1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes. Previously debugging stopped any time Exim
74 re-execs, or for processing a queued message.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
86 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
87 supplied as the db file-pair basename. Previously, if a directory
88 path was given, for example via the autoreply "once" option, the DB
89 file.pag and file.dir files would be created in that directory's
92 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
93 "taint" log_selector. These were previously deprecated.
95 JH/22 Fix static address-list lookups to properly return the matched item.
96 Previously only the domain part was returned.
98 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
99 the call into OpenSSL to send a TLS Close was being repeated; this
100 resulted in the library waiting for the peer's Close. If that was never
101 sent we waited forever. Fix by tracking send calls.
103 JH/24 The ${run} expansion item now expands its command string elements after
104 splitting. Previously it was before; the new ordering makes handling
105 zero-length arguments simpler. The old ordering can be obtained by
106 appending a new option "preexpand", after a comma, to the "run".
108 JH/25 Taint-check exec arguments for transport-initiated external processes.
109 Previously, tainted values could be used. This affects "pipe", "lmtp" and
110 "queryprogram" transport, transport-filter, and ETRN commands.
111 The ${run} expansion is also affected: in "preexpand" mode no part of
112 the command line may be tainted, in default mode the executable name
115 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
116 the the facility was not passed across execs, and only the first message
117 passed over a connection could use BDAT; any further ones using DATA.
119 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
120 uses $sending_ip_address and an interface is specified.
121 Previously any use of the local address in the EHLO name disabled
122 PIPECONNECT, the common case being to use the rDNS of it.
124 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
125 resumption. Previously verify failed because no certificate status is
126 passed on the wire for the restarted session. Fix by using the recorded
127 ocsp status of the stored session for the new connection.
129 JH/29 TLS resumption: the key for session lookup in the client now includes
130 more info that a server could potentially use in configuring a TLS
131 session, avoiding oferring mismatching sessions to such a server.
132 Previously only the server IP was used.
134 JH/30 Fix string_copyn() for limit greater than actual string length.
135 Previously the copied amount was the limit, which could result in a
136 overlapping memcpy for newly allocated destination soon after a
137 source string shorter than the limit. Found/investigated by KM.
139 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
140 close; it may be needed for a subsequent connection. This caused a
141 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
147 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
148 related applications. Previously an "H" was used where available info
149 says that "M" should be, so change to match.
151 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
152 as arguments, so an implementation trying to copy these into a local
153 buffer was taking a taint-enforcement trap. Fix by using dynamically
154 created buffers. Similar fix for radius expansion condition.
156 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
157 reasonable, eg. to count headers. Fix by using dynamically created
158 buffers rather than a local. Do similar fixes for ACL actions "dcc",
159 "log_reject_target", "malware" and "spam"; the arguments are expanded
160 so could be handling tainted values.
162 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
163 broken the (no-op) support for this sendmail command. Restore it
164 to doing nothing, silently, and returning good status.
166 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
167 record path was given (or the default used) without a leading directory
168 path, an error occurred on trying to open it. Use the transport's working
171 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
172 transport. Previously it was the name on the DNS A-record; use instead
173 the head of the CNAME chain leading there (if there is one). This seems
174 to align better with RFC 6125.
176 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
177 smtp_accept_max_per_host allocated resources which were not released
178 when the limit was exceeded. This eventually crashed the daemon. Fix
179 by adding a release action in that path.
181 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
182 expanded; previously using tainted values was rejected. Fix by using
183 dynamically-created buffers.
185 JH/09 Relax restrictions on ACL verify condition needing access to message
186 headers. Previously they were only permitted in data and non-smtp ACLs;
187 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
188 not_blind, header_sender and header_names_ascii verification.
190 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
191 Previously a macro used one argument twice; when called with the
192 argument as an expression having side-effects, incorrect operation
193 resulted. Use an inlineable function.
195 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
196 held open for a verify callout. Previously this wan not accounted for
197 and a corrupt onward SMTP conversation resulted.
199 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
200 Previously they were embedded naively in the constructed address; when
201 needed, strip the quoting and quote the entire local_part.
202 Also make the inbound_srs expansion condition handle quoting.
204 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
205 excluded, not matching the documentation.
207 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
208 was given for the sqlite_dbfile a trap resulted.
210 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
211 "name" argument resulted in a trap. There is no reason to disallow such;
212 this was a coding error.
214 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
215 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
216 spent suspended, ignoring the POSIX definition. Previously we assumed
217 it did and a constant offset from real time could be used as a correction.
218 Change to using the same clock source for the start-of-message and the
219 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
220 exists, just to get a clock slightly more aligned to reality.
222 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
223 RFC says it is optional some validators care. The missing char was not
224 intended but triggered by a line-wrap alignment. Discovery and fix by
225 Guillaume Outters, hacked on by JH.
227 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
228 name being quoted was tainted a trap would be taken. Fix by using
229 dynamically created buffers. The routine could have been called by a
230 rewrite with the "h" flag, by using the "-F" command-line option, or
231 by using a "name=" option on a control=submission ACL modifier.
233 JH/19 SPF: change the Authentication-Results expansion component to give
234 smtp.helo when the sender domain is empty. Previously it gave
237 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
238 not in 127.0.0.0/8 to help in spotting list domains taken over by a
239 domain-parking registrar.
241 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
242 Previously when a whitespace character was specified it was not inserted
243 after removing the newline.
245 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
246 the domain part of the recipient address. This overrides any tls_sni
247 option set, which was previously used.
249 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
252 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
253 is_tainted() had an off-by-one error in the overenthusiastic direction.
254 Find and fix by Gavan. Although NetBSD is not a supported platform for
255 4.94 this bug could affect other platforms.
257 PP/01 Fix default prime selection to be consistent.
258 One path used ike23 still, instead of exim.dev.20160529.3; now both
259 execution flows will use the same DH primes (currently
260 exim.dev.20160529.3).
262 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
263 option in smtp transport, to match the documentation. Previously
264 verification was not being done in some cases where it should have been.
266 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
267 than one server was defined and depending on the platform memory layout
268 details, an internal consistency trap could be hit while walking the list
271 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
272 files. The value is used by the authresults expansion item. Previously
273 if this was used in a router or transport, a crash could result.
