lookups in the parent process. There should also be a search_tidyup() at
5005 the end of the subprocess.
5007 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5008 only if the host matched helo_try_verify_hosts, which caused the
5009 verification to occur when the EHLO/HELO command was issued. The ACL just
5010 tested the remembered result. Now, if a previous verification attempt has
5011 not happened, "verify = helo" does it there and then.
5013 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5015 TK/03 Fix log output including CR from clamd.
5017 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5018 exist provoked a memory error which could cause a segfault.
5020 PH/15 Installed PCRE 6.2
5022 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5024 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5025 of the problem. Specifically, suggested +O2 rather than +O1 for the
5028 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5030 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5031 record, the port was not correctly logged when the outgoing_port log
5032 selector was set (it logged the transort's default port).
5034 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5035 fallback_hosts, and "hosts" in the smtp transport.
5037 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5038 host errors such as "Connection refused".
5040 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5041 authentication with radiusclient 0.4.9:
5043 - Error returned from rc_read_config was caught wrongly
5044 - Username/password not passed on to radius server due to wrong length.
5046 The presumption is that some radiusclient API changes for 4.51/PH/17
5047 were not taken care of correctly. The code is still untested by me (my
5048 Linux distribution still has 0.3.2 of radiusclient), but it was
5049 contributed by a Radius user.
5051 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5052 expanding the "port" option of the smtp transport.
5054 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5055 while reading a MIME header. Thanks to Tom Hughes for a patch.
5057 PH/24 Include config.h inside local_scan.h so that configuration settings are
5060 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5061 that in an ACL for RCPT (for example), you can examine exactly what was
5064 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5065 commands, but it was not correctly comparing the address with the actual
5066 client host address. Thus, it would show the EHLO address in Received:
5067 header lines when this was not necessary.
5069 PH/27 Added the % operator to ${eval:}.
5071 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5072 it should be ignoring failures (because with -C, for example, it has lost
5073 privilege). It wasn't ignoring creation failures other than "already
5076 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5079 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5080 because an input line was too long, either on its own, or by virtue of
5081 too many continuations, the temporary file was not being removed, and the
5082 return code was incorrect.
5084 PH/31 Missing "BOOL" in function definition in filtertest.c.
5086 PH/32 Applied Sieve patches from the maintainer.
5088 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5090 PH/33 Added "verify = not_blind".
5092 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5093 Local/Makefile (with some defaults set). These are used in built scripts
5094 such as exicyclog, but they have never been used in the exim_install
5095 script (though there are many overriding facilities there). I have
5096 arranged that the exim_install script now takes note of these two
5099 PH/35 Installed configuration files for Dragonfly.
5101 PH/36 When a locally submitted message by a trusted user did not contain a
5102 From: header, and the sender address was obtained from -f or from an SMTP
5103 MAIL command, and the trusted user did not use -F to supply a sender
5104 name, $originator_name was incorrectly used when constructing a From:
5105 header. Furthermore, $originator_name was used for submission mode
5106 messages from external hosts without From: headers in a similar way,
5107 which is clearly wrong.
5109 PH/37 Added control=suppress_local_fixups.
5111 PH/38 When log_selector = +received_sender was set, and the addition of the
5112 sender made the log line's construction buffer exactly full, or one byte
5113 less than full, an overflow happened when the terminating "\n" was
5116 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5117 when the result of a list match is failure because a DNS lookup failed.
5119 PH/40 RM_COMMAND is now used in the building process.
5121 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5122 the "build-* directories that it finds.
5124 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5125 address in a domain literal was a prefix of an interface address.
5127 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5128 when verifying a sender address, unless rewrite_headers is false.
5130 PH/44 Wrote a long comment about why errors_to addresses are verified as
5131 recipients, not senders.
5133 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5134 the ratelimit ACL was added.
5136 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5138 PH/46 Added extra information about PostgreSQL errors to the error string.
5140 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5141 avoiding the need to install special zones in a real server. This is
5142 backwards compatible; if it can't find the fake resolver, it drops back.
5143 Thus, both old and new test suites can be run.
5145 TF/02 Added util/ratelimit.pl
5147 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5148 clock is set back in time.
5150 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5151 Candler <B.Candler@pobox.com>.
5153 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5154 OK for addresses that are the result of redirections.
5156 PH/48 A number of further additions for the benefit of the new test suite,
5157 including a fake gethostbyname() that interfaces to the fake DNS resolver
5160 TF/06 The fix for widen_domains has also been applied to qualify_single and
5161 search_parents which are the other dnslookup options that can cause
5164 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5167 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5169 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5170 service" if it encountered an MX record with an empty target hostname.
5171 The message is now "an MX or SRV record indicated no SMTP service".
5173 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5174 if the DNS of the sending site is misconfigured. This is quite a
5175 common situation. This change restores the behaviour of treating a
5176 helo verification defer as a failure.
5178 PH/51 If self=fail was set on a router, the bounce message did not include the
5179 actual error message.
5185 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5187 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5188 command itself ended in error, the underlying message about the transport
5189 filter timeout was being overwritten with the pipe command error. Now the
5190 underlying error message should be appended to the second error message.
5192 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5194 PH/02 Reduce the amount of output that "make" produces by default. Full output
5195 can still be requested.
5197 PH/03 The warning log line about a condition test deferring for a "warn" verb
5198 was being output only once per connection, rather than after each
5199 occurrence (because it was using the same function as for successful
5200 "warn" verbs). This seems wrong, so I have changed it.
5202 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5203 it should not have, which might have caused a crash in the right
5204 circumstances, but probably never did.
5206 PH/04 Installed a modified version of Tony Finch's patch to make submission
5207 mode fix the return path as well as the Sender: header line, and to
5208 add a /name= option so that you can make the user's friendly name appear
5211 TF/03 Added the control = fakedefer ACL modifier.
5213 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5214 Mark Lowes for thorough testing.
5216 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5218 TK/03 Merged latest SRS patch from Miles Wilton.
5220 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5221 with the definition in sysexits.h (which is #included earlier).
5222 Fortunately, Exim does not actually use EX_OK. The code used to try to
5223 preserve the sysexits.h value, by assuming that macro definitions were
5224 scanned for macro replacements. I have been disabused of this notion,
5225 so now the code just undefines EX_OK before #including unistd.h.
5227 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5228 in the smtp transport. When a block could not be written in a single
5229 write() function, the timeout was being re-applied to each part-write.
5230 This seems wrong - if the receiver was accepting one byte at a time it
5231 would take for ever. The timeout is now adjusted when this happens. It
5232 doesn't have to be particularly precise.
5234 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5235 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5237 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5238 without a domain if the check to <postmaster@domain> fails.
5240 SC/01 Eximstats: added -xls and the ability to specify output files
5241 (patch written by Frank Heydlauf).
5243 SC/02 Eximstats: use FileHandles for outputting results.
5245 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5247 SC/04 Eximstats: fixed display of large numbers with -nvr option
5249 SC/05 Eximstats: fixed merging of reports with empty tables.
5251 SC/06 Eximstats: added the -include_original_destination flag
5253 SC/07 Eximstats: removed tabs and trailing whitespace.
5255 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5257 TK/06 MBOX spool code: Add real "From " MBOX separator line
5258 so the .eml file is really in mbox format (even though
5259 most programs do not really care). Patch from Alex Miller.
5261 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5262 The latter is generated from $received_to and is only set if the
5263 message has one envelope recipient. SA can use these headers,
5264 obviously out-of-the-box. Patch from Alex Miller.
5266 PH/08 The ${def test on a variable was returning false if the variable's
5267 value was "0", contrary to what the specification has always said!
5268 The result should be true unless the variable is empty.
5270 PH/09 The syntax error of a character other than { following "${if
5271 def:variable_name" (after optional whitespace) was not being diagnosed.
5272 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5273 accidental colon was present, for example, could give incorrect results.
5275 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5276 result is used (not including appendfile, where other changes are about
5279 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5280 This involved changing a lot of size variables from int to off_t. It
5281 should work with maildirs and everything.
5283 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5284 spamd dying while we are connected to it.
5286 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5289 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5290 function for BDB 4.3.
5292 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5294 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5295 the sources for PCRE have been re-arranged and more files are now
5298 PH/15 The code I had for printing potentially long long variables in PH/11
5299 above was not the best (it lost precision). The length of off_t variables
5300 is now inspected at build time, and an appropriate printing format (%ld
5301 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5302 to be "long long int" or "long int". This is needed for the internal
5303 formatting function string_vformat().
5305 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5306 the configuration file to be ":syslog", then the script "guesses" where
5307 the logs files are, rather than using the compiled in default. In our
5308 case the guess is not the same as the compiled default, so the script
5309 suddenly stopped working when I started to use syslog. The patch checks
5310 to see if log_file_path is "". If so, it attempts to read it from exim
5311 with no configuration file to get the compiled in version, before it
5312 falls back to the previous guessing code."
5314 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5315 implementing BATV in an Exim configuration. See NewStuff for the gory
5318 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5319 Makefile that are specific to HP-UX.
5321 PH/18 If the "use_postmaster" option was set for a recipient callout together
5322 with the "random" option, the postmaster address was used as the MAIL
5323 FROM address for the random test, but not for the subsequent recipient
5324 test. It is now used for both.
5326 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5327 patch removes a few documentation additions to RFC 3028, because the
5328 latest draft now contains them. It adds the new en;ascii-case comparator
5329 and a new error check for 8bit text in MIME parts. Comparator and
5330 require names are now matched exactly. I enabled the subaddress
5331 extension, but it is not well tested yet (read: it works for me)."
5333 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5334 rework some of the code of TK/09 above to avoid the hardwired use of
5335 "%lld" and "long long". Replaced the call to snprintf() with a call to
5338 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5339 records point to non-existent hosts", "retry timeout exceeded", and
5340 "retry time not reached for any host after a long failure period".
5342 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5343 experimental DomainKeys support:
5345 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5346 (2) On an error such as an illegally used "control", the wrong name for
5347 the control was given.
5349 These problems did NOT occur unless DomainKeys support was compiled.
5351 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5353 PH/24 Added ${if match_ip condition.
5355 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5356 databases so that it will be absolutely obvious if a crash occurs in the
5357 DB library. This is a regular occurrence (often caused by mis-matched
5360 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5361 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5362 picked up on a user's system that detects such things. There doesn't seem
5363 to be a gcc warning option for this - only an attribute that has to be
5364 put on the function's prototype. It seems that in Fedora Core 4 they have
5365 set this on a number of new functions. No doubt there will be more in due
5368 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5369 specify a transport. However, if an address that was verified by such a
5370 router was the subject of a callout, Exim crashed because it tried to
5371 read the rcpt_include_affixes from the non-existent transport. Now it
5372 just assumes that the setting of that option is false. This bug was
5373 introduced by 4.51/PH/31.
5375 PH/28 Changed -d+all to exclude +memory, because that information is very
5376 rarely of interest, but it makes the output a lot bigger. People tend to
5377 do -d+all out of habit.
5379 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5380 code in os-type was giving problems when libc.so lives in lib64, like on
5383 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5384 aren't the modern standard, and it seems that some systems' include files
5385 don't always have them. Exim was already checking for some of the newer
5386 ones like T_AAAA, and defining it itself. I've added checks for all the
5387 record types that Exim uses.
5389 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5390 not automatically generating a new one, as it is supposed to. This
5391 prevented TLS from working. If the file did exist, but contained invalid
5392 data, a new version was generated, as expected. It was only the case of a
5393 non-existent file that was broken.
5395 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5396 with a change in libdomainkeys > 0.64.
5398 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5399 from DNS. If the selector record carries the flag, it now has
5400 precedence over the domain-wide flag.
5402 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5404 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5405 the use of a port name that isn't defined in /etc/services) occurred, the
5406 message was deferred as in a normal delivery, and thus remained on the
5407 spool, instead of being failed because of the mua_wrapper setting. This
5408 is now fixed, and I tidied up some of the mua_wrapper messages at the
5411 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5412 the messages in a hash of arrays rather than using individual hashes.
5413 This is a bit cleaner and results in dramatic memory savings, albeit
5414 at a slight CPU cost.
5416 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5417 as requested by Marc Sherman.
5419 SC/10 Eximstats: added histograms for user specified patterns as requested
5422 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5424 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5425 fopen() in the content-scanning modules that did not already have it.
5431 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5432 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5434 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5436 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5438 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5439 to test Sieve filters that use "vacation".
5441 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5442 that changes the way the GnuTLS parameters are stored in the cache file.
5443 The new format can be generated externally. For backward compatibility,
5444 if the data in the cache doesn't make sense, Exim assumes it has read an
5445 old-format file, and it generates new data and writes a new file. This
5446 means that you can't go back to an older release without removing the
5449 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5450 work if there are any delivery delays because "one_time" forces the
5451 parent to be marked "delivered", so its unseen clone is never tried
5452 again. For this reason, Exim now forbids the simultaneous setting of
5455 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5456 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5457 entry if you want to know the details. The fix, however, neglected to
5458 consider the case when local delivery batching is involved. The test for
5459 "previously delivered" was not happening when checking to see if an
5460 address could be batched with a previous (undelivered) one; under
5461 certain circumstances this could lead to multiple deliveries to the same
5464 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5465 in its include files, and this causes problems building Exim.
5467 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5468 header_syntax) but Exim was just ignoring anything given after a slash.
5469 In particular, this caused confusion with an attempt to use "verify =
5470 reverse_host_lookup/defer_ok". An error is now given when options are
5471 supplied for verify items that do not have them. (Maybe reverse_host_
5472 lookup should have a defer_ok option, but that's a different point.)
5474 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5475 defined by RFC 821) to 2048, because there were problems with some AUTH
5476 commands, and RFC 1869 says the size should be increased for extended
5477 SMTP commands that take arguments.
5479 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5482 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5483 "unknown" error; now it says that the functionality isn't in the binary.
5485 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5486 an address' error message when a string expansion fails (syntax or
5487 whatever). Otherwise the password may appear in the log. Following change
5488 PH/42 below, there is no longer a chance of it appearing in a bounce
5491 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5493 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5494 of its addresses was ever tried. (Bugzilla bug #2.)