275 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
276 have a sufficient incidence of EINTR returns from select that an
277 interaction with other operations done by the main daemon loop exposed
278 a bug in the error-handling. This was benign apart from the log
281 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
282 for consistency with delivered "=>" and failed "**" lines. While we're
283 there, handle PRX and TFO.
285 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
286 applied. This resulted, if any header-line rewrite rules were configured,
287 in a panic-log triggerable by sending a message with a long address in
288 a header. Fix by increasing the arbitrary limit to larger than a single
289 (dewrapped) 5322 header line maximum size.
291 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
292 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
293 RFC 6648 which deprecates X- options in protocols as a general practice.
294 Changeover between the implementations is handled by the mechanisms
297 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
298 a remote which uses callout/hold. Previously the local was queued.
300 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
303 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
304 log line. The asterisk is now consistently appended to the remote IP
305 (and port, if given), and will also be provided on defer and fail log
306 lines. Previously it could be placed on the local IP if that was being
307 logged, and was only provided on delivery lines.
309 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
311 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
312 authenticator client_send option. Previously the next char, after a pair
313 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
314 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
315 documentation. There is still no way to get a leading ^ immediately
316 after a NUL (ie. for the password of a PLAIN method authenticator.
318 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
319 files. For bad sizes read, delete the record and whine to paniclog.
321 JH/38 When logging an AUTH failure, as server, do not include sensitive
322 information. Previously, the credentials would be included if given
323 as part of the AUTH command line and an ACL denied authentication.
325 JH/39 Bug 2691: fix $local_part_data. When the matching list element
326 referred to a file, bad data was returned. This likely also affected
329 JH/40 The gsasl authenticator now supports caching of the salted password
330 generated by the client-side implementation. This required the addition
331 of a new variable: $auth4.
333 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
334 left undeleted; the attempt to re-create it then failed - resulting in
335 the usual "SIGHUP tp have daemon reload configuration" to not work.
336 This affected any platform not supporting "abstract" Unix-domain
337 sockets (i.e. not Linux).
339 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
340 recipients" response to RCPT in a later response, with a 250. The
341 previous coding assumed this would not happen, and under PIPELINING
342 would result in both lost and duplicate recipients for a message.
344 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
345 Previously the weighting was incorrectly applied. Similar fix for socks
346 proxies. Found and fixed by Heiko Schlichting.
348 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
349 not handle sub-lists included using the +namedlist syntax. While
350 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
351 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
354 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
355 Previously the permanent pool was used, so the sore could not be freed.
356 This meant a connection with many messages would use continually-growing
359 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
360 per-pool so as not to waste too much space. Previously a constant size
361 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
362 DOS attacks harder. The cost is wasted memory use in the larger blocks.
364 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
365 same space repeatedly, and a smaller process footprint.
367 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
368 for GnuTLS. Previously the name started "%s" which, while not a bug,
369 looked as if if might be one.
371 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
372 "continued connection") make the $proxy_* variables available. Previously
373 the information was not passed across the exec() call for subsequent
374 transport executions. This also mean that the log lines for the
375 messages can show the proxy information.
377 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
378 receive time. With modern systems the difference is significant.
379 The historical behaviour can be restored by disabling (a new) log_selector
380 "queue_time_exclusive".
382 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
383 content) were specifically tested for. Now, also cover expansions
384 resulting in ACL names and inline ACL content.
386 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
387 making it unusable in complex expressions.
389 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
390 fail with a taint-check on the spoolfile name, and leave the message
393 HS/01 Enforce absolute PID file path name.
395 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
397 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
399 PP/02 Bug 2643: Correct TLS DH constants.
400 A missing NUL termination in our code-generation tool had led to some
401 incorrect Diffie-Hellman constants in the Exim source.
402 Reported by kylon94, code-gen tool fix by Simon Arlott.
404 PP/03 Impose security length checks on various command-line options.
405 Fixes CVE-2020-SPRSS reported by Qualys.
407 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
408 better. Reported by Qualys.
410 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
411 providing a particularly obnoxious sender full name.
414 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
416 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
419 PP/08 Change default for recipients_max from unlimited to 50,000.
421 PP/09 Fix security issue with too many recipients on a message (to remove a
422 known security problem if someone does set recipients_max to unlimited,
423 or if local additions add to the recipient list).
424 Fixes CVE-2020-RCPTL reported by Qualys.
426 PP/10 Fix security issue in SMTP verb option parsing
427 Fixes CVE-2020-EXOPT reported by Qualys.
429 PP/11 Fix security issue in BDAT state confusion.
430 Ensure we reset known-good where we know we need to not be reading BDAT
431 data, as a general case fix, and move the places where we switch to BDAT
432 mode until after various protocol state checks.
433 Fixes CVE-2020-BDATA reported by Qualys.
435 HS/03 Die on "/../" in msglog file names
437 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
438 the Exim runtime user are allowed to create files.
440 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
443 QS/03 When reading the output from interpreted forward files we do not
444 pass the pipe between the parent and the interpreting process to
445 executed child processes (if any).
447 QS/04 Always die if requested from internal logging, even is logging is
450 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
451 the API; compilation noo longer completes with DMARC support included.
452 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
453 been reported on other platforms.
455 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
457 JH/56 Make the majority of info read from config files readonly, for defence-in-
458 depth against exploits. Suggestion by Qualys.
459 Not supported on Solaris 10.
461 JH/57 Fix control=fakreject for a custom message containing tainted data.
462 Previously this resulted in a log complaint, due to a re-expansion present
463 since fakereject was originally introduced.
465 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
466 modifier was given, a loop resulted.
468 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
469 pipelined SMTP command followed the BDAT LAST then it would be
470 incorrectly treated as part of the message body, causing a verification
473 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
474 require looser validation rules than those for 821-level addresses,
475 which only permit IP addresses.
481 JH/01 Avoid costly startup code when not strictly needed. This reduces time
482 for some exim process initialisations. It does mean that the logging
483 of TLS configuration problems is only done for the daemon startup.