5496 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5497 the result incorrectly in the debug output. (It correctly added a newline
5498 to what was transported.)
5500 TF/01 Added $received_time.
5502 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5503 commented out examples of how to interface to a virus scanner and to
5504 SpamAssassin. Also added commented examples of av_scanner and
5505 spamd_address settings.
5507 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5508 and controls are allowed in which ACLs. There were a couple of minor
5509 errors. Some of the entries in the conditions table (which is a table of
5510 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5511 negation of where the condition IS allowed.
5513 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5515 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5516 header file does not have a version number, so I've had to invent a new
5517 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5518 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5519 radiusclient), but it was contributed by a Radius user.
5521 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5522 files or directories, for OpenSSL.
5524 PH/19 When an Exim process that is running as root has to create an Exim log
5525 file, it does so in a subprocess that runs as exim:exim so as to get the
5526 ownership right at creation (otherwise, other Exim processes might see
5527 the file with the wrong ownership). There was no test for failure of this
5528 fork() call, which would lead to the process getting stuck as it waited
5529 for a non-existent subprocess. Forks do occasionally fail when resources
5530 run out. I reviewed all the other calls to fork(); they all seem to check
5533 PH/20 When checking for unexpected SMTP input at connect time (before writing
5534 the banner), Exim was not dealing correctly with a non-positive return
5535 from the read() function. If the client had disconnected by this time,
5536 the result was a log entry for a synchronization error with an empty
5537 string after "input=" when read() returned zero. If read() returned -1
5538 (an event I could not check), uninitialized data bytes were printed.
5539 There were reports of junk text (parts of files, etc) appearing after
5542 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5544 PH/22 Added support for macro redefinition, and (re)definition in between
5545 driver and ACL definitions.
5547 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5548 forgetting to use the resulting value; it was using the unexpanded value.
5550 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5551 hadn't been configured. The fix is from Juergen Kreileder, who
5552 understands it better than I do:
5554 "Here's what I see happening with three configured cyrus_sasl
5555 authenticators configured (plain, login, cram-md5):
5557 On startup auth_cyrus_sasl_init() gets called for each of these.
5558 This means three calls to sasl_listmech() without a specified mech_list.
5559 => SASL tests which mechs of all available mechs actually work
5560 => three warnings about OTP not working
5561 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5563 With the patch, sasl_listmech() also gets called three times. But now
5564 SASL's mech_list option is set to the server_mech specified in the the
5565 authenticator. Or in other words, the answer from sasl_listmech()
5566 gets limited to just the mech you're testing for (which is different
5568 => the return list contains just 'plain' or 'login', 'cram-md5' or
5569 nothing depending on the value of ob->server_mech.
5571 I've just tested the patch: Authentication still works fine,
5572 unavailable mechs specified in the exim configuration are still
5573 caught, and the auth.log warnings about OTP are gone."
5575 PH/25 When debugging is enabled, the contents of the command line are added
5576 to the debugging output, even when log_selector=+arguments is not
5579 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5580 answer is "GNU", and only if the return is "GNU/something" is the answer
5583 PH/27 $acl_verify_message is now set immediately after the failure of a
5584 verification in an ACL, and so is available in subsequent modifiers. In
5585 particular, the message can be preserved by coding like this:
5587 warn !verify = sender
5588 set acl_m0 = $acl_verify_message
5590 Previously, $acl_verify_message was set only while expanding "message"
5591 and "log_message" when a very denied access.
5593 PH/28 Modified OS/os.c-Linux with
5595 -#ifndef OS_LOAD_AVERAGE
5596 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5598 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5599 nomenclature these days.)
5601 PH/29 Installed patch from the Sieve maintainer that adds the options
5602 sieve_useraddress and sieve_subaddress to the redirect router.
5604 PH/30 In these circumstances:
5605 . Two addresses routed to the same list of hosts;
5606 . First host does not offer TLS;
5607 . First host accepts first address;
5608 . First host gives temporary error to second address;
5609 . Second host offers TLS and a TLS session is established;
5610 . Second host accepts second address.
5611 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5612 and peerdn, if requested) that were in fact used only for the second
5615 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5616 attention to any local part prefix or suffix that was matched by the
5617 router that accepted the address. It now behaves in the same way as it
5618 does for delivery: the affixes are removed from the local part unless
5619 rcpt_include_affixes is set on the transport.
5621 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5622 timeout during the DATA phase of an incoming message.
5624 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5625 applied a patch sanctioned by the Sieve maintainer.
5627 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5628 the uid or gid is negative. A case of a negative gid caused this to be
5629 noticed. The fix allows for either to be negative.
5631 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5632 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5635 PH/36 The change PH/12 above was broken. Fixed it.
5637 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5638 the grounds that routing the same address twice would always produce the
5639 same answer. This might have been true once, but it is certainly no
5640 longer true now. Routing a child address may depend on the previous
5641 routing that produced that child. Some complicated redirection strategies
5642 went wrong when messages had multiple recipients, and made Exim's
5643 behaviour dependent on the order in which the addresses were given.
5645 I have moved the duplicate checking until after the routing is complete.
5646 Exim scans the addresses that are assigned to local and remote
5647 transports, and removes any duplicates. This means that more work will be
5648 done, as duplicates will always all be routed, but duplicates are
5649 presumably rare, so I don't expect this is of any significance.
5651 For deliveries to pipes, files, and autoreplies, the duplicate checking
5652 still happens during the routing process, since they are not going to be
5655 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5656 It corrects a timeout issue with spamd. This is Ian's comment: "The
5657 background is that sometimes spamd either never reads data from a
5658 connection it has accepted, or it never writes response data. The exiscan
5659 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5660 blindly assumes that writes won't block so it may never time out."
5662 PH/39 Allow G after quota size as well as K and M.
5664 PH/40 The value set for $authenticated_id in an authenticator may not contain
5665 binary zeroes or newlines because the value is written to log lines and
5666 to spool files. There was no check on this. Now the value is run through
5667 the string_printing() function so that such characters are converted to
5668 printable escape sequences.
5670 PH/41 $message_linecount is a new variable that contains the total number of
5671 lines in the message. Compare $body_linecount, which is the count for the
5674 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5675 bounce and delay warning messages, except in certain special cases, which
5678 (a) An SMTP error message from a remote host;
5679 (b) A message specified in a :fail: redirection;
5680 (c) A message specified in a "fail" command in a system filter;
5681 (d) A message specified in a FAIL return from the queryprogram router;
5682 (e) A message specified by the cannot_route_message router option.
5684 In these cases only, Exim does include the error details in bounce and
5685 warning messages. There are also a few cases where bland messages such
5686 as "unrouteable address" or "local delivery error" are given.
5688 PH/43 $value is now also set for the "else" part of a ${run expansion.
5690 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5691 being worked on, but at least Exim now implements the latest version to
5694 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5695 process to complete was treated as a delivery failure, a timeout while
5696 writing the message to the pipe was logged, but erroneously treated as a
5697 successful delivery. Such timeouts include transport filter timeouts. For
5698 consistency with the overall process timeout, these timeouts are now
5699 treated as errors, giving rise to delivery failures by default. However,
5700 there is now a new Boolean option for the pipe transport called
5701 timeout_defer, which, if set TRUE, converts the failures into defers for
5702 both kinds of timeout. A transport filter timeout is now identified in
5705 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5706 systems where "make" and "gmake" are different, calling "gmake" at top
5707 level broke things. I've arranged for the value of $(MAKE) to be passed
5708 from the Makefile to this script so that it can call the same version of
5712 A note about Exim versions 4.44 and 4.50
5713 ----------------------------------------
5715 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5716 changes of various kinds. As a consequence, a big documentation update was
5717 needed. This delayed the release for rather longer than seemed good, especially
5718 in the light of a couple of (minor) security issues. Therefore, the changes
5719 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5720 release. So 4.44 and 4.50 are in effect two different branches that both start
5723 I have left the 4.50 change log unchanged; it contains all the changes since
5724 4.43. The change log for 4.44 is below; many of its items are identical to
5725 those for 4.50. This seems to be the most sensible way to preserve the
5726 historical information.
5732 1. Minor wording change to the doc/README.SIEVE file.
5734 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5735 computation of the current number of files was incorrect.
5737 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5738 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5741 4. Give more explanation in the error message when the command for a transport
5742 filter fails to execute.
5744 5. There are several places where Exim runs a non-Exim command in a
5745 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5746 was being done only for the command run by the queryprogram router. It is
5747 now done for all such subprocesses. The other cases are: ${run, transport
5748 filters, and the commands run by the lmtp and pipe transports.
5750 6. Added CONFIGURE_GROUP build-time option.
5752 7. Some older OS have a limit of 256 on the maximum number of file
5753 descriptors. Exim was using setrlimit() to set 1000 as a large value
5754 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5755 systems. I've change it so that if it can't get 1000, it tries for 256.
5757 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5758 was an oversight, and furthermore, ever since the addition of extra
5759 controls (e.g. 4.43/32), the checks on when to allow different forms of
5760 "control" were broken. There should now be diagnostics for all cases when a
5761 control that does not make sense is encountered.
5763 9. Added the /retain_sender option to "control=submission".
5765 10. $recipients is now available in the predata ACL (oversight).
5767 11. Tidy the search cache before the fork to do a delivery from a message
5768 received from the command line. Otherwise the child will trigger a lookup
5769 failure and thereby defer the delivery if it tries to use (for example) a
5770 cached ldap connection that the parent has called unbind on.
5772 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5773 of $address_data from the recipient verification was clobbered by the
5774 sender verification.
5776 13. The value of address_data from a sender verification is now available in
5777 $sender_address_data in subsequent conditions in the ACL statement.
5779 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5781 15. Added a new option "connect=<time>" to callout options, to set a different
5784 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5785 was its contents. (It was OK if the option was not defined at all.)
5787 17. A "Completed" log line is now written for messages that are removed from
5788 the spool by the -Mrm option.
5790 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5791 information about exactly what failed.
5793 19. Added -dd to debug only the daemon process.
5795 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5796 handles timeouts, both on the server side and network timeouts. Renamed the
5797 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5799 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5800 It is now set to "smtps".
5802 22. $host_address is now set to the target address during the checking of
5803 ignore_target_hosts.
5805 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5806 being passed; this would have caused $sender_host_name to have been used if
5807 matching the list had actually called for a host name (not very likely,
5808 since this list is usually IP addresses). A host name is now passed as
5811 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5812 code that specifies a non-restarting handler (typically sigaction() in
5813 modern systems) in an attempt to fix a rare and obscure crash bug.
5815 25. Narrowed the window for a race in the daemon that could cause it to ignore
5816 SIGCHLD signals. This is not a major problem, because they are used only to
5817 wake it up if nothing else does.
5819 26. A malformed maildirsize file could cause Exim to calculate negative values
5820 for the mailbox size or file count. Odd effects could occur as a result.
5821 The maildirsize information is now recalculated if the size or filecount
5824 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5825 support for a long time. Removed HAVE_SYS_VFS_H.
5827 28. Installed the latest version of exipick from John Jetmore.
5829 29. In an address list, if the pattern was not a regular expression, an empty
5830 subject address (from a bounce message) matched only if the pattern was an
5831 empty string. Non-empty patterns were not even tested. This was the wrong
5832 because it is perfectly reasonable to use an empty address as part of a
5833 database query. An empty address is now tested by patterns that are
5834 lookups. However, all the other forms of pattern expect the subject to
5835 contain a local part and a domain, and therefore, for them, an empty
5836 address still always fails if the pattern is not itself empty.
5838 30. Exim went into a mad DNS loop when attempting to do a callout where the
5839 host was specified on an smtp transport, and looking it up yielded more
5840 than one IP address.
5842 31. Re-factored the code for checking spool and log partition space into a
5843 function that finds that data and another that does the check. The former
5844 is then used to implement four new variables: $spool_space, $log_space,
5845 $spool_inodes, and $log_inodes.
5847 32. The RFC2047 encoding function was originally intended for short strings
5848 such as real names; it was not keeping to the 75-character limit for
5849 encoded words that the RFC imposes. It now respects the limit, and
5850 generates multiple encoded words if necessary. To be on the safe side, I
5851 have increased the buffer size for the ${rfc2047: expansion operator from
5854 33. It is now permitted to omit both strings after an "if" condition; if the
5855 condition is true, the result is "true". As before, when the second string
5856 is omitted, a false condition yields an empty string. This makes it less
5857 cumbersome to write custom ACL and router conditions.
5859 34. Failure to deliver a bounce message always caused it to be frozen, even if
5860 there was an errors_to setting on the router. The errors_to setting is now
5863 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5864 canonical form (fully expanded) before being placed in
5865 $sender_host_address.
5867 36. The table in the code that translates DNS record types into text (T_A to
5868 "A" for instance) was missing entries for NS and CNAME. It is just possible
5869 that this could have caused confusion if both these types were looked up
5870 for the same domain, because the text type is used as part of Exim's
5871 per-process caching. But the chance of anyone hitting this buglet seems
5874 37. The dnsdb lookup has been extended in a number of ways.
5876 (1) There is a new type, "zns", which walks up the domain tree until it
5877 finds some nameserver records. It should be used with care.
5879 (2) There is a new type, "mxh", which is like "mx" except that it returns
5880 just the host names, not the priorities.
5882 (3) It is now possible to give a list of domains (or IP addresses) to be
5883 looked up. The behaviour when one of the lookups defers can be
5884 controlled by a keyword.
5886 (4) It is now possible to specify the separator character for use when
5887 multiple records are returned.
5889 38. The dnslists ACL condition has been extended: it is now possible to supply
5890 a list of IP addresses and/or domains to be looked up in a particular DNS
5893 39. Added log_selector=+queue_time_overall.
5895 40. When running the queue in the test harness, wait just a tad after forking a
5896 delivery process, to get repeatability of debugging output.
5898 41. Include certificate and key file names in error message when GnuTLS fails
5899 to set them up, because the GnuTLS error message doesn't include the name
5900 of the failing file when there is a problem reading it.
5902 42. Allow both -bf and -bF in the same test run.
5904 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5906 44. The "Exiscan patch" is now merged into the mainline Exim source.
5908 45. Sometimes the final signoff response after QUIT could fail to get
5909 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5910 < 0 before doing a fflush(). This bug looks as though it goes back to the
5911 introduction of TLS in release 3.20, but "sometimes" must have been rare
5912 because the tests only now provoked it.
5914 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5915 (this can affect the format of dates).
5917 47. exim_tidydb, when checking for the continued existence of a message for
5918 which it has found a message-specific retry record, was not finding
5919 messages that were in split spool directories. Consequently, it was
5920 deleting retry records that should have stayed in existence.
5922 48. Steve fixed some bugs in eximstats.
5924 49. The SPA authentication driver was not abandoning authentication and moving
5925 on to the next authenticator when an expansion was forced to fail,
5926 contradicting the general specification for all authenticators. Instead it
5927 was generating a temporary error. It now behaves as specified.
5929 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5930 (the order specifies the preference for clients). The order is now AES256,
5931 AES128, 3DES, ARCFOUR128.
5933 51. Small patch to Sieve code - explicitly set From: when generating an
5936 52. Exim crashed if a remote delivery caused a very long error message to be
5937 recorded - for instance if somebody sent an entire SpamAssassin report back
5938 as a large number of 550 error lines. This bug was coincidentally fixed by
5939 increasing the size of one of Exim's internal buffers (big_buffer) that
5940 happened as part of the Exiscan merge. However, to be on the safe side, I
5941 have made the code more robust (and fixed the comments that describe what
5944 53. Now that there can be additional text after "Completed" in log lines (if
5945 the queue_time_overall log selector is set), a one-byte patch to exigrep
5946 was needed to allow it to recognize "Completed" as not the last thing in
5949 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5950 patch that reportedly fixes this has been added. I am not expert enough to
5951 create a test for it. This is what the patch creator wrote:
5953 "I found a little strange behaviour of ldap code when working with
5954 Windows 2003 AD Domain, where users was placed in more than one
5955 Organization Units. When I tried to give exim partial DN, the exit code
5956 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5957 But simultaneously result of request was absolutely normal ldap result,
5958 so I produce this patch..."
5960 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5961 REFERENCE, so I have modified the code to exclude the patch when that macro
5964 55. Some experimental protocols are using DNS PTR records for new purposes. The
5965 keys for these records are domain names, not reversed IP addresses. The
5966 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5967 leaves it alone. Component reversal etc. now happens only for IP addresses.
5970 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5972 57. Double the size of the debug message buffer (to 2048) so that more of very
5973 long debug lines gets shown.
5975 58. The exicyclog utility now does better if the number of log files to keep
5976 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5978 59. Two changes related to the smtp_active_hostname option:
5980 (1) $smtp_active_hostname is now available as a variable.
5981 (2) The default for smtp_banner uses $smtp_active_hostname instead
5982 of $primary_hostname.
5984 60. The host_aton() function is supposed to be passed a string that is known
5985 to be a valid IP address. However, in the case of IPv6 addresses, it was
5986 not checking this. This is a hostage to fortune. Exim now panics and dies
5987 if the condition is not met. A case was found where this could be provoked
5988 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5989 components; fortuitously, this particular loophole had already been fixed
5990 by change 4.50/55 above.
5992 If there are any other similar loopholes, the new check in host_aton()
5993 itself should stop them being exploited. The report I received stated that
5994 data on the command line could provoke the exploit when Exim was running as
5995 exim, but did not say which command line option was involved. All I could
5996 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5997 running as the user.
6000 61. There was a buffer overflow vulnerability in the SPA authentication code
6001 (which came originally from the Samba project). I have added a test to the
6002 spa_base64_to_bits() function which I hope fixes it.
6005 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6006 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6008 63. The daemon start-up calls getloadavg() while still root for those OS that
6009 need the first call to be done as root, but it missed one case: when
6010 deliver_queue_load_max is set with deliver_drop_privilege. This is
6011 necessary for the benefit of the queue runner, because there is no re-exec
6012 when deliver_drop_privilege is set.
6014 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6015 This has been fixed.
6017 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6018 was in use, was not putting the data itself into the right store pool;
6019 consequently, it could be overwritten for a subsequent message in the same
6020 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6023 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6025 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6026 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6027 the function were treating the return as a boolean value, which happened to
6028 work because 0=false and not-0=true, but is not correct code.
6030 68. The host_aton() function was not handling scoped IPv6 addresses (those
6031 with, for example, "%eth0" on the end) correctly.
6033 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6034 negated items (that is, ~something) in unsigned ints. Some compilers
6035 apparently mutter when there is no cast.
6037 70. If an address verification called from an ACL failed, and did not produce a
6038 user-specific message (i.e. there was only a "system" message), nothing was
6039 put in $acl_verify_message. In this situation, it now puts the system
6042 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6043 session; change 4.31/43 added the unwanted input to the log line - except
6044 that it did not do this in the start of session case. It now does.
6046 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6047 This is wrong and can cause the other end to generate a synchronization
6048 error if it is another Exim or anything else that does the synchronization
6049 check. A QUIT command is no longer sent after a timeout.
6051 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6052 during host lookups.
6054 74. The defer_ok option of callout verification was not working if it was used
6055 when verifying addresses in header lines, that is, for this case:
6057 verify = header_sender/callout=defer_ok
6059 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6060 those file descriptors could be used for SMTP connections. If anything
6061 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6062 could be sent to the SMTP client, causing chaos. The daemon now opens
6063 stdin, stdout, and stderr to /dev/null when it puts itself into the
6066 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6067 log by default. The user can override this with suitable Perl magic.
6069 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6070 the log message when discard triggers, was not working for the DATA ACL or
6071 for the non-SMTP ACL.
6073 78. Error message wording change in sieve.c.
6075 79. If smtp_accept_max_per_host was set, the number of connections could be
6076 restricted to fewer than expected, because the daemon was trying to set up
6077 a new connection before checking whether the processes handling previous
6078 connections had finished. The check for completed processes is now done
6079 earlier. On busy systems, this bug wouldn't be noticed because something
6080 else would have woken the daemon, and it would have reaped the completed
6083 80. If a message was submitted locally by a user whose login name contained one
6084 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6085 It caused a spool format error. I have fixed the spool reading code. A
6086 related problem was that the "from" clause in the Received: line became
6087 illegal because of the space(s). It is now covered by ${quote_local_part.
6089 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6092 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6094 83. Patch from Sieve maintainer to fix unterminated string problem in
6095 "vacation" handling.
6097 84. Some minor changes to the Linux configuration files to help with other
6098 OS variants using glibc.
6100 85. One more patch for Sieve to update vacation handling to latest spec.
6103 ----------------------------------------------------
6104 See the note above about the 4.44 and 4.50 releases.
6105 ----------------------------------------------------
6111 1. Change 4.43/35 introduced a bug that caused file counts to be
6112 incorrectly computed when quota_filecount was set in an appendfile
6115 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6116 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6119 3. Give more explanation in the error message when the command for a transport
6120 filter fails to execute.
6122 4. There are several places where Exim runs a non-Exim command in a
6123 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6124 was being done only for the command run by the queryprogram router. It is
6125 now done for all such subprocesses. The other cases are: ${run, transport
6126 filters, and the commands run by the lmtp and pipe transports.
6128 5. Some older OS have a limit of 256 on the maximum number of file
6129 descriptors. Exim was using setrlimit() to set 1000 as a large value
6130 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6131 systems. I've change it so that if it can't get 1000, it tries for 256.
6133 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6134 was an oversight, and furthermore, ever since the addition of extra
6135 controls (e.g. 4.43/32), the checks on when to allow different forms of
6136 "control" were broken. There should now be diagnostics for all cases when a
6137 control that does not make sense is encountered.
6139 7. $recipients is now available in the predata ACL (oversight).
6141 8. Tidy the search cache before the fork to do a delivery from a message
6142 received from the command line. Otherwise the child will trigger a lookup
6143 failure and thereby defer the delivery if it tries to use (for example) a
6144 cached ldap connection that the parent has called unbind on.
6146 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6147 of $address_data from the recipient verification was clobbered by the
6148 sender verification.
6150 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6151 was its contents. (It was OK if the option was not defined at all.)
6153 11. A "Completed" log line is now written for messages that are removed from
6154 the spool by the -Mrm option.
6156 12. $host_address is now set to the target address during the checking of
6157 ignore_target_hosts.
6159 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6160 being passed; this would have caused $sender_host_name to have been used if
6161 matching the list had actually called for a host name (not very likely,
6162 since this list is usually IP addresses). A host name is now passed as
6165 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6166 code that specifies a non-restarting handler (typically sigaction() in
6167 modern systems) in an attempt to fix a rare and obscure crash bug.
6169 15. Narrowed the window for a race in the daemon that could cause it to ignore
6170 SIGCHLD signals. This is not a major problem, because they are used only to
6171 wake it up if nothing else does.
6173 16. A malformed maildirsize file could cause Exim to calculate negative values
6174 for the mailbox size or file count. Odd effects could occur as a result.
6175 The maildirsize information is now recalculated if the size or filecount
6178 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6179 support for a long time. Removed HAVE_SYS_VFS_H.
6181 18. Updated exipick to current release from John Jetmore.
6183 19. Allow an empty sender to be matched against a lookup in an address list.
6184 Previously the only cases considered were a regular expression, or an
6187 20. Exim went into a mad DNS lookup loop when doing a callout where the
6188 host was specified on the transport, if the DNS lookup yielded more than
6191 21. The RFC2047 encoding function was originally intended for short strings
6192 such as real names; it was not keeping to the 75-character limit for
6193 encoded words that the RFC imposes. It now respects the limit, and
6194 generates multiple encoded words if necessary. To be on the safe side, I
6195 have increased the buffer size for the ${rfc2047: expansion operator from
6198 22. Failure to deliver a bounce message always caused it to be frozen, even if
6199 there was an errors_to setting on the router. The errors_to setting is now
6202 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6203 canonical form (fully expanded) before being placed in
6204 $sender_host_address.
6206 24. Updated eximstats to version 1.33
6208 25. Include certificate and key file names in error message when GnuTLS fails
6209 to set them up, because the GnuTLS error message doesn't include the name
6210 of the failing file when there is a problem reading it.
6212 26. Expand error message when OpenSSL has problems setting up cert/key files.
6215 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6216 (this can affect the format of dates).
6218 28. exim_tidydb, when checking for the continued existence of a message for
6219 which it has found a message-specific retry record, was not finding
6220 messages that were in split spool directories. Consequently, it was
6221 deleting retry records that should have stayed in existence.
6223 29. eximstats updated to version 1.35
6224 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6225 1.35 - bugfix such that pie charts by volume are generated correctly
6227 30. The SPA authentication driver was not abandoning authentication and moving
6228 on to the next authenticator when an expansion was forced to fail,
6229 contradicting the general specification for all authenticators. Instead it
6230 was generating a temporary error. It now behaves as specified.
6232 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6233 (the order specifies the preference for clients). The order is now AES256,
6234 AES128, 3DES, ARCFOUR128.
6236 31. Small patch to Sieve code - explicitly set From: when generating an
6239 32. Exim crashed if a remote delivery caused a very long error message to be
6240 recorded - for instance if somebody sent an entire SpamAssassin report back
6241 as a large number of 550 error lines. This bug was coincidentally fixed by
6242 increasing the size of one of Exim's internal buffers (big_buffer) that
6243 happened as part of the Exiscan merge. However, to be on the safe side, I
6244 have made the code more robust (and fixed the comments that describe what
6247 33. Some experimental protocols are using DNS PTR records for new purposes. The
6248 keys for these records are domain names, not reversed IP addresses. The
6249 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6250 leaves it alone. Component reversal etc. now happens only for IP addresses.
6253 34. The host_aton() function is supposed to be passed a string that is known
6254 to be a valid IP address. However, in the case of IPv6 addresses, it was
6255 not checking this. This is a hostage to fortune. Exim now panics and dies
6256 if the condition is not met. A case was found where this could be provoked
6257 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6258 components; fortuitously, this particular loophole had already been fixed
6259 by change 4.50/55 or 4.44/33 above.
6261 If there are any other similar loopholes, the new check in host_aton()
6262 itself should stop them being exploited. The report I received stated that
6263 data on the command line could provoke the exploit when Exim was running as
6264 exim, but did not say which command line option was involved. All I could
6265 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6266 running as the user.
6269 35. There was a buffer overflow vulnerability in the SPA authentication code
6270 (which came originally from the Samba project). I have added a test to the
6271 spa_base64_to_bits() function which I hope fixes it.
6274 36. The daemon start-up calls getloadavg() while still root for those OS that
6275 need the first call to be done as root, but it missed one case: when
6276 deliver_queue_load_max is set with deliver_drop_privilege. This is
6277 necessary for the benefit of the queue runner, because there is no re-exec
6278 when deliver_drop_privilege is set.
6280 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6281 was in use, was not putting the data itself into the right store pool;
6282 consequently, it could be overwritten for a subsequent message in the same
6283 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6286 38. Sometimes the final signoff response after QUIT could fail to get
6287 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6288 < 0 before doing a fflush(). This bug looks as though it goes back to the
6289 introduction of TLS in release 3.20, but "sometimes" must have been rare
6290 because the tests only now provoked it.
6296 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6297 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6298 Now it returns an integer. A number of calls were still expecting a T/F
6299 return. Fortuitously, in all cases, the tests worked in OK situations,
6300 which is the norm. However, things would have gone wrong on any write
6301 failures on the smtp file descriptor. This function is used when sending
6302 messages over SMTP and also when doing verify callouts.
6304 2. When Exim is called to do synchronous delivery of a locally submitted
6305 message (the -odf or -odi options), it no longer closes stderr before doing
6308 3. Implemented the mua_wrapper option.