485 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
487 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
488 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
491 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
492 DKIM signing is being done. Previously a transport_filter would always
493 disable CHUNKING, falling back to traditional DATA.
495 JH/05 Regard command-line recipients as tainted.
497 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
499 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
500 PAM library frees one of the arguments given to it, despite the
501 documentation. Therefore a plain malloc must be used.
503 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
504 on-stack buffers were used, resulting in a taint trap when DSN information
505 copied from a received message was written into the buffer.
507 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
508 the ordering of its ARC headers. This caused a crash.
510 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
511 a new record was being constructed with information from the peer, a trap
514 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
515 installation would get error messages from DMARC verify, when it hit the
516 nonexistent file indicated by the default. Distros wanting DMARC enabled
517 should both provide the file and set the option.
518 Also enforce no DMARC verification for command-line sourced messages.
520 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
521 could, depending on the platform, hang at the STARTTLS response.
523 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
524 message on a connection. Previously if one message had ARC headers and
525 the following one did not, a crash could result when adding an
526 Authentication-Results: header.
528 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
529 Exim main code and Exim-related utities. The introduction of taint
530 tracking also did many adjustments to string handling. Since then, eximon
531 frequently terminated with an assert failure.
533 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
534 check for 452 responses. This slightly helps the inefficieny of doing
535 a large alias-expansion into a recipient-limited target. The max_rcpt
536 transport option still applies (and at the current default, will override
537 the new feature). The check is done for either cause of synch, and forces
538 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
539 connection. The new facility is not tunable at this time.
541 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
542 library live data was being used, so the results became garbage. Make
543 copies while it is still usable.
545 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
546 on delivery deferred (==) and failed (**) lines (if a delivery was
547 attemtped). Previously it was only on completion (=>) lines.
549 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
550 for the expansion of the server_scram_iter and server_scram_salt options.
552 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
553 are now specifically given a NO_DATA response without hitting the system
554 resolver. The library goes on to do the now-standard TXT lookup.
555 Use of dnsdb lookups is not affected.
557 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
558 only retrieve the errormessage once. Previously two calls to dlerror()
559 were used, and the second one (for mainlog/paniclog) retrieved null
562 JH/20 Taint checking: disallow use of tainted data for
563 - the appendfile transport file and directory options
564 - the pipe transport command
565 - the autoreply transport file, log and once options
566 - file names used by the redirect router (including filter files)
568 - paths used by single-key lookups
569 Previously this was permitted.
571 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
572 adjusted the size of a major service buffer; this failed because the
573 buffer was in use at the time. Change to a compile-time increase in the
574 buffer size, when this authenticator is compiled into exim.
576 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
577 previous fast-mode was untenable in the face of glibs using mmap to
578 support larger malloc requests.
580 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
581 New values supported, if defined on system where compiled:
582 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
583 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
585 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
586 running a limited number of proceses in parallel, a benefit is gained. The
587 amount varies with the platform hardware and load. The use of the option
588 queue_run_in_order means we cannot do this, as ordering becomes
591 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
592 had introduced a string-copy (for ensuring NUL-termination) which was not
593 appropriate for that case, which can include embedded NUL bytes in the
594 block of data. Investigation showed the copy to actually be needless, the
595 data being length-specified.
597 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
598 done during a receiving connection, and both used TLS, global info was
599 used rather than per-connection info for tracking the state of data
600 queued for transmission. This could result in a connection hang.
602 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
603 Previously, when delivering serveral messages down a single connection
604 only the first would provide a SIZE. This was due to the size information
605 not being properly tracked.
607 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
608 TAI (at 37 seconds currently), pretend to be in UTC for time-related
609 expansion and logging. Previously, spurious values such as a future
610 minute could be seen.
612 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
613 it could crash from a null-deref. This could also affect the
614 ${addresses: } operator and ${readsock } item.
616 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
617 message following a mime one, the variable was not reset.
619 JH/30 When an pipelined-connect fails at the first response, assume incorrect
620 cached capability (perhaps the peer reneged?) and immediately retry in
623 JH/31 Fix spurious detection of timeout while writing to transport filter.
625 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
626 an attempt to copy the string was made before checking it.
628 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
629 taint of the lookup key was maintained; we now regard the presence in the
630 filesystem as sufficient validation.
632 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
633 argument is supplied.
635 JH/35 The dsearch lookup now requires that the directory is an absolute path.
636 Previously this was not checked, and nonempty relative paths made an
637 access under Exim's current working directory.
639 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
640 Previously no event was raised.
642 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
643 parameter supplied by the sender MAIL FROM command. Previously it was
644 ignored, and only the check_spool_space option value for the required
647 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
648 the size of the signing public-key. Previously it was instead giving
649 the size of the signature hash.
651 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
652 the default. See the (new) dkim_verify_min_keysizes option.
654 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
655 and an ACL use a lookup for checking either the local_part or domain,
656 stale data could be accessed. Ensure that variable references are
657 dropped between messages.
659 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
660 by the client was not checked as pointing within response data before
661 being used. A malicious client could thus cause an out-of-bounds read and
662 possibly gain authentication. Fix by adding the check.
664 JH/42 Internationalisation: change the default for downconversion in the smtp
665 transport to be "if needed". Previously it was "as previously set" for
666 the message, which usually meant "if needed" for message-submission but
667 "no" for everything else. However, MTAs have been seen using SMTPUTF8
668 even when the envelope addresses did not need it, resulting in forwarding
669 failures to non-supporting MTAs. A downconvert in such cases will be
670 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
671 transport. The change does mean that addresses needing conversion will
672 be converted when previously a delivery failure would occur.
674 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
675 response was received it would be used unchecked in a fail-DSN, violating
676 standards on line-length limits. Truncate if needed.
678 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
679 doesn't work. (Additionally add a new main config option to configure the
686 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
687 side, to decode a TLS 1.3 packet capture.
689 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
690 Previously the default library behaviour applied, sending two, each in
693 JH/03 Debug output for ACL now gives the config file name and line number for
696 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
698 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
700 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
701 buffer overrun for (non-chunking) other transports.
703 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
704 TLS1.3, means that a server rejecting a client certificate is not visible
705 to the client until the first read of encrypted data (typically the
706 response to EHLO). Add detection for that case and treat it as a failed
707 TLS connection attempt, so that the normal retry-in-clear can work (if
708 suitably configured).