6310 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6312 5. Implemented the functions header_remove(), header_testname(),
6313 header_add_at_position(), and receive_remove_recipient(), and exported them
6316 6. If an ACL "warn" statement specified the addition of headers, Exim already
6317 inserted X-ACL-Warn: at the start if there was no header name. However, it
6318 was not making this test for the second and subsequent header lines if
6319 there were newlines in the string. This meant that an invalid header could
6320 be inserted if Exim was badly configured.
6322 7. Allow an ACL "warn" statement to add header lines at the start or after all
6323 the Received: headers, as well as at the end.
6325 8. Added the rcpt_4xx retry error code.
6327 9. Added postmaster_mailfrom=xxx to callout verification option.
6329 10. Added mailfrom=xxxx to the callout verification option, for verify=
6332 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6333 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6335 12. Inserted some casts to stop certain compilers warning when using pointer
6336 differences as field lengths or precisions in printf-type calls (mostly
6337 affecting debugging statements).
6339 13. Added optional readline() support for -be (dynamically loaded).
6341 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6342 same clock tick as a message's arrival, so that its received time was the
6343 same as the "first fail" time on the retry record, and that message
6344 remained on the queue past the ultimate address timeout, every queue runner
6345 would try a delivery (because it was past the ultimate address timeout) but
6346 after another failure, the ultimate address timeout, which should have then
6347 bounced the address, did not kick in. This was a "< instead of <=" error;
6348 in most cases the first failure would have been in the next clock tick
6349 after the received time, and all would be well.
6351 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6352 being recognized when the domain list was tested by the match_domain
6353 condition in an expansion string.
6355 16. Added the ${str2b64: operator.
6357 17. Exim was always calling setrlimit() to set a large limit for the number of
6358 processes, without checking whether the existing limit was already
6359 adequate. (It did check for the limit on file descriptors.) Furthermore,
6360 errors from getrlimit() and setrlimit() were being ignored. Now they are
6361 logged to the main and panic logs, but Exim does carry on, to try to do its
6362 job under whatever limits there are.
6364 18. Imported PCRE 5.0.
6366 19. Trivial typo in log message " temporarily refused connection" (the leading
6369 20. If the log selector return_path_on_delivery was set and an address was
6370 redirected to /dev/null, the delivery process crashed because it assumed
6371 that a return path would always be set for a "successful" delivery. In this
6372 case, the whole delivery is bypassed as an optimization, and therefore no
6375 21. Internal re-arrangement: the function for sending a challenge and reading
6376 a response while authentication was assuming a zero-terminated challenge
6377 string. It's now changed to take a pointer and a length, to allow for
6378 binary data in such strings.
6380 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6382 23. Exim was not respecting finduser_retries when seeking the login of the
6383 uid under which it was called; it was always trying 10 times. (The default
6384 setting of finduser_retries is zero.) Also, it was sleeping after the final
6385 failure, which is pointless.
6387 24. Implemented tls_on_connect_ports.
6389 25. Implemented acl_smtp_predata.
6391 26. If the domain in control=submission is set empty, Exim assumes that the
6392 authenticated id is a complete email address when it generates From: or
6393 Sender: header lines.
6395 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6396 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6397 chown and chgrp in /bin and hostname in /usr/bin.
6399 28. Exim was keeping the "process log" file open after each use, just as it
6400 does for the main log. This opens the possibility of it remaining open for
6401 long periods when the USR1 signal hits a daemon. Occasional processlog
6402 errors were reported, that could have been caused by this. Anyway, it seems
6403 much more sensible not to leave this file open at all, so that is what now
6406 29. The long-running daemon process does not normally write to the log once it
6407 has entered its main loop, and it closes the log before doing so. This is
6408 so that log files can straightforwardly be renamed and moved. However,
6409 there are a couple of unusual error situations where the daemon does write
6410 log entries, and I had neglected to close the log afterwards.
6412 30. The text of an SMTP error response that was received during a remote
6413 delivery was being truncated at 512 bytes. This is too short for some of
6414 the long messages that one sometimes sees. I've increased the limit to
6417 31. It is now possible to make retry rules that apply only when a message has a
6418 specific sender, in particular, an empty sender.
6420 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6421 it possible to be selective about when SMTP synchronization is enforced.
6423 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6425 32. Implemented hosts_connection_nolog.
6427 33. Added an ACL for QUIT.
6429 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6432 35. Added mailbox_size and mailbox_filecount to appendfile.
6434 36. Added control = no_multiline_responses to ACLs.
6436 37. There was a bug in the logic of the code that waits for the clock to tick
6437 in the case where the clock went backwards by a substantial amount such
6438 that the microsecond fraction of "now" was more than the microsecond
6439 fraction of "then" (but the whole seconds number was less).
6441 38. Added support for the libradius Radius client library this is found on
6442 FreeBSD (previously only the radiusclient library was supported).
6448 1. When certain lookups returned multiple values in the form name=value, the
6449 quoting of the values was not always being done properly. Specifically:
6450 (a) If the value started with a double quote, but contained no whitespace,
6452 (b) If the value contained whitespace other than a space character (i.e.
6453 tabs or newlines or carriage returns) it was not quoted.
6454 This fix has been applied to the mysql and pgsql lookups by writing a
6455 separate quoting function and calling it from the lookup code. The fix
6456 should probably also be applied to nisplus, ibase and oracle lookups, but
6457 since I cannot test any of those, I have not disturbed their existing code.
6459 2. A hit in the callout cache for a specific address caused a log line with no
6460 reason for rejecting RCPT. Now it says "Previous (cached) callout
6461 verification failure".
6463 3. There was an off-by-one bug in the queryprogram router. An over-long
6464 return line was truncated at 256 instead of 255 characters, thereby
6465 overflowing its buffer with the terminating zero. As well as fixing this, I
6466 have increased the buffer size to 1024 (and made a note to document this).
6468 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6469 when Exim is waiting for an SMTP response from a remote server, Exim
6470 restarts its select() call on the socket, thereby resetting its timeout.
6471 This is not a problem when such interrupts are rare. Somebody set up a cron
6472 job to run exiwhat every 2 minutes, which is less than the normal select()
6473 timeout (5 or 10 minutes). This meant that the select() timeout never
6474 kicked in because it was always reset. I have fixed this by comparing the
6475 time when an interrupt arrives with the time at the start of the first call
6476 to select(). If more time than the timeout has elapsed, the interrupt is
6477 treated as a timeout.
6479 5. Some internal re-factoring in preparation for the addition of Sieve
6480 extensions (by MH). In particular, the "personal" test is moved to a
6481 separate function, and given an option for scanning Cc: and Bcc: (which is
6482 not set for Exim filters).
6484 6. When Exim created an email address using the login of the caller as the
6485 local part (e.g. when creating a From: or Sender: header line), it was not
6486 quoting the local part when it contained special characters such as @.
6488 7. Installed new OpenBSD configuration files.
6490 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6491 try to make them clearer.
6493 9. Callout options, other than the timeout value, were being ignored when
6494 verifying sender addresses in header lines. For example, when using
6496 verify = header_sender/callout=no_cache
6498 the cache was (incorrectly) being used.
6500 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6501 only the Cygwin environment.
6503 11. When return_path_on_delivery was set as a log selector, if different remote
6504 addresses in the same message used different return paths and parallel
6505 remote delivery occurred, the wrong values would sometimes be logged.
6506 (Whenever a remote delivery process finished, the return path value from
6507 the most recently started remote delivery process was logged.)
6509 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6510 lines when AUTH and/or TLS are in use. This is the "received protocol"
6511 field. Exim used to use "asmtp" for authenticated SMTP, without any
6512 indication (in the protocol name) for TLS use. Now it follows the RFC and
6513 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6514 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6515 names appear in log lines as well as in Received: header lines.
6517 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6518 extensions, and comparison tests, and to fix some bugs.
6520 14. Changes to the "personal" filter test:
6522 (1) The test was buggy in that it was just doing the equivalent of
6523 "contains" tests on header lines. For example, if a user's address was
6524 anne@some.where, the "personal" test would incorrectly be true for
6526 To: susanne@some.where
6528 This test is now done by extracting each address from the header in turn,
6529 and checking the entire address. Other tests that are part of "personal"
6530 are now done using regular expressions (for example, to check local parts
6531 of addresses in From: header lines).
6533 (2) The list of non-personal local parts in From: addresses has been
6534 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6535 taken from the Sieve specification recommendations.
6537 (3) If the message contains any header line starting with "List-" it is
6538 treated as non-personal.
6540 (4) The test for "circular" in the Subject: header line has been removed
6541 because it now seems ill-conceived.
6543 15. Minor typos in src/EDITME comments corrected.
6545 16. Installed latest exipick from John Jetmore.
6547 17. If headers_add on a router specified a text string that was too long for
6548 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6549 of string_sprintf() is now avoided.
6551 18. $message_body_size was not set (it was always zero) when running the DATA
6552 ACL and the local_scan() function.
6554 19. For the "mail" command in an Exim filter, no default was being set for
6555 the once_repeat time, causing a random time value to be used if "once" was
6556 specified. (If the value happened to be <= 0, no repeat happened.) The
6557 default is now 0s, meaning "never repeat". The "vacation" command was OK
6558 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6559 (I found it when inspecting the code).
6561 20. There is now an overall timeout for performing a callout verification. It
6562 defaults to 4 times the callout timeout, which applies to individual SMTP
6563 commands during the callout. The overall timeout applies when there is more
6564 than one host that can be tried. The timeout is checked before trying the
6565 next host. This prevents very long delays if there are a large number of
6566 hosts and all are timing out (e.g. when the network connections are timing
6567 out). The value of the overall timeout can be changed by specifying an
6568 additional sub-option for "callout", called "maxwait". For example:
6570 verify = sender/callout=5s,maxwait=20s
6572 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6573 to the end before writing, but this should make it even safer).
6575 22. Exim was forgetting that it had advertised PIPELINING for the second and
6576 subsequent messages on an SMTP connection. It was also not resetting its
6577 memory on STARTTLS and an internal HELO.
6579 23. When Exim logs an SMTP synchronization error within a session, it now
6580 records whether PIPELINING has been advertised or not.
6582 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6583 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6584 rather than long int.
6586 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6588 26. Added the never_mail option to autoreply.
6594 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6595 crash if the getsockname() call failed; this can happen if a connection is
6596 closed very soon after it is established. The problem was simply in the
6597 order in which certain operations were done, causing Exim to try to write
6598 to the SMTP stream before it had set up the file descriptor. The bug has
6599 been fixed by making things happen in the correct order.
6605 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6606 before the connection was closed, thus losing the rejection response.
6608 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6609 some early Solaris releases, but causes trouble in current releases where
6610 socklen_t is defined.
6612 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6615 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6618 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6619 that releasing the top of it at the end releases what was used for sub-
6620 expansions (unless the block got too big). However, discard this block if
6621 the first thing is a variable or header, so that we can use its block when
6622 it is dynamic (useful for very large $message_headers, for example).
6624 6. Lookups now cache *every* query, not just the most recent. A new, separate
6625 store pool is used for this. It can be recovered when all lookup caches are
6626 flushed. Lookups now release memory at the end of their result strings.
6627 This has involved some general refactoring of the lookup sources.
6629 7. Some code has been added to the store_xxx() functions to reduce the amount
6630 of flapping under certain conditions.
6632 8. log_incoming_interface used to affect only the <= reception log lines. Now
6633 it causes the local interface and port to be added to several more SMTP log
6634 lines, for example "SMTP connection from", and rejection lines.
6636 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6638 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6640 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6641 could be overwritten at the end of the current message (or the start of a
6642 new message if it was set in a HELO ACL). The value is now preserved for
6643 the duration of the SMTP connection.
6645 12. If a transport had a headers_rewrite setting, and a matching header line
6646 contained an unqualified address, that address was qualified, even if it
6647 did not match any rewriting rules. The underlying bug was that the values
6648 of the flags that permit the existence of unqualified sender and recipient
6649 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6650 non-local messages, and by -bnq for local messages) were not being
6651 preserved with the message after it was received.
6653 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6654 "next input=" as part of the text comprising the host identity instead of
6655 the correct text. The code was using the same buffer for two different
6656 strings. However, depending on which order the printing function evaluated
6657 its arguments, the bug did not always show up. Under Linux, for example, my
6658 test suite worked just fine.
6660 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6661 doesn't work with some older versions of Perl. It has been changed to "my",
6662 which in any case is probably the better facility to use.
6664 15. A really picky compiler found some instances of statements for creating
6665 error messages that either had too many or two few arguments for the format
6668 16. The size of the buffer for calls to the DNS resolver has been increased
6669 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6670 for addresses that have a lot of PTR records. This alleviates a problem; it
6671 does not fully solve it.
6673 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6674 buffer now truncates the list and logs the incident, which is the same
6675 action as happens when Exim is looking up a host name and its aliases.
6676 Previously in this situation something unpredictable would happen;
6677 sometimes it was "internal error: store_reset failed".
6679 18. If a server dropped the connection unexpectedly when an Exim client was
6680 using GnuTLS and trying to read a response, the client delivery process
6681 crashed while trying to generate an error log message.
6683 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6684 string, for example:
6686 warn message = H1: something\nH2: something
6688 the text was added as a single header line from Exim's point of view
6689 though it ended up OK in the delivered message. However, searching for the
6690 second and subsequent header lines using $h_h2: did not work. This has been
6691 fixed. Similarly, if a system filter added multiple headers in this way,
6692 the routers could not see them.
6694 20. Expanded the error message when iplsearch is called with an invalid key to
6695 suggest using net-iplsearch in a host list.
6697 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6698 ACLs are no longer actually imposed (and a message to that effect is
6701 22. If a "gecos" field in a passwd entry contained escaped characters, in
6702 particular, if it contained a \" sequence, Exim got it wrong when building
6703 a From: or a Sender: header from that name. A second bug also caused
6704 incorrect handling when an unquoted " was present following a character
6705 that needed quoting.
6707 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6708 was not being matched caselessly.
6710 24. Arranged for all hyphens in the exim.8 source to be escaped with
6713 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6714 properly. Recipient callout cache records were still being keyed to include
6715 the sender, even when use_sender was set false. This led to far more
6716 callouts that were necessary. The sender is no longer included in the key
6717 when use_sender is false.
6719 26. Added "control = submission" modifier to ACLs.
6721 27. Added the ${base62d: operator to decode base 62 numbers.
6723 28. dnsdb lookups can now access SRV records.
6725 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6726 the configuration file.
6728 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6729 (-v) mode. This makes the output for a verbose queue run more intelligible.
6731 31. Added a use_postmaster feature to recipient callouts.
6733 32. Added the $body_zerocount variable, containing the number of binary zero
6734 bytes in the message body.
6736 33. The time of last modification of the "new" subdirectory is now used as the
6737 "mailbox time last read" when there is a quota error for a maildir
6740 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6742 35. Added +ignore_unknown as a special item in host lists.
6744 36. Code for decoding IPv6 addresses in host lists is now included, even if
6745 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6746 address was recognized as an IP address, but was then not correctly decoded
6747 into binary, causing unexpected and incorrect effects when compared with
6754 1. Very minor rewording of debugging text in manualroute to say "list of
6755 hosts" instead of "hostlist".
6757 2. If verify=header_syntax was set, and a header line with an unqualified
6758 address (no domain) and a large number of spaces between the end of the
6759 name and the colon was received, the reception process suffered a buffer
6760 overflow, and (when I tested it) crashed. This was caused by some obsolete
6761 code that should have been removed. The fix is to remove it!
6763 3. When running in the test harness, delay a bit after writing a bounce
6764 message to get a bit more predictability in the log output.
6766 4. Added a call to search_tidyup() just before forking a reception process. In
6767 theory, someone could use a lookup in the expansion of smtp_accept_max_
6768 per_host which, without the tidyup, could leave open a database connection.
6770 5. Added the variables $recipient_data and $sender_data which get set from a
6771 lookup success in an ACL "recipients" or "senders" condition, or a router
6772 "senders" option, similar to $domain_data and $local_part_data.
6774 6. Moved the writing of debug_print from before to after the "senders" test
6777 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6778 problems for message scanning, either using a data ACL, or using
6779 local_scan() because the Received: header was not generated till after they
6780 were called (in order to set the time as the time of reception completion).
6781 I have revised the way this works. The header is now generated after the
6782 body is received, but before the ACL or local_scan() are called. After they
6783 are run, the timestamp in the header is updated.
6789 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6790 before starting a queue runner without re-exec. This happened only when
6791 deliver_drop_privilege was set or when the Exim user was set to root. The
6792 effect of the bug was that timeouts during subsequent deliveries caused
6793 crashes instead of being properly handled. The handler is now left at its
6794 default (and expected) setting.
6796 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6797 message, again when deliver_drop_privilege is set or Exim is run as root.
6798 The bug described in (1) was not present in this case, but the tidying up
6799 of the other signals was missing. I have made the two cases consistent.
6801 3. The ignore_target_hosts setting on a manualroute router was being ignored
6802 for hosts that were looked up using the /MX notation.
6804 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6807 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6808 operated on the sender address. After changing the $sender_address to <>
6809 for the sender address verify, Exim was re-instated it as the original
6810 (before rewriting) address, but remembering that it had rewritten it, so it
6811 wasn't rewriting it again. This bug also had the effect of breaking the
6812 sender address verification caching when the sender address was rewritten.
6814 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6815 This has been changed so that if the ip literal address matches
6816 ignore_target_hosts, the router declines.
6818 7. Added expansion conditions match_domain, match_address, and match_local_
6819 part (NOT match_host).
6821 8. The placeholder for the Received: header didn't have a length field set.
6823 9. Added code to Exim itself and to exim_lock to test for a specific race
6824 condition that could lead to file corruption when using MBX delivery. The
6825 issue is with the lockfile that is created in /tmp. If this file is removed
6826 after a process has opened it but before that process has acquired a lock,
6827 there is the potential for a second process to recreate the file and also
6828 acquire a lock. This could lead to two Exim processes writing to the file
6829 at the same time. The added code performs the same test as UW imapd; it
6830 checks after acquiring the lock that its file descriptor still refers to
6831 the same named file.
6833 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6834 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6837 11. Added the smtp_active_hostname option. If used, this will typically be made
6838 to depend on the incoming interface address. Because $interface_address is
6839 not set up until the daemon has forked a reception process, error responses
6840 that can happen earlier (such as "too many connections") no longer contain
6843 12. If an expansion in a condition on a "warn" statement fails because a lookup
6844 defers, the "warn" statement is abandoned, and the next ACL statement is
6845 processed. Previously this caused the whole ACL to be aborted.
6847 13. Added the iplsearch lookup type.
6849 14. Added ident_timeout as a log selector.
6851 15. Added tls_certificate_verified as a log selector.
6853 16. Added a global option tls_require_ciphers (compare the smtp transport
6854 option of the same name). This controls incoming TLS connections.
6856 17. I finally figured out how to make tls_require_ciphers do a similar thing
6857 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6858 before starting the TLS session.
6860 18. Tabs are now shown as \t in -bP output.
6862 19. If the log selector return_path_on_delivery was set, Exim crashed when
6863 bouncing a message because it had too many Received: header lines.
6865 20. If two routers both had headers_remove settings, and the first one included
6866 a superfluous trailing colon, the final name in the first list and the
6867 first name in the second list were incorrectly joined into one item (with a
6868 colon in the middle).
6874 1. Added -C and -D options to the exinext utility, mainly to make it easier
6875 to include in the automated testing, but these could be helpful when
6876 multiple configurations are in use.
6878 2. The exinext utility was not formatting the output nicely when there was
6879 an alternate port involved in the retry record key, nor when there was a
6880 message id as well (for retries that were specific to a specific message
6881 and a specific host). It was also confused by IPv6 addresses, because of
6882 the additional colons they contain. I have fixed the IPv4 problem, and
6883 patched it up to do a reasonable job for IPv6.
6885 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6886 delivery, the log line now contains "pipelined" if PIPELINING was used.
6888 4. An SMTP transport process used to panic and die if the bind() call to set
6889 an explicit outgoing interface failed. This has been changed; it is now
6890 treated in the same way as a connect() failure.
6892 5. A reference to $sender_host_name in the part of a conditional expansion
6893 that was being skipped was still causing a DNS lookup. This no longer
6896 6. The def: expansion condition was not recognizing references to header lines
6897 that used bh_ and bheader_.
6899 7. Added the _cache feature to named lists.
6901 8. The code for checking quota_filecount in the appendfile transport was
6902 allowing one more file than it should have been.
6904 9. For compatibility with Sendmail, the command line option
6912 and sets the incoming protocol and host name (for trusted callers). The
6913 host name and its colon can be omitted when only the protocol is to be set.
6914 Note the Exim already has two private options, -pd and -ps, that refer to
6915 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6916 "s", but I don't think that's a major issue.
6918 10. A number of refactoring changes to the code, none of which should affect
6921 (a) The number of logging options was getting close to filling up the
6922 32-bit word that was used as a bit map. I have split them into two classes:
6923 those that are passed in the argument to log_write(), and those that are
6924 only ever tested independently outside of that function. These are now in
6925 separate 32-bit words, so there is plenty of room for expansion again.
6926 There is no change in the user interface or the logging behaviour.
6928 (b) When building, for example, log lines, the code previously used a
6929 macro that called string_cat() twice, in order to add two strings. This is
6930 not really sufficiently general. Furthermore, there was one instance where
6931 it was actually wrong because one of the argument was used twice, and in
6932 one call a function was used. (As it happened, calling the function twice
6933 did not affect the overall behaviour.) The macro has been replaced by a
6934 function that can join an arbitrary number of extra strings onto a growing
6937 (c) The code for expansion conditions now uses a table and a binary chop
6938 instead of a serial search (which was left over from when there were very
6939 few conditions). Also, it now recognizes conditions like "pam" even when
6940 the relevant support is not compiled in: a suitably worded error message is
6941 given if an attempt is made to use such a condition.
6943 11. Added ${time_interval:xxxxx}.
6945 12. A bug was causing one of the ddress fields not to be passed back correctly
6946 from remote delivery subprocesses. The field in question was not being
6947 subsequently used, so this caused to problems in practice.
6949 13. Added new log selectors queue_time and deliver_time.
6951 14. Might have fixed a bug in maildirsizefile handling that threw up
6952 "unexpected character" debug warnings, and recalculated the data
6953 unnecessarily. In any case, I expanded the warning message to give more
6956 15. Added the message "Restricted characters in address" to the statements in
6957 the default ACL that block characters like @ and % in local parts.
6959 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6960 Three changes have been made:
6962 (a) There was a serious bug; a negative response to MAIL caused the whole
6963 recipient domain to be cached as invalid, thereby blocking all messages
6964 to all local parts at the same domain, from all senders. This bug has
6965 been fixed. The domain is no longer cached after a negative response to
6966 MAIL if the sender used is not empty.
6968 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6971 (c) A new callout option, "use_sender" has been added for people who want
6972 the modified behaviour.
6978 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6981 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6982 indeed breaks things for older releases.
6984 3. Added additional logging to the case where there is a problem reading data
6985 from a filter that is running in a subprocess using a pipe, in order to
6986 try to track down a specific problem.
6988 4. Testing facility fudge: when running in the test harness and attempting
6989 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6990 getting "No route to host". Convert this to a timeout.
6992 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6995 6. Some OS don't have socklen_t but use size_t instead. This affects the
6996 fifth argument of getsockopt() amongst other things. This is now
6997 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6998 can be set for individual OS. I have set it for SunOS5, OSF1, and
6999 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7000 some earlier ones do not.
7002 7. Change 4.30/15 was not doing the test caselessly.
7004 8. The standard form for an IPv6 address literal was being rejected by address
7005 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7006 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7007 this, as well as the form without the "IPv6" on the front (but only when
7008 address literals are enabled, of course).
7010 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7012 10. Exim crashed if a message with an empty sender address specified by -f
7013 encountered a router with an errors_to setting. This could be provoked only
7014 by a command such as
7018 where an empty string was supplied; "<>" did not hit this bug.
7020 11. Installed PCRE release 4.5.
7022 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7023 remained set. It is now erased.
7025 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7026 times from message ids (which are base 36 rather than the normal 62).
7028 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7029 were being counted as actual protocol errors, and logged if the log
7030 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7031 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7032 and DATA following a set of rejected RCPTs do not count as protocol errors.
7033 In other words, Exim assumes they were pipelined, though this may not
7034 actually be the case. Of course, in all cases the client gets an
7035 appropriate error code.
7037 15. If a lookup fails in an ACL condition, a message about the failure may
7038 be available; it is used if testing the ACL cannot continue, because most
7039 such messages specify what the cause of the deferral is. However, some
7040 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7041 that caused an old message to be retained and used if a later statement
7042 caused a defer, replacing the real cause of the deferral.
7044 16. If an IP address had so many PTR records that the DNS lookup buffer
7045 was not large enough to hold them, Exim could crash while trying to process
7046 the truncated data. It now detects and logs this case.
7048 17. Further to 4.21/58, another change has been made: if (and only if) the
7049 first line of a message (the first header line) ends with CRLF, a bare LF
7050 in a subsequent header line has a space inserted after it, so as not to
7051 terminate the header.
7053 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7054 unnecessarily, used atoi() instead of strtol(), and didn't check the
7055 termination when getting file sizes from file names by regex.
7057 19. Completely re-implemented the support for maildirsize files, in the light
7058 of a number of problems with the previous contributed implementation
7059 (4.30/29). In particular:
7061 . If the quota is zero, the maildirsize file is maintained, but no quota is
7064 . If the maildir directory does not exist, it is created before any attempt
7065 to write a maildirsize file.
7067 . The quota value in the file is just a cache; if the quota is changed in
7068 the transport, the new value overrides.
7070 . A regular expression is available for excluding directories from the
7073 20. The autoreply transport checks the characters in options that define the
7074 message's headers; it allows continued headers, but it was checking with
7075 isspace() after an embedded newline instead of explicitly looking for a
7078 21. If all the "regular" hosts to which an address was routed had passed their
7079 expiry times, and had not reached their retry times, the address was
7080 bounced, even if fallback hosts were defined. Now Exim should go on to try
7083 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7084 equivalent code in the SMTP transport. Some hosts send humungous responses
7085 to HELO/EHLO, more than 1024 it seems.
7087 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7088 gives compiler warnings in some environments. I've now done it "properly",
7091 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7092 (because of problems with the built-in one) was declared to return uschar *
7093 instead of char *, causing compiler failure.
7095 25. Fixed a file descriptor leak when processing alias/forward files.
7097 26. Fixed a minor format string issue in dbfn.c.
7099 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7101 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7102 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7105 29. When a sender address is verified, it is cached, to save repeating the test
7106 when there is more than one recipient in a message. However, when the
7107 verification involves a callout, it is possible for different callout
7108 options to be set for different recipients. It is too complicated to keep
7109 track of this in the cache, so now Exim always runs a verification when a
7110 callout is required, relying on the callout cache for the optimization.
7111 The overhead is duplication of the address routing, but this should not be
7114 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7115 to be verified with callout=postmaster, and the main callout worked but the
7116 postmaster check failed, the verification correctly failed. However, if a
7117 subsequent RCPT command asked for sender verification *without* the
7118 postmaster check, incorrect caching caused this verification also to fail,
7121 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7122 it was not caching the DNS options (qualify_single, search_parents) that
7123 were used when the lookup failed. A subsequent lookup with different
7124 options therefore always gave the same answer, though there were cases
7125 where it should not have. (Example: a "domains = !$mx_any" option on a
7126 dnslookup router: the "domains" option is always processed without any
7127 widening, but the router might have qualify_single set.) Now Exim uses the
7128 cached value only when the same options are set.
7130 32. Added John Jetmore's "exipick" utility to the distribution.
7132 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7133 than a timeout (e.g. a certificate is required, and is not provided), an
7134 Exim server now closes the connection immediately. Previously it waited for
7135 the client to close - but if the client is SSL, it seems that they each
7136 wait for each other, leading to a delay before one of them times out.
7138 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7139 maintained 0.8.x compatibility because I don't think many are using it, and
7140 it is clearly obsolete.