710 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
711 and/or domain. Found and fixed by Jason Betts.
713 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
714 configuration). If a CNAME target was not a wellformed name pattern, a
717 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
718 the OS reports them interleaved with other addresses.
720 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
721 used both for input and for a verify callout, both encrypted, SMTP
722 responses being sent by the server could be lost. This resulted in
723 dropped connections and sometimes bounces generated by a peer sending
726 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
727 string. Previously it was possible to cause undefined behaviour in a
728 library routine (usually a crash). Found by "zerons".
730 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
733 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
734 API was removed, so update to use the newer ones.
736 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
737 any timeout set, is taking a long time. Previously we would hang on to a
738 rotated logfile "forever" if the input was arriving with long gaps
739 (a previous attempt to fix addressed lack, for a long time, of initial
742 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
743 shared (NFS) environment. The length of the tempfile name is now
744 4 + 16 ("hdr.$message_exim_id") which might break on file
745 systems which restrict the file name length to lower values.
748 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
749 shared (NFS) environment.
751 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
752 did for all versions <4.90). Notably -M, -m, --invert, -I may be
755 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
756 on some platforms for bit 31.
758 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
759 to changes apparently associated with TLS1.3 handling some of the APIs
760 previously used were either nonfunctional or inappropriate. Strings
761 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
762 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
763 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
764 This affects log line X= elements, the $tls_{in,out}_cipher variables,
765 and the use of specific cipher names in the encrypted= ACL condition.
767 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
769 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
770 verification result was not updated unless hosts_require_ocsp applied.
772 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
773 queue_list_requires_admin set to false, non-admin users were denied the
776 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
777 directory-of-certs mode. Previously they were advertised despite the
780 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
781 A single TCP connection by a client will now hold a TLS connection open
782 for multiple message deliveries, by default. Previously the default was to
785 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
786 default. If built with the facility, DANE will be used. The facility
787 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
789 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
790 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
791 must be defined and you must still, unless you define DISABLE_TLS, manage
792 the the include-dir and library-file requirements that go with that
793 choice. Non-TLS builds are still supported.
795 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
798 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
799 default. If the platform supports and has the facility enabled, it will
800 be requested on all coneections.
802 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
803 controlled by the build-time option SUPPORT_PIPE_CONNECT.
805 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
807 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
808 success-DSN messages. Previously the From: header was always the default
809 one for these; the option was ignored.
811 JH/28 Fix the timeout on smtp response to apply to the whole response.
812 Previously it was reset for every read, so a teergrubing peer sending
813 single bytes within the time limit could extend the connection for a
814 long time. Credit to Qualsys Security Advisory Team for the discovery.
816 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
817 delivery address, which leaked information of the results of local
818 forwarding. Change to the original envelope recipient address, per
821 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
822 requested. Previously not bounce was generated and a log entry of
823 error ignored was made.
825 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
827 JH/32 Introduce a general tainting mechanism for values read from the input
828 channel, and values derived from them. Refuse to expand any tainted
829 values, to catch one form of exploit.
831 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
832 was unused and the unexpanded text used for the test. Found and
833 fixed by Ruben Jenster.
835 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
836 an attempt to use a TLS library read routine dereffed a nul pointer,
839 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
840 them in our smtp response.
842 JH/36 Have the general router option retry_use_local_part default to true when
843 any of the restrictive preconditions are set (to anything). Previously it
844 was only for check_local user. The change removes one item of manual
845 configuration which is required for proper retries when a remote router
846 handles a subset of addresses for a domain.
848 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
849 link count into consideration.
851 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
852 caused the extension of big_buffer, the following lines were ignored.
854 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
855 accordance with RFC 2308. Previously there was no expiry, so a longlived
856 receive process (eg. due to ACL delays) versus a short SOA value could
859 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
861 JH/39 Promote DMARC support to mainline.
863 JH/40 Bug 2452: Add a References: header to DSNs.
865 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
866 parameters. The relevant library call is documented as "Deprecated: This
867 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
868 3.6.0, DH parameters are negotiated following RFC7919."
870 HS/06 Change the default of dnssec_request_domains to "*"
872 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
873 carried on and emitted a BDAT command, even when PIPELINING was not
876 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
877 buffer was used for the filename, resulting in a trap when tainted
878 arguments (eg. $domain) were used.
880 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
881 recommended to avoid a possible server-load attack. The feature can be
882 re-enabled via the openssl_options main cofiguration option.
884 JH/45 local_scan API: documented the current smtp_printf() call. This changed
885 for version 4.90 - adding a "more data" boolean to the arguments.
886 Bumped the ABI version number also, this having been missed previously;
887 release versions 4.90 to 4.92.3 inclusive were effectively broken in
888 respect of usage of smtp_printf() by either local_scan code or libraries
889 accessed via the ${dlfunc } expansion item. Both will need coding
890 adjustment for any calls to smtp_printf() to match the new function
891 signature; a FALSE value for the new argument is always safe.
893 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
894 the file-offset (which the Linux syscall does, and exim expects); this
895 resulted in an indefinite loop.
897 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
898 to do ARC verification. The Authentication-Results: header line added
899 by the configuration then had no ARC item.
901 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
902 an incorrect mode was used for reading the result, resulting in it being
909 JH/01 Remove code calling the customisable local_scan function, unless a new
910 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
912 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
913 non-signal-safe functions being used.
915 JH/03 Bug 2269: When presented with a received message having a stupidly large
916 number of DKIM-Signature headers, disable DKIM verification to avoid
917 a resource-consumption attack. The limit is set at twenty.
919 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
920 report of oldest_pass in ${authres } in consequence, and separate out
921 some descriptions of reasons for verification fail.
923 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
924 files in the spool were present and unlocked. A queue-runner could spot
925 them, resulting in a duplicate delivery. Fix that by doing the unlock
926 after the unlink. Investigation by Tim Stewart. Take the opportunity to
927 add more error-checking on spoolfile handling while that code is being
930 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
931 No known attacks, no CVE, this is defensive hardening.
933 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
934 a queue-runner could start a delivery while other operations were ongoing.