7142 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7145 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7146 client certificate was expired. A simple patch fixes this, though I don't
7147 understand the full logic of why the verify callback is called multiple
7150 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7151 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7152 which causes problems with some clients (such as the Certicom SSL Plus
7153 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7154 disables the coutermeasure allowing Eudora to connect."
7156 38. Exim was not checking that a write() to a log file succeeded. This could
7157 lead to Bad Things if a log got too big, in particular if it hit a file
7158 size limit. Exim now panics and dies if it cannot write to a log file, just
7159 as it does if it cannot open a log file.
7161 39. Modified OS/Makefile-Linux so that it now contains
7163 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7165 The two -D definitions ensure that Exim is compiled with large file
7166 support, which makes it possible to handle log files that are bigger than
7169 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7170 instance) a domain was checked against a named list that involved a lookup,
7171 causing $domain_data to be set, then another domain was checked against the
7172 same list, then the first domain was re-checked, the value of $domain_data
7173 after the final check could be wrong. In particular, if the second check
7174 failed, it could be set empty. This bug probably also applied to
7177 41. The strip_trailing_dot option was not being applied to the address given
7178 with the -f command-line option.
7180 42. The code for reading a message's header from the spool was incrementing
7181 $received_count, but never initializing it. This meant that the value was
7182 incorrect (doubled) while delivering a message in the same process in which
7183 it was received. In the most common configuration of Exim, this never
7184 happens - a fresh exec is done - but it can happen when
7185 deliver_drop_privilege is set.
7187 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7188 it now includes up to 150 characters of the unexpected data in the log
7191 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7192 and building data strings. The size of both of these buffers was 10 000
7193 bytes - far larger than anybody would *ever* want, thought I. Needless to
7194 say, somebody hit the limit. I have increased the maximum line length to
7195 20 000 and the maximum data length of concatenated lines to 100 000. I have
7196 also fixed two bugs, because there was no checking on these buffers. Tsk,
7197 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7198 buffer is too small.
7200 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7201 lsearch lookups. Now it does.
7203 46. When parsing a route_list item in a manualroute router, a fixed-length
7204 buffer was used for the list of hosts. I made this 1024 bytes long,
7205 thinking that nobody would ever have a list of hosts that long. Wrong.
7206 Somebody had a whole pile of complicated expansion conditions, and the
7207 string was silently truncated, leading to an expansion error. It turns out
7208 that it is easier to change to an unlimited length (owing to other changes
7209 that have happened since this code was originally written) than to build
7210 structure for giving a limitation error. The length of the item that
7211 expands into the list of hosts is now unlimited.
7213 47. The lsearch lookup could not handle data where the length of text line was
7214 more than 4095 characters. Such lines were truncated, leading to shortened
7215 data being returned. It should now handle lines of any length.
7217 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7218 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7221 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7222 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7223 no longer happens in comment lines. A list of replacements is now placed
7224 at the head of all of the source files, except those whose only change is
7225 to replace PERL_COMMAND in the very first #! line.
7227 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7228 messages on the queue, with a bottom-up merge sort, using code contributed
7229 by Michael Haardt. This should make operations like -bp somewhat faster on
7230 large queues. It won't affect queue runners, except when queue_run_in_order
7233 51. Installed eximstats 1.31 in the distribution.
7235 52. Added support for SRV lookups to the dnslookup router.
7237 53. If an ACL referred to $message_body or $message_body_end, the value was not
7238 reset for any messages that followed in the same SMTP session.
7240 54. The store-handling optimization for building very long strings was not
7241 differentiating between the different store pools. I don't think this
7242 actually made any difference in practice, but I've tidied it.
7244 55. While running the routers to verify a sender address, $sender_address
7245 was still set to the sender address. This is wrong, because when routing to
7246 send a bounce to the sender, it would be empty. Therefore, I have changed
7247 it so that, while verifying a sender address, $sender_address is set to <>.
7248 (There is no change to what happens when verifying a recipient address.)
7250 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7251 target A or AAAA records (if not already returned) without resetting the
7252 qualify_single or search_parents options of the DNS resolver. These are
7253 inappropriate in this case because the targets of MX and SRV records must
7254 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7255 target that, when qualified, matched something in the local domain. These
7256 two options are now turned off when doing these lookups.
7258 57. It seems that at least some releases of Reiserfs (which does not have the
7259 concept of a fixed number of inodes) returns zero and not -1 for the
7260 number of available inodes. This interacted badly with check_spool_inodes,
7261 which assumed that -1 was the "no such thing" setting. What I have done is
7262 to check that the total number of inodes is greater than zero before doing
7263 the test of how many are available.
7265 58. When a "warn" ACL statement has a log_message modifier, the message is
7266 remembered, and not repeated. This is to avoid a lot of repetition when a
7267 message has many recipients that cause the same warning to be written.
7268 However, Exim was preserving the list of already written lines for an
7269 entire SMTP session, which doesn't seem right. The memory is now reset if a
7270 new message is started.
7272 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7273 debugging output unless log_rewrite was also set.
7275 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7276 of (int)(handle) when we know that handle contains (void *)(-1).
7278 61. The Exim daemon panic-logs an error return when it closes the incoming
7279 connection. However "connection reset by peer" seems to be common, and
7280 isn't really an error worthy of noting specially, so that particular error
7283 62. When Exim is trying to find all the local interfaces, it used to panic and
7284 die if the ioctl to get the interface flags failed. However, it seems that
7285 on at least one OS (Solaris 9) it is possible to have an interface that is
7286 included in the list of interfaces, but for which you get a failure error
7287 for this call. This happens when the interface is not "plumbed" into a
7288 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7289 failure of the "get flags" call assumes that the interface is down.
7291 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7292 makes life easier for people who are doing arithmetic on fields extracted
7293 from dates, where you often get leading zeros that should not be
7294 interpreted as octal.
7296 64. Added qualify_domain to the redirect router, to override the global
7299 65. If a pathologically long header line contained very many addresses (the
7300 report of this problem mentioned 10 000) and each of them was rewritten,
7301 Exim could use up a very large amount of memory. (It kept on making new
7302 copies of the header line as it rewrote, and never released the old ones.)
7303 At the expense of a bit more processing, the header rewriting function has
7304 been changed so that it no longer eats memory in this way.
7306 66. The generation of the Received: header has been moved from the time that a
7307 message starts to be received, to the time that it finishes. The timestamp
7308 in the Received: header should now be very close to that of the <= log
7309 line. There are two side-effects of this change:
7311 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7312 logged header lines no longer include the local Received: line, because
7313 it has not yet been created. The same applies to a copy of the message
7314 that is returned to a non-SMTP sender when a message is rejected.
7316 (b) When a filter file is tested using -bf, no additional Received: header
7317 is added to the test message. After some thought, I decided that this
7320 This change does not affect the value of $received_for. It is still set
7321 after address rewriting, but before local_scan() is called.
7323 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7325 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7326 gave an unhelpful panic error message, and a defer error. I have managed to
7327 change this behaviour so that it now rejects any supplied certificate,
7328 which seems right, as the list of acceptable certificates is empty.
7330 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7331 gave an unhelpful defer error. I have not managed to make this reject any
7332 supplied certificates, but the error message it gives is "no certificate
7333 supplied", which is not helpful.
7335 70. exigrep's output now also includes lines that are not associated with any
7336 message, but which match the given pattern. Implemented by a patch from
7337 Martin Sluka, which also tidied up the Perl a bit.
7339 71. Recipient callout verification, like sender verification, was using <> in
7340 the MAIL FROM command. This isn't really the right thing, since the actual
7341 sender may affect whether the remote host accepts the recipient or not. I
7342 have changed it to use the actual sender in the callout; this means that
7343 the cache record is now keyed on a recipient/sender pair, not just the
7344 recipient address. There doesn't seem to be a real danger of callout loops,
7345 since a callout by the remote host to check the sender would use <>.
7346 [SEE ABOVE: changed after hitting problems.]
7348 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7349 temporary errors. However, in the case of such a code being given after
7350 the end of a data transmission (i.e. after ".") Exim was failing to write
7351 a retry record for the message. (Yes, there was some broken host that was
7352 actually sending 8xx at this point.)
7354 73. An unknown lookup type in a host list could cause Exim to panic-die when
7355 the list was checked. (An example that provoked this was putting <; in the
7356 middle of a list instead of at the start.) If this happened during a DATA
7357 ACL check, a -D file could be left lying around. This kind of configuration
7358 error no longer causes Exim to die; instead it causes a defer error. The
7359 incident is still logged to the main and panic logs.
7361 74. Buglet left over from Exim 3 conversion. The message "too many messages
7362 in one connection" was written to the rejectlog but not the mainlog, except
7363 when address rewriting (yes!) was being logged.
7365 75. Added write_rejectlog option.
7367 76. When a system filter was run not as root (that is, when system_filter_user
7368 was set), the values of the $n variables were not being returned to the
7369 main process; thus, they were not subsequently available in the $sn
7372 77. Added +return_path_on_delivery log selector.
7374 78. A connection timeout was being treated differently from recipients deferred
7375 when testing hosts_max_try with a message that was older than the host's
7376 retry timeout. (The host should not be counted, thus allowing all hosts to
7377 be tried at least once before bouncing.) This may have been the cause of an
7378 occasionally reported bug whereby a message would remain on the queue
7379 longer than the retry timeout, but would be bounced if a delivery was
7380 forced. I say "may" because I never totally pinned down the problem;
7381 setting up timeout/retry tests is difficult. See also the next item.
7383 79. The ultimate address timeout was not being applied to errors that involved
7384 a combination of host plus message (for example, a timeout on a MAIL
7385 command). When an address resolved to a number of possible hosts, and they
7386 were not all tried for each delivery (e.g. because of hosts_max_try), a
7387 message could remain on the queue longer than the retry timeout.
7389 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7392 81. Fixed an obscure SMTP outgoing bug which required at least the following
7393 conditions: (a) there was another message waiting for the same server;
7394 (b) the server returned 5xx to all RCPT commands in the first message so
7395 that the message was not completed; (c) the server dropped the connection
7396 or gave a negative response to the RSET that Exim sends to abort the
7397 transaction. The observed case was a dropped connection after DATA that had
7398 been sent in pipelining mode. That is, the server had advertised PIPELINING
7399 but was not implementing it correctly. The effect of the bug was incorrect
7400 behaviour, such as trying another host, and this could lead to a crash.
7406 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7407 and daemon.c were passed as pointers to ints; they should have been
7408 pointers to socklen_t variables (which are typically unsigned ints).
7410 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7413 3. Fixed a really odd bug that affected only the testing scheme; patching a
7414 certain fixed string in the binary changed the value of another string that
7415 happened to be identical to the end of the original first string.
7417 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7418 name", it returns that address as the IP address. On some operating
7419 systems (e.g. Solaris), it also passes back the IP address string as the
7420 "host name". However, on others (e.g. Linux), it passes back an empty
7421 string. Exim wasn't checking for this, and was changing the host name to an
7422 empty string, assuming it had been canonicalized.
7424 5. Although rare, it is permitted to have more than one PTR record for a given
7425 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7426 all the names associated with an address, because they do in Solaris.
7427 However, it seems that they do not in Linux for data that comes from the
7428 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7429 I found this out when I moved to a new Linux workstation and tried to run
7430 the Exim test suite.
7432 To get round this problem I have changed the code so that it now does its
7433 own call to the DNS to look up PTR records when searching for a host name.
7434 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7435 addresses that are only in /etc/hosts are still found.
7437 This behaviour is, however, controlled by an option called host_lookup_
7438 order, which defaults to "bydns:byaddr". If people want to use the other
7439 order, or indeed, just use one or the other means of lookup, they can
7440 specify it in this variable.
7442 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7443 some operating systems, this comes back from gethostbyaddr() as an empty
7444 string, and this is what Exim used to test for. However, it seems that in
7445 other systems, "." is yielded. Exim now tests for this case too.
7447 7. The values of check_spool_space and check_log_space are now held internally
7448 as a number of kilobytes instead of an absolute number of bytes. If a
7449 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7450 kilobyte. This means that much larger values can be stored.
7452 8. Exim monitor: an attempt to get the action menu when not actually pointing
7453 at a message produces an empty menu entitled "No message selected". This
7454 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7455 no entries in it ("Shell widget menu has zero width and/or height"). So I
7456 have added a single, blank menu entry in this case.
7458 9. Added ${quote_local_part.
7460 10. MIME decoding is now applied to the contents of Subject: header lines when
7463 11. Now that a reference to $sender_host_address automatically causes a reverse
7464 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7465 host lookup before query-style lookups in lists that might use this
7466 variable. This has therefore been abolished, and the "net-" prefix is no
7467 longer necessary for query-style lookups.
7469 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7470 have been a typo for LFLAGS, so it has been changed.
7472 13. The install script calls Exim with "-C /dev/null" in order to find the
7473 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7474 to be output. However, since Exim outputs its version number before the
7475 error, it didn't break the script. It just looked ugly. I fixed this by
7476 always allowing "-C /dev/null" if the caller is root.
7478 14. Ignore overlarge ACL variable number when reading spool file - insurance
7479 against a later release with more variables having written the file.
7481 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7482 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7483 this, as well as the form without the "IPv6" on the front.
7485 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7486 OS/Makefile-Darwin file.
7488 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7489 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7491 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7492 variable. However, some OS use ints for inodes. I've added cast to long int
7493 to get rid of the compiler warning.
7495 19. I had forgotten to lock out "/../" in configuration file names when
7496 ALT_CONFIG_PREFIX was set.
7498 20. Routers used for verification do not need to specify transports. However,
7499 if such a router generated a host list, and callout was configured, Exim
7500 crashed, because it could not find a port number from the (non-existent)
7501 transport. It now assumes port 25 in this circumstance.
7503 21. Added the -t option to exigrep.
7505 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7506 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7507 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7508 them is set, LOOKUP_LSEARCH is forced.
7510 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7512 24. Added sender and host information to the "rejected by local_scan()" log
7513 line; previously there was no indication of these.