935 Cutthrough delivery was a common victim, resulting in duplicate delivery.
936 Found and investigated by Tim Stewart. Fix by using the open message data
937 file handle rather than opening another, and not locally closing it (which
938 releases a lock) for that case, while creating the temporary .eml format
939 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
941 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
942 $sender_verify_failure/$recipient_verify_failure to "random".
944 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
947 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
948 Previously this would segfault.
950 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
953 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
954 like zero, since the resolver should be doing this for us, But we need one
955 as a CNAME but no MX presence gets the CNAME returned; we need to check
956 that doesn't point to an MX to declare it "no MX returned" rather than
957 "error, loop". A new main option is added so the older capability of
958 following some limited number of chain links is maintained.
960 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
962 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
963 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
964 now-deprecated earlier definitions used only specified the range up to TLS
965 1.2 (in the older-version library docs).
967 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
969 JH/15 Rework TLS client-side context management. Stop using a global, and
970 explicitly pass a context around. This enables future use of TLS for
971 connections to service-daemons (eg. malware scanning) while a client smtp
972 connection is using TLS; with cutthrough connections this is quite likely.
974 JH/16 Fix ARC verification to do AS checks in reverse order.
976 JH/17 Support a "tls" option on the ${readsocket } expansion item.
978 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
979 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
980 Previously the "utf8" would be re-prepended for every additional message.
982 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
983 Previously thery were accepted, resulting in issues when attempting to
984 forward messages to a non-supporting MTA.
986 PP/02 Let -n work with printing macros too, not just options.
988 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
989 one parent address was copied, and bogus data was used at delivery-logging
990 time. Either a crash (after delivery) or bogus log data could result.
991 Discovery and analysis by Tim Stewart.
993 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
994 Previously if the string ended mid-character, we did not insert the
995 promised '?' replacement.
997 PP/04 Documentation: current string operators work on bytes, not codepoints.
999 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1000 should pack well giving a smaller memory footprint so better caching and
1001 therefore performance. Group the declarations where this can't be done so
1002 that the byte-sized flag variables are not interspersed among pointer
1003 variables, giving a better chance of good packing by the compiler.
1005 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1006 non-null, to avoid issues with sites running BATV. Previously reports were
1007 sent with an empty envelope sender so looked like bounces.
1009 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1010 The ignore_error flag wasn't being returned from the filter subprocess so
1011 was not set for later routers. Investigation and fix by Matthias Kurz.
1013 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1014 and a msg:complete for the whole, when a message is manually removed using
1015 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1017 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1018 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1019 hope that is portable enough.
1021 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1022 requiring a known-CA anchor certificate; make it now rely entirely on the
1023 TLSA as an anchor. Checking the name on the leaf cert against the name
1024 on the A-record for the host is still done for TA (but not for EE mode).
1026 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1027 included in delivery lines for non-proxied connections, when compiled with
1028 SUPPORT_SOCKS and running with proxy logging enabled.
1030 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1031 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1032 move the existing event to fire before the normal logging of message
1033 failure so that custom logging is bracketed by normal logging.
1035 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1036 msg:fail:internal event. Developement by Matthias Kurz.
1038 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1039 far too small for todays use of crypto signatures stored there. Go all
1040 the way to the max DNS message size of 64kB, even though this might be
1041 overmuch for IOT constrained device use.
1043 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1044 copy a string over itself. The library routine is documented as not
1045 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1047 JH/32 For main options check_spool_space and check_inode_space, where the
1048 platform supports 64b integers, support more than the previous 2^31 kB
1049 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1050 the previous G, M, k.
1052 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1053 $authenticated_fail_id variable on authentication failure. Previously
1056 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1057 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1058 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1059 GNUTLS_SEC_PARAM_MEDIUM.
1061 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1062 side. Previously we would continue as if no SNI had been received.
1064 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1065 "<" character, which should be a list-separator specification, we walked
1066 off past the nul-terimation.
1068 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1069 causes) even when the retry time is not yet met. Previously they were
1070 not, meaning that when (say) an account was over-quota and temp-rejecting,
1071 and multiple senders' messages were queued, only one sender would get
1072 notified on each configured delay_warning cycle.
1074 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1076 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1077 cipher-suites, an error can be left on the stack even for a succeeding
1078 accept; this results in impossible error messages when a later operation
1081 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1082 return error codes indicating retry. Under TLS1.3 this becomes required.
1084 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1085 it only wrote the new authenticators, resulting in a lack of tracking of
1086 peer changes of ESMTP extensions until the next cache flush.
1088 JH/41 Fix the loop reading a message header line to check for integer overflow,
1089 and more-often against header_maxsize. Previously a crafted message could
1090 induce a crash of the recive process; now the message is cleanly rejected.
1092 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1093 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1099 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1100 When redis_servers is set to a list of > 1 element, and the Redis servers
1101 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1102 case of MOVED into a DEFER case instead, thus moving the query onto the
1103 next server in the list. For a cluster of N elements, all N servers must
1104 be defined in redis_servers.
1106 GF/02 Catch and remove uninitialized value warning in exiqsumm
1107 Check for existence of @ARGV before looking at $ARGV[0]
1109 JH/01 Replace the store_release() internal interface with store_newblock(),
1110 which internalises the check required to safely use the old one, plus
1111 the allocate and data copy operations duplicated in both (!) of the
1112 extant use locations.
1114 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1115 modifier. This matches the restriction on the commandline.
1117 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1118 Previously only the last row was returned.
1120 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1121 we assumed that tags in the header were well-formed, and parsed the
1122 element content after inspecting only the first char of the tag.
1123 Assumptions at that stage could crash the receive process on malformed
1126 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1127 While running the DKIM ACL we operate on the Permanent memory pool so that
1128 variables created with "set" persist to the DATA ACL. Also (at any time)
1129 DNS lookups that fail create cache records using the Permanent pool. But
1130 expansions release any allocations made on the current pool - so a dnsdb
1131 lookup expansion done in the DKIM ACL releases the memory used for the
1132 DNS negative-cache, and bad things result. Solution is to switch to the
1133 Main pool for expansions.
1134 While we're in that code, add checks on the DNS cache during store_reset,
1135 active in the testsuite.
1136 Problem spotted, and debugging aided, by Wolfgang Breyha.
1138 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1139 When none of the hosts presented to a transport match an already-open
1140 connection, close it and proceed with the list. Previously we would
1141 queue the message. Spotted by Lena with Yahoo, probably involving
1144 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1145 Previously a spurious "250 OK id=" response was appended to the proper
1148 JH/08 The "support for" informational output now, which built with Content
1149 Scanning support, has a line for the malware scanner interfaces compiled
1150 in. Interface can be individually included or not at build time.
1152 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1153 by the template makefile "src/EDITME". The "STREAM" support for an older
1154 ClamAV interface method is removed.
1156 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1157 rows affected is given instead).
1159 JH/11 The runtime Berkeley DB library version is now additionally output by
1160 "exim -d -bV". Previously only the compile-time version was shown.
1162 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1163 SMTP connection. Previously, when one had more recipients than the
1164 first, an abortive onward connection was made. Move to full support for
1165 multiple onward connections in sequence, handling cutthrough connection
1166 for all multi-message initiating connections.
1168 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1169 routers. Previously, a multi-recipient message would fail to match the
1170 onward-connection opened for the first recipient, and cause its closure.
1172 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1173 a timeout on read on a GnuTLS initiating connection, resulting in the
1174 initiating connection being dropped. This mattered most when the callout
1175 was marked defer_ok. Fix to keep the two timeout-detection methods
1178 JH/15 Relax results from ACL control request to enable cutthrough, in
1179 unsupported situations, from error to silently (except under debug)
1180 ignoring. This covers use with PRDR, frozen messages, queue-only and
1183 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1185 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1186 metadata, resulting in a crash in free().
1188 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1189 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1190 Broken also in d185889f4, with init system revamp.
1192 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1193 Previously we abruptly closed the connection after reading a malware-
1194 found indication; now we go on to read the "scan ok" response line,
1197 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1198 ACL. Previously, a crash would result.
1200 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1201 macro text after a replacement (previously it was only once per line) and
1202 by skipping builtin macros when searching for an uppercase lead character.
1204 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1205 for the build is renamed.
1207 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1208 was allocated for every new TLS startup, meaning one per message. Fix
1209 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1211 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1212 reported the original. Fix to report (as far as possible) the ACL
1213 result replacing the original.
1215 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1216 OpenSSL. Certificate information is loaded for every new TLS startup,
1217 and the resources needed to be freed.
1219 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1221 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1224 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1225 DATA response info to the (existing) per-recipient response info for
1226 the "C=" log element. It can have useful tracking info from the
1227 destination system. Patch from Simon Arlott.
1229 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1230 length value. Previously this would segfault.
1232 HS/02 Support Avast multiline protoocol, this allows passing flags to
1233 newer versions of the scanner.
1235 JH/28 Ensure that variables possibly set during message acceptance are marked
1236 dead before release of memory in the daemon loop. This stops complaints
1237 about them when the debug_store option is enabled. Discovered specifically
1238 for sender_rate_period, but applies to a whole set of variables.
1239 Do the same for the queue-runner and queue-list loops, for variables set
1240 from spool message files. Do the same for the SMTP per-message loop, for
1241 certain variables indirectly set in ACL operations.
1243 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1244 as a multi-recipient message from a mailinglist manager). The coding had
1245 an arbitrary cutoff number of characters while checking for more input;
1246 enforced by writing a NUL into the buffer. This corrupted long / fast
1247 input. The problem was exposed more widely when more pipelineing of SMTP
1248 responses was introduced, and one Exim system was feeding another.
1249 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1250 receiving system, and refused recipients seen by the sending system
1251 (propating to people being dropped from mailing lists).
1252 Discovered and pinpointed by David Carter.
1254 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1255 replaced by the ${authresults } expansion.
1257 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1259 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1260 allows proper process termination in container environments.
1262 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1263 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1265 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1266 and "err_perm", deprecated since 4.83 when the RFC-defined words
1267 "temperror" and "permerror" were introduced.
1269 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1270 transport-filters or DKIM-signing. The restriction was lost in the
1271 consolidation of verify-callout and delivery SMTP handling.
1272 Extend the restriction to also cover ARC-signing.
1274 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1275 in defer=pass mode supply a 450 to the initiator. Previously the message
1278 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1279 tls_require_ciphers is used as before.
1281 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1282 "pass_unscanned". Only tmpfails from the scanner are written to
1283 the paniclog, as they may require admin intervention (permission
1284 denied, license issues). Other scanner errors (like decompression
1285 bombs) do not cause a paniclog entry.
1287 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1288 Previously it was possible to log spurious information in receive log
1291 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1292 triggered odd behaviour from Outlook Express clients.
1294 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1297 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1298 since the IETF WG has not yet settled on that versus the original
1299 "bare" representation.
1301 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1302 Previously the millisecond value corrupted the output.
1303 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1304 corrupted the output.
1310 JH/01 Rework error string handling in TLS interface so that the caller in
1311 more cases is responsible for logging. This permits library-sourced
1312 string to be attached to addresses during delivery, and collapses
1313 pairs of long lines into single ones.
1315 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1316 during configuration. Wildcards are allowed and expanded.
1318 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1319 This permits better logging.
1321 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1322 do not close it down and have the child transport start it up again on
1323 the passed-on TCP connection. Instead, proxy the child (and any
1324 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1325 affected: the continued delivery log lines do not have any DNSSEC, TLS
1326 Certificate or OCSP information. TLS cipher information is still logged.
1328 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1329 identical IP addresses on different listening ports. Will also affect
1332 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1333 add noisy ifdef guards to special-case this sillyness.
1334 Patch from Bernd Kuhls.
1336 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1337 than 255 are no longer allowed.
1339 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1340 Disable the session-cache too, which might reduce our load. Since we
1341 currrectly use a new context for every connection, both as server and
1342 client, there is no benefit for these.
1343 GnuTLS appears to not support tickets server-side by default (we don't
1344 call gnutls_session_ticket_enable_server()) but client side is enabled
1345 by default on recent versions (3.1.3 +) unless the PFS priority string
1348 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1349 <https://reproducible-builds.org/specs/source-date-epoch/>.