7515 25. Added .include_if_exists.
7517 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7518 Linux. It turns out that not all file systems support this. Apparently some
7519 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7520 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7521 not supported on the file descriptor, is now ignored when Exim is trying to
7522 sync a directory. This applies only to Linux.
7524 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7526 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7527 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7530 29. Added support for maildirsize files from supplied patch (modified a bit).
7532 30. The use of :fail: followed by an empty string could lead Exim to respond to
7533 sender verification failures with (e.g.):
7535 550 Verification failed for <xxx>
7536 550 Sender verify failed
7538 where the first response line was missing the '-' that indicates it is not
7539 the final line of the response.
7541 31. The loop for finding the name of the user that called Exim had a hardwired
7542 limit of 10; it now uses the value of finduser_retries, which is used for
7543 all other user lookups.
7545 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7548 33. Exim was neglecting to zero errno before one call of strtol() when
7549 expanding a string and expecting an integer value. On some systems this
7550 resulted in spurious "integer overflow" errors. Also, it was casting the
7551 result into an int without checking.
7553 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7554 did not work. The code looks as if it has *never* worked, though it appears
7555 to have been documented since at least release 1.62. I have made it work.
7557 35. The "timeout_DNS" error in retry rules, also documented since at least
7558 1.62, also never worked. As it isn't clear exactly what this means, and
7559 clearly it isn't a major issue, I have abolished the feature by treating it
7560 as "timeout", and writing a warning to the main and panic logs.
7562 36. The display of retry rules for -brt wasn't always showing the error code
7565 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7566 timeout_connect_A, timeout_connect_MX.
7568 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7569 to the empty sender.
7571 39. The daemon was not analysing the content of -oX till after it had closed
7572 stderr and disconnected from the controlling terminal. This meant that any
7573 syntax errors were only noted on the panic log, and the return code from
7574 the command was 0. By re-arranging the code a little, I've made the
7575 decoding happen first, so such errors now appear on stderr, and the return
7576 code is 1. However, the actual setting up of the sockets still happens in
7577 the disconnected process, so errors there are still only recorded on the
7580 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7581 connections (as happens on some IP stacks) was logged at start up time as
7582 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7583 it from "IPv6 and IPv4", which means that two separate sockets are being
7586 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7587 says whether AF_INET or AF_INET6 was passed as an argument.
7589 42. Exiwhat output was messed up when time zones were included in log
7592 43. Exiwhat now gives more information about the daemon's listening ports,
7593 and whether -tls-on-connect was used.
7595 44. The "port" option of the smtp transport is now expanded.
7597 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7598 silently ignored. Now an error message is written to the main and panic
7601 46. There's a new ACL modifier called "logwrite" which writes to a log file
7602 as soon as it is encountered.
7604 47. Added $local_user_uid and $local_user_gid at routing time.
7606 48. Exim crashed when trying to verify a sender address that was being
7609 49. Exim was recognizing only a space character after ".include". It now also
7610 recognizes a tab character.
7612 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7613 extracting the relevant information from the specification. The man page no
7614 longer contains scrambled data for the -d option, and I've added a section
7615 at the front about calling Exim under different names.
7617 51. Added "extra_headers" argument to the "mail" command in filter files.
7619 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7622 53. Installed eximstats 1.29.
7624 54. Added transport_filter_timeout as a generic transport option.
7626 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7627 Cc: header lines. This was required by RFC 822, but it not required by RFC
7630 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7631 incoming messages that did not have them. Now it does so only if the
7632 message originates locally, that is, if there is no associated remote host
7633 address. When Resent- header lines are present, this applies to the Resent-
7634 lines rather than the non-Resent- lines.
7636 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7637 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7639 58. Messages for configuration errors now include the name of the main
7640 configuration file - useful now that there may be more than one file in a
7641 list (.included file names were always shown).
7643 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7644 for those rare installations that do not start the daemon as root or run it
7645 setuid root. I've cut out the call to initgroups() if the daemon is not
7648 60. The Exim user and group can now be bound into the binary as text strings
7649 that are looked up at the start of Exim's processing.
7651 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7653 62. Added $mailstore_basename variable.
7655 63. Installed patch to sieve.c from Michael Haardt.
7657 64. When Exim failed to open the panic log after failing to open the main log,
7658 the original message it was trying to log was written to stderr and debug
7659 output, but if they were not available (the usual case in production), it
7660 was lost. Now it is written to syslog before the two lines that record the
7661 failures to open the logs.
7663 65. Users' Exim filters run in subprocesses under the user's uid. It is
7664 possible for a "deliver" command or an alias in a "personal" command to
7665 provoke an address rewrite. If logging of address rewriting is configured,
7666 this fails because the process is not running as root or exim. There may be
7667 a better way of dealing with this, but for the moment (because 4.30 needs
7668 to be released), I have disabled address rewrite logging when running a
7669 filter in a non-root, non-exim process.
7675 1. The buildconfig auxiliary program wasn't quoting the value set for
7676 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7677 not defined. This bug was masked in 4.22 by the effect that was fixed in
7680 2. Some messages that were rejected after a message id was allocated were
7681 shown as "incomplete" by exigrep. It no longer does this for messages that
7682 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7684 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7685 have allow_domain_literals set, the ID did not get logged in the <= line.
7686 Domain literals are now always recognized in Message-ID: header lines.
7688 4. The first argument for a ${extract expansion item is the key name or field
7689 number. Leading and trailing spaces in this item were not being ignored,
7690 causing some misleading effects.
7692 5. When deliver_drop_privilege was set, single queue runner processes started
7693 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7694 same command in the process it spins off) were not dropping privilege.
7696 6. When the daemon running as "exim" started a queue runner, it always
7697 re-executed Exim in the spun-off process. This is a waste of effort when
7698 deliver_drop_privilege is set. The new process now just calls the
7699 queue-runner function directly.
7705 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7708 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7709 ignored. Though the use of -oP was forcing the writing of a pid file, it
7710 was always written to the default place.
7712 3. If the message "no IP address found for host xxxx" is generated during
7713 incoming verification, it is now followed by identification of the incoming
7714 connection (so you can more easily find what provoked it).
7716 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7718 5. Added some features to "harden" Exim a bit more against certain attacks:
7720 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7721 be put in Local/Makefile. This is like the never_users runtime option,
7722 but it cannot be overridden. The default setting is "root".
7724 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7725 prefix string with which any file named in a -C command line option
7728 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7729 is retained for -C and -D only if the caller of Exim is root. Without
7730 it, the exim user may also use -C and -D and retain privilege.
7732 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7733 command line option is disabled.
7735 6. Macro names set by the -D option must start with an upper case letter, just
7736 like macro names defined in the configuration file.
7738 7. Added "dereference=" facility to LDAP.
7740 8. Two instances of the typo "uknown" in the source files are fixed.
7742 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7743 the Configure-Makefile script screwed up while processing it.
7745 10. Incorporated PCRE 4.4.
7747 11. The SMTP synchronization check was not operating right at the start of an
7748 SMTP session. For example, it could not catch a HELO sent before the client
7749 waited for the greeting. There is now a check for outstanding input at the
7750 point when the greeting is written. Because of the duplex, asynchronous
7751 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7752 way, but not yet received, when the check is performed.
7754 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7755 on TCP/IP sockets, because this apparently causes some broken clients to
7758 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7759 unchanged) from the Cygwin maintainer.
7761 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7762 was supported instead of testing for mbx. Effectively a typo.
7764 15. The spa authenticator server code was not checking that the input it
7765 received was valid base64.
7767 16. The debug output line for the "set" modifier in ACLs was not showing the
7768 name of the variable that was being set.
7770 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7772 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7773 lookup on its own. Something else was needed to trigger the lookup. For
7774 example, a match in host_lookup or the need for a host name in a host list.
7775 Now, if $sender_host_name is referenced and the host name has not yet been
7776 looked up, a lookup is performed. If the lookup fails, the variable remains
7777 empty, and $host_lookup_failed is set to "1".
7779 19. Added "eqi" as a case-independent comparison operator.
7781 20. The saslauthd authentication condition could segfault if neither service
7782 nor realm was specified.
7784 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7785 error message that was logged was misleading, and incoming SMTP
7786 connections were dropped. The message is now more accurate, and temporary
7787 errors are given to SMTP connections.
7789 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7790 (except RSET) with a 421 error, until QUIT is received. However, it was
7791 failing to send a response to QUIT.
7793 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7794 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7795 ACL is now run after the helo_try_verify_hosts code.
7797 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7798 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7799 case-independent, so other case variants are also recognized). Apparently
7800 some systems use these upper case variants.
7802 25. If more than two messages were waiting for the same host, and a transport
7803 filter was specified for the transport, Exim sent two messages over the
7804 same TCP/IP connection, and then failed with "socket operation on non-
7805 socket" when it tried to send the third.
7807 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7809 27. The extern definition of crypt16() in expand.c was not being excluded when
7810 the OS had its own crypt16() function.
7812 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7813 as a preferred synonym for return_size_limit, both as an option and as an
7816 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7818 30. Changed the default configuration ACL to relax the local part checking rule
7819 for addresses that are not in any local domains. For these addresses,
7820 slashes and pipe symbols are allowed within local parts, but the sequence
7821 /../ is explicitly forbidden.
7823 31. SPA server authentication was not clearing the challenge buffer before
7826 32. log_message in a "warn" ACL statement was writing to the reject log as
7827 well as to the main log, which contradicts the documentation and doesn't
7828 seem right (because no rejection is happening). So I have stopped it.
7830 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7831 However, I am unable to do any testing of this.
7833 34. Fixed an infelicity in the appendfile transport. When checking directories
7834 for a mailbox, to see if any needed to be created, it was accidentally
7835 using path names with one or more superfluous leading slashes; tracing
7836 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7838 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7839 contents are added to the log line that is written for every discarded
7840 recipient. (Previously a log_message setting was ignored.)
7842 36. The ${quote: operator now quotes the string if it is empty.
7844 37. The install script runs exim in order to find its version number. If for
7845 some reason other than non-existence or emptiness, which it checks, it
7846 could not run './exim', it was installing it with an empty version number,
7847 i.e. as "exim-". This error state is now caught, and the installation is
7850 38. An argument was missing from the function that creates an error message
7851 when Exim fails to connect to the socket for saslauthd authentication.
7852 This could cause Exim to crash, or give a corrupted message.
7854 39. Added isip, isip4, and isip6 to ${if conditions.
7856 40. The ACL variables $acl_xx are now saved with the message, and can be
7857 accessed later in routers, transports, and filters.
7859 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7860 strings in the file are not string-expanded.
7862 42. If a MAIL command specified a SIZE value that was too large to fit into an
7863 int variable, the check against message_size_limit failed. Such values are
7864 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7865 day this will have to be increased, but I don't think I want to be around
7866 when emails are that large.
7873 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7874 iconv() is not standard in FreeBSD.
7876 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7877 IPv6 enabled. The observed symptom was a segmentation fault on return from
7878 the function os_common_find_running_interfaces() in src/os.c.
7880 3. In the check_special_case() function in daemon.c I had used "errno" as an
7881 argument name, which causes warnings on some systems. This was basically a
7882 typo, since it was named "eno" in the comments!
7884 4. The code that waits for the clock to tick (at a resolution of some fraction
7885 of a second) so as to ensure message-id uniqueness was always waiting for
7886 at least one whole tick, when it could have waited for less. [This is
7887 almost certainly not relevant at current processor speeds, where it is
7888 unlikely to ever wait at all. But we try to future-proof.]
7890 5. The function that sleeps for a time interval that includes fractions of a
7891 second contained a race. It did not block SIGALRM between setting the
7892 timer, and suspending (a couple of lines later). If the interval was short
7893 and the sigsuspend() was delayed until after it had expired, the suspension
7894 never ended. On busy systems this could lead to processes getting stuck for
7897 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7898 process, before it forks any delivery processes. The open lookup caching
7899 mechanism meant that the open file or database connection was passed into
7900 the delivery process. The problem was that delivery processes always tidy
7901 up cached lookup data. This could cause a problem for the next delivery
7902 process started by the queue runner, because the external queue runner
7903 process does not know about the closure. So the next delivery process
7904 still has data in the lookup cache. In the case of a file lookup, there was
7905 no problem because closing a file descriptor in a subprocess doesn't affect
7906 the parent. However, if the lookup was caching a connection to a database,
7907 the connection was closed, and the second delivery process was likely to
7908 see errors such as "PGSQL: query failed: server closed the connection
7909 unexpectedly". The problem has been fixed by closing all cached lookups
7910 in a queue runner before running a delivery process.
7912 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7913 seem to have the "const" qualifier which it has on other OS. I've
7916 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7917 *of the same type* (client or server) with the same public name that an
7918 error should be diagnosed.
7920 9. When Exim looked up a host name for an IP address, but failed to find the
7921 original IP address when looking up the host name (a safety check), it
7922 output the message "<ip address> does not match any IP for NULL", which was
7923 confusing, to say the least. The bug was that the host name should have
7924 appeared instead of "NULL".
7926 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7927 user that is built into the binary, and the -C or -D options is used, root
7928 privilege is dropped before the configuration file is read. In addition,
7929 logging is switched to stderr instead of the normal log files. If the
7930 configuration then re-defines the Exim user, the unprivileged environment
7931 is probably not what is expected, so Exim logs a panic warning message (but
7934 However, if deliver_drop_privilege is set, the unprivileged state may well
7935 be exactly what is intended, so the warning has been cut out in that case,
7936 and Exim is allowed to try to write to its normal log files.
7942 1. smtp_return_error_details was not giving details for temporary sender
7943 or receiver verification errors.
7945 2. Diagnose a configuration error if two authenticators have the same public
7948 3. Exim used not to create the message log file for a message until the first
7949 delivery attempt. This could be confusing when incoming messages were held
7950 for policy or load reasons. The message log file is now created at the time
7951 the message is received, and an initial "Received" line is written to it.
7953 4. The automatically generated man page for command line options had a minor
7954 bug that caused no ill effects; however, a more serious problem was that
7955 the procedure for building the man page automatically didn't always
7956 operate. Consequently, release 4.20 contains an out-of-date version. This
7957 shouldn't happen again.