1351 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1352 the check for any unsuccessful recipients did not notice the limit, and
1353 erroneously found still-pending ones.
1355 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1356 MSG_MORE. Only in-clear (not on TLS connections).
1358 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1359 transport-filter is involved we can buffer the headers in memory for
1360 creating the signature, and read the spool data file once for the
1361 signature and again for transmission.
1363 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1364 4.77 as the kernel support then wasn't solid, having issues in 64bit
1365 mode. Now, it's been long enough. Add support for FreeBSD also.
1367 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1368 case where the routing stage had gathered several addresses to send to
1369 a host before calling the transport for the first, we previously failed
1370 to close down TLS in the old transport process before passing the TCP
1371 connection to the new process. The new one sent a STARTTLS command
1372 which naturally failed, giving a failed delivery and bloating the retry
1373 database. Investigation and fix prototype from Wolfgang Breyha.
1375 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1376 false-negatives in the check that the sender had not preempted a response
1377 or prompt from Exim (running as a server), due to that code's lack of
1378 awareness of the SMTP input buffering.
1380 PP/04 Add commandline_checks_require_admin option.
1381 Exim drops privileges sanely, various checks such as -be aren't a
1382 security problem, as long as you trust local users with access to their
1383 own account. When invoked by services which pass untrusted data to
1384 Exim, this might be an issue. Set this option in main configuration
1385 AND make fixes to the calling application, such as using `--` to stop
1388 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1389 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1390 responses to those, into a single TLS record each way (this usually means
1391 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1394 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1395 affects you only if you're dancing at the edge of the param size limits.
1396 If you are, and this message makes sense to you, then: raise the
1397 configured limit or use OpenSSL 1.1. Nothing we can do for older
1400 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1401 cmdline element to get the documented default one. Previously it was
1404 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1405 are unsuable for non-message-specific reasons - eg. connection timeout,
1406 banner-time rejection.
1408 JH/16 Fix logging of delivery remote port, when specified by router, under
1411 PP/06 Repair manualroute's ability to take options in any order, even if one
1412 is the name of a transport.
1415 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1417 JH/17 Change the list-building routines interface to use the expanding-string
1418 triplet model, for better allocation and copying behaviour.
1420 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1421 Previously it was constructed the first time a possibly-matching string
1422 was met in the configuration file input during startup; now it is done
1425 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1426 compatible one, to avoid the (poorly documented) possibility of a config
1427 file in the working directory redirecting the DB files, possibly correpting
1428 some existing file. CVE-2017-10140 assigned for BDB.
1430 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1431 cache-hot. Previously, although the result was properly cached, the
1432 initial verify call returned a defer.
1434 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1435 the main verify for receipient in uncached-mode.
1437 JH/22 Retire historical build files to an "unsupported" subdir. These are
1438 defined as "ones for which we have no current evidence of testing".
1440 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1441 if present. Previously it was ignored.
1443 JH/24 Start using specified-initialisers in C structure init coding. This is
1444 a C99 feature (it's 2017, so now considered safe).
1446 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1447 if was a fixed-sized field and bitmask ops via macros; it is now more
1450 PP/07 GitHub PR 56: Apply MariaDB build fix.
1451 Patch provided by Jaroslav Škarvada.
1453 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1454 during Coverity cleanups [4.87 JH/47]
1455 Diagnosis and fix provided by Michael Fischer v. Mollard.
1457 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1458 the right size to place the terminating semicolon on its own folded
1459 line, the header hash was calculated to an incorrect value thanks to
1460 the (relaxed) space the fold became.
1462 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1463 and confused the parent.
1465 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1466 which could crash as a result. This could lead to undeliverable messages.
1468 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1471 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1472 matters on fast-turnover and PID-randomising systems, which were getting
1473 out-of-order delivery.
1475 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1476 a possibly-overlapping copy. The symptom was that "Remote host closed
1477 connection in response to HELO" was logged instead of the actual 4xx
1480 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1481 Previously only that bufferd was discarded, resulting in SYMTP command
1484 JH/32 DKIM: when a message has multiple signatures matching an identity given
1485 in dkim_verify_signers, run the dkim acl once for each. Previously only
1486 one run was done. Bug 2189.
1488 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1489 "panic the current process" to "deliberately defer". The panic log is
1490 still written with the problem list name; the mail and reject logs now
1491 get a temp-reject line for the message that was being handled, saying
1492 something like "domains check lookup or other defer". The SMTP 451
1493 message is still "Temporary local problem".
1495 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1496 A crafted sequence of BDAT commands could result in in-use memory beeing
1497 freed. CVE-2017-16943.
1499 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1500 from SMTP input. Previously it was always done; now only done for DATA
1501 and not BDAT commands. CVE-2017-16944.
1503 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1504 to the message (such as an overlong header line). Previously this was
1505 not done and we did not exit BDAT mode. Followon from the previous item
1506 though a different problem.
1512 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1513 than -2003 did; needs libidn2 in addition to libidn.
1515 JH/02 The path option on a pipe transport is now expanded before use.
1517 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1518 Patch provided by "Björn", documentation fix added too.
1520 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1521 missing a wire-to-host endian conversion.
1523 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1524 close after a BDAT command line could be taken as a following command,
1525 giving a synch failure. Fix by only checking for synch immediately
1526 before acknowledging the chunk.
1528 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1529 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1530 macro. Patches provided by Josh Soref.
1532 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1533 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1534 by IANA as a service extension. However, John Klensin suggests that we
1537 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1538 the dkim code may be unix-mode line endings rather than smtp wire-format
1539 CRLF, so prepend a CR to any bare LF.
1541 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1542 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1544 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1545 the first were themselves being wrongly included in the feed into dkim
1546 processing; with most chunk sizes in use this resulted in an incorrect
1547 body hash calculated value.
1549 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1550 DKIM signature block, for verification. Although advised against by
1551 standards it is specifically not ruled illegal.
1553 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1555 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1556 missing a body hash (the bh= tag).
1558 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1559 It seems that HAProxy sends the Proxy Protocol information in clear and
1560 only then does a TLS startup, so do the same.
1562 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1563 TCP connections (such as for Spamd) unless the daemon successfully set
1564 Fast Open mode on its listening sockets. This fixes breakage seen on
1565 too-old kernels or those not configured for Fast Open, at the cost of
1566 requiring both directions being enabled for TFO, and TFO never being used
1567 by non-daemon-related Exim processes.
1569 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1570 endings, at least on the first header line. Try to canonify any that get
1571 past that check, despite the cost.
1573 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1574 now limited to an arbitrary five deep, while parsing addresses with the
1575 strip_excess_angle_brackets option enabled.
1577 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1578 instead leave the unprompted TLS handshake in socket buffer for the
1579 TLS library to consume.
1581 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1583 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1585 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1586 region in one operation, for speed, and this leaves assigned pointers
1587 dangling. Add checks run only under the testsuite which checks all
1588 variables at a store-reset and panics on a dangling pointer; add code
1589 explicitly nulling out all the variables discovered. Fixes one known
1590 bug: a transport crash, where a dangling pointer for $sending_ip_address
1591 originally assigned in a verify callout, is re-used.
1593 PP/06 Drop '.' from @INC in various Perl scripts.
1595 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1597 PP/08 Reduce a number of compilation warnings under clang; building with
1598 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1599 should be warning-free.
1601 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1603 HS/01 Fix portability problems introduced by PP/08 for platforms where
1604 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1606 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1607 chunk. This allows us to accept broken chunked messages. We need a more
1608 general solution here.
1610 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1611 already-broken messages in the queue.
1613 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1615 JH/19 Fix reference counting bug in routing-generated-address tracking.
1621 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1622 supports it and a size is available (ie. the sending peer gave us one).
1624 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1625 years of being deprecated). The replacements are the ACLs
1626 acl_smtp_mime and acl_not_smtp_mime.
1628 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1629 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1630 or even an in-clear connection were permitted. Now, if the host lookup
1631 was dnssec and dane was requested then the host is only used if the
1632 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1633 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1634 if one fails this test.
1635 This means that a poorly-configured remote DNS will make it incommunicado;
1636 but it protects against a DNS-interception attack on it.
1638 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1639 non-noisy when a race steals the message being considered.
1641 JH/05 If main configuration option tls_certificate is unset, generate a
1642 self-signed certificate for inbound TLS connections.
1644 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1645 in rewrites and routers.
1647 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1648 and logged a warning sing 4.83; now they are a configuration file error.
1650 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1651 (lacking @domain). Apply the same qualification processing as RCPT.
1653 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1655 JH/10 Support ${sha256:} applied to a string (as well as the previous
1658 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1659 a cutthrough deliver is pending, as we always want to make a connection.
1660 This also avoids re-routing the message when later placing the cutthrough
1661 connection after a verify cache hit.
1662 Do not update it with the verify result either.
1664 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1665 when routing results in more than one destination address.
1667 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1668 signing (which inhibits the cutthrough capability). Previously only
1669 the presence of an option was tested; now an expansion evaluating as
1670 empty is permissible (obviously it should depend only on data available
1671 when the cutthrough connection is made).
1673 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1674 the relevant preceding SMTP command did not note the pipelining mode.
1676 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1677 Previously they were not counted.
1679 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1680 as one having no matching records. Previously we deferred the message
1681 that needed the lookup.
1683 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1684 distinguished as "(=".
1686 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1687 for missing MX records. Previously it only worked for missing A records.
1689 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1691 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1692 after the data-go-ahead and data-ack. Patch from Jason Betts.
1694 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1695 even for a "none" policy. Patch from Tony Meyer.
1697 JH/22 Fix continued use of a connection for further deliveries. If a port was
1698 specified by a router, it must also match for the delivery to be
1701 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1702 When one of the recipients of a message was unsuitable for the connection
1703 (has no matching addresses), we lost track of needing to mark it
1704 deferred. As a result mail would be lost.
1706 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1708 JH/25 Decoding ACL controls is now done using a binary search; the source code
1709 takes up less space and should be simpler to maintain. Merge the ACL
1710 condition decode tables also, with similar effect.
1712 JH/26 Fix problem with one_time used on a redirect router which returned the
1713 parent address unchanged. A retry would see the parent address marked as
1714 delivered, so not attempt the (identical) child. As a result mail would
1717 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1718 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1719 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1722 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1723 Main config options check_{log,spool}_{inodes,space} are now
1724 100 inodes, 10MB unless set otherwise in the configuration.
1726 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1727 Previously it only applied to the main-section connection policy
1730 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1732 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1733 by me. Added RFC7919 DH primes as an alternative.
1735 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1736 are not in the system include path.
1738 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1739 GnuTLS, when a session startup failed (eg because the client disconnected)
1740 Exim did stdio operations after fclose. This was exposed by a recent
1741 change which nulled out the file handle after the fclose.
1743 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1744 signed directly by the cert-signing cert, rather than an intermediate
1745 OCSP-signing cert. This is the model used by LetsEncrypt.
1747 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1749 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1750 an incoming connection.
1752 HS/02 Bug 1802: Do not half-close the connection after sending a request
1755 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1756 fallback to "prime256v1".
1758 JH/34 SECURITY: Use proper copy of DATA command in error message.
1759 Could leak key material. Remotely exploitable. CVE-2016-9963.
1765 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1766 and 3.4.4 - once the server is enabled to respond to an OCSP request
1767 it does even when not requested, resulting in a stapling non-aware
1768 client dropping the TLS connection.
1770 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1771 support variable-length bit vectors. No functional change.
1773 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1774 The I= interface field on outgoing lines is now after the H= remote
1775 host field, same as incoming lines. There is a separate
1776 outgoing_interface log selector which allows you to disable the
1779 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1780 If not running log_selector +smtp_connection the mainlog would be held
1781 open indefinitely after a "too many connections" event, including to a
1782 deleted file after a log rotate. Leave the per net connection logging
1783 leaving it open for efficiency as that will be quickly detected by the
1784 check on the next write.
1786 HS/01 Bug 1671: Fix post transport crash.
1787 