7959 5. When building Exim with embedded Perl support, the script that builds the
7960 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7961 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7963 6. The freeze_tell option was not being used for messages that were frozen on
7964 arrival, either by an ACL or by local_scan().
7966 7. Added the smtp_incomplete_transaction log selector.
7968 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7969 was accepting AUTH without a new EHLO.
7971 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7972 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7975 10. Logging of TCP/IP connections (when configured) now happens in the main
7976 daemon process instead of the child process, so that the TCP/IP connection
7977 count is more accurate (but it can never be perfect).
7979 11. The use of "drop" in a nested ACL was not being handled correctly in the
7980 outer ACL. Now, if condition failure induced by the nested "drop" causes
7981 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7982 or "require"), the connection is dropped.
7984 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7985 that yield "discard" can now be used with an "accept" or a "discard" verb,
7986 but an error is generated for any others (because I can't see a useful way
7987 to define what should happen).
7989 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7990 are now processed in the same way as lines in the Exim configuration file.
7991 In particular, continuation lines are supported.
7993 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7995 15. Added -ti meaning -t -i.
7997 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7998 domains, and warn by logging if others are found.
8000 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8001 returned from GIFCONF when Exim is trying to find the list of interfaces on
8002 a host. The code in os.c has been modified to copy each ifreq to an aligned
8003 structure in all cases.
8005 Also, in some cases, the returned ifreq's were being copied to a 'struct
8006 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8007 means the last couple of bytes of an IPv6 address could be chopped if the
8008 ifreq contained only a normal sockaddr (14 bytes storage).
8010 18. Named domain lists were not supported in the hosts_treat_as_local option.
8011 An entry such as +xxxx was not recognized, and was treated as a literal
8014 19. Ensure that header lines added by a DATA ACL are included in the reject log
8015 if the ACL subsequently rejects the message.
8017 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8018 MD5 (which is deprecated).
8020 21. When testing a filter file using -bf, Exim was writing a message when it
8021 took the sender from a "From " line in the message, but it was not doing so
8022 when it took $return_path from a Return-Path: header line. It now does.
8024 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8025 with a valid header line field name (a series of printing characters
8026 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8028 23. Changed "disc" in the source to "disk" to conform to the documentation and
8029 the book and for uniformity.
8031 24. Ignore Sendmail's -Ooption=value command line item.
8033 25. When execve() failed while trying to run a command in a pipe transport,
8034 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8035 could be confused with a return value of 69 from the command itself. This
8036 has been changed to 127, the value the shell returns if it is asked to run
8037 a non-existent command. The wording for the related log line suggests a
8038 non-existent command as the problem.
8040 26. If received_header_text expands to an empty string, do not add a Received:
8041 header line to the message. (Well, it adds a token one on the spool, but
8042 marks it "old" so that it doesn't get used or transmitted.)
8044 27. Installed eximstats 1.28 (addition of -nt option).
8046 28. There was no check for failure on the call to getsockname() in the daemon
8047 code. This can fail if there is a shortage of resources on the system, with
8048 ENOMEM, for example. A temporary error is now given on failure.
8050 29. Contrary to the C standard, it seems that in some environments, the
8051 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8052 program. Exim now does this explicitly; it affects the formatting of
8053 timestamps using strftime().
8055 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8056 complaints. I've now initialized all the variables, to avoid this.
8058 32. Header lines added by a system filter were not being "seen" during
8059 transport-time rewrites.
8061 33. The info_callback() function passed to OpenSSL is set up with type void
8062 (*)(SSL *, int, int), as described somewhere. However, when calling the
8063 function (actually a macro) that sets it up, the type void(*)() is
8064 expected. I've put in a cast to prevent warnings from picky compilers.
8066 34. If a DNS black list lookup found a CNAME record, but there were no A
8067 records associated with the domain it pointed at, Exim crashed.
8069 35. If a DNS black list lookup returned more than one A record, Exim ignored
8070 all but the first. It now scans all returned addresses if a particular IP
8071 value is being sought. In this situation, the contents of the
8072 $dnslist_value variable are a list of all the addresses, separated by a
8075 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8076 to accept a host name and all its aliases if the forward lookup for any of
8077 them yielded the IP address of the incoming connection. Now it accepts only
8078 those names whose forward lookup yields the correct IP address. Any other
8079 names are discarded. This closes a loophole whereby a rogue DNS
8080 administrator could create reverse DNS records to break through a
8081 wildcarded host restriction in an ACL.
8083 37. If a user filter or a system filter that ran in a subprocess used any of
8084 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8085 the wrong values were passed to the pipe command ($thisaddress had the
8086 value of $0, $0 had the value of $1, etc). This bug was introduced by
8087 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8089 38. Improved the line breaking for long SMTP error messages from ACLs.
8090 Previously, if there was no break point between 40 and 75 characters, Exim
8091 left the rest of the message alone. Two changes have been made: (a) I've
8092 reduced the minimum length to 35 characters; (b) if it can't find a break
8093 point between 35 and 75 characters, it looks ahead and uses the first one
8094 that it finds. This may give the occasional overlong line, but at least the
8095 remaining text gets split now.
8097 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8098 file descriptors might be low, and that setting 1000 would always raise it.
8099 It turns out that in some environments, the limit is already over 1000 and
8100 that lowering it causes trouble. So now Exim takes care not to decrease it.
8102 40. When delivering a message, the value of $return_path is set to $sender_
8103 address at the start of routing (routers may change the value). By an
8104 oversight, this default was not being set up when an address was tested by
8105 -bt or -bv, which affected the outcome if any router or filter referred to
8108 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8109 log only when "message" or "log_message" are set. However, if one of the
8110 conditions was an address verification, or a call to a nested ACL, the
8111 messages generated by the underlying test were being passed through. This
8112 no longer happens. The underlying message is available in $acl_verify_
8113 message for both "message" and "log_message" expansions, so it can be
8114 passed through if needed.
8116 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8117 new expansion $bh_ to give the encoded byte string without charset
8118 translation. Translation happens only if iconv() is available; HAVE_ICONV
8119 indicates this at build time. HEADERS_CHARSET gives the charset to
8120 translate to; headers_charset can change it in the configuration, and
8121 "headers charset" can change it in an individual filter file.
8123 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8124 that creates RFC 2047 encoded "words" labels them as that charset instead
8125 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8126 expansion operator; (ii) when Exim creates a From: line for a local
8127 message; (iii) when a header line is rewritten to include a "phrase" part.
8129 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8130 buggy, causing it to skip the first lines of messages whose message ID
8131 ended in 'D'. This would not have bitten before Exim release 4.14, because
8132 message IDs were unlikely to end in 'D' before then. The effect was to have
8133 incorrect size information for certain domains.
8135 45. #include "config.h" was missing at the start of the crypt16.c module. This
8136 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8139 46. If there was a timeout during a "random" callout check, Exim treated it as
8140 a failure of the random address, and carried on sending RSET and the real
8141 address. If the delay was just some slowness somewhere, the response to the
8142 original RCPT would be taken as a response to RSET and so on, causing
8143 mayhem of various kinds.
8145 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8146 when I implemented it. It didn't allow for the fact that some option values
8147 may legitimately be negative (e.g. size_addition), and it didn't even do
8148 the right test for positive values.
8150 48. Domain names in DNS records are case-independent. Exim always looks them up
8151 in lower case. Some resolvers return domain names in exactly the case they
8152 appear in the zone file, that is, they may contain uppercase letters. Not
8153 all resolvers do this - some return always lower case. Exim was treating a
8154 change of case by a resolver as a change of domain, similar to a widening
8155 of a domain abbreviation. This triggered its re-routing code and so it was
8156 trying to route what was effectively the same domain again. This normally
8157 caused routing to fail (because the router wouldn't handle the domain
8158 twice). Now Exim checks for this case specially, and just changes the
8159 casing of the domain that it ultimately uses when it transmits the message
8162 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8165 50. If a filter generated a file delivery with a non-absolute name (possible if
8166 no home directory exists for the router), the forbid_file option was not
8169 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8170 the existing equality matching.
8172 52. Exim was using ints instead of ino_t variables in some places where it was
8173 dealing with inode numbers.
8175 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8176 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8177 called TMPDIR, and if it finds it is different, it changes its value.
8179 54. The smtp_printf() function is now made available to local_scan() so
8180 additional output lines can be written before returning. There is also an
8181 smtp_fflush() function to enable the detection of a dropped connection.
8182 The variables smtp_input and smtp_batched_input are exported to
8185 55. Changed the default runtime configuration: the message "Unknown user"
8186 has been removed from the ACL, and instead placed on the localuser router,
8187 using the cannot_route_message feature. This means that any verification
8188 failures that generate their own messages won't get overridden. Similarly,
8189 the "Unrouteable address" message that was in the ACL for unverifiable
8190 relay addresses has also been removed.
8192 56. Added hosts_avoid_esmtp to the smtp transport.
8194 57. The exicyclog script was not checking for the esoteric option
8195 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8196 will work only if exicyclog is run under the appropriate euid.
8198 58. Following a discussion on the list, the rules by which Exim recognises line
8199 endings on incoming messages have been changed. The -dropcr and drop_cr
8200 options are now no-ops, retained only for backwards compatibility. The
8201 following line terminators are recognized: LF CRLF CR. However, special
8202 processing applies to CR:
8204 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8205 nor a local message in the state where . is a terminator.
8207 (ii) If a bare CR is encountered in a header line, an extra space is added
8208 after the line terminator so as not to end the header. The reasoning
8209 behind this is that bare CRs in header lines are most likely either
8210 to be mistakes, or people trying to play silly games.
8212 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8213 was being incorrectly given as 18 bytes larger than it should have been.
8214 This is a VOB (very old bug).
8216 60. This may never have affected anything current, but just in case it has:
8217 When the local host is found other than at the start of a list of hosts,
8218 the local host, those with the same MX, and any that follow, are discarded.
8219 When the list in question was part of a longer list of hosts, the following
8220 hosts (not currently being processed) were also being discarded. This no
8221 longer happens. I'm not sure if this situation could ever has previously
8224 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8227 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8228 header. This is something that is recommended in a new Internet Draft, and
8229 is something that is documented as being done by Sendmail. There are two
8230 possible values. For messages generated by the autoreply transport, Exim
8233 Auto-Submitted: auto-replied
8235 whereas for all other generated messages (e.g. bounces) it adds
8237 Auto-Submitted: auto-generated
8239 63. The "personal" condition in filters now includes a test for the
8240 Auto-Submitted: header. If it contains the string "auto-" the message it
8241 not considered personal.
8243 64. Added rcpt_include_affixes as a generic transport option.
8245 65. Added queue_only_override (default true).
8247 66. Added the syslog_duplication option.
8249 67. If what should have been the first header line of a message consisted of
8250 a space followed by a colon, Exim was mis-interpreting it as a header line.
8251 It isn't of course - it is syntactically invalid and should therefore be
8252 treated as the start of the message body. The misbehaviour could have
8253 caused a number of strange effects, including loss of data in subsequent
8254 header lines, and spool format errors.
8256 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8257 client host had authenticated. This control can now be exercised by an ACL
8258 for more flexibility.
8260 69. By default, callouts do not happen when testing with -bh. There is now a
8261 variant, -bhc, which does actually run the callout code, including
8262 consulting and updating the callout cache.
8264 70. Added support for saslauthd authentication, courtesy of Alexander
8267 71. If statvfs() failed on the spool or log directories while checking their
8268 size for availability, Exim confusingly gave the error "space shortage".
8269 Furthermore, in debugging mode it crashed with a floating point exception.
8270 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8271 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8272 serious problem, Exim now writes to the main and panic logs when this
8273 happens, with details of the failure. It then refuses to accept the
8274 incoming message, giving the message "spool directory problem" or "log
8275 directory problem" with a 421 code for SMTP messages.
8277 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8278 0, 1, and 2 exist, because some OS complain for execs without them (see
8279 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8280 descriptors. However, the code omitted to check that the open succeeded,
8281 causing mysterious errors if for some reason the permissions on /dev/null
8282 got screwed. Now Exim writes a message to the main and panic logs, and
8283 bombs out if it can't open /dev/null.
8285 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8286 interact so that it is all more flexible. It is supposed to remain
8287 backwards compatible. Also added extra_local_interfaces.
8289 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8290 to bomb out with an assertion failure - to the client this appears as a
8291 connection drop. This problem occurs in the part of the code that was taken
8292 from the Samba project. Fortunately, the assertion is in a very simple
8293 function, so I have fixed this by reproducing the function inline in the
8294 one place where it is called, and arranging for authentication to fail
8295 instead of killing the process with assert().
8297 75. The SPA client code was not working when the server requested OEM rather
8298 than Unicode encoding.
8300 76. Added code to make require_files with a specific uid setting more usable in
8301 the case where statting the file as root fails - usually a non-root-mounted
8302 NFS file system. When this happens and the failure is EACCES, Exim now
8303 forks a subprocess and does the per-uid checking as the relevant uid.
8305 77. Added process_log_path.
8307 78. If log_file_path was not explicitly set, a setting of check_log_space or
8308 check_log_inodes was ignored.
8310 79. If a space check for the spool or log partitions fails, the incident is now
8311 logged. Of course, in the latter case the data may get lost...
8313 80. Added the %p formatting code to string_format() so that it can be used to
8314 print addresses in debug_print(). Adjusted all the address printing in the
8315 debugging in store.c to use %p rather than %d.
8317 81. There was a concern that a line of code in smtp_in.c could overflow a
8318 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8319 initially expressed, the concern was not well-founded, because trailing
8320 spaces are removed early. However, if the trailing spaces were followed by
8321 a NULL, they did not get removed, so the overflow was possible. Two fixes
8324 (a) I re-wrote the offending code in a cleaner fashion.
8325 (b) If an incoming SMTP command contains a NULL character, it is rejected
8328 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8329 runs initgroups(), so that if the Exim user is in any additional groups,
8330 they will be used during message reception.
8336 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob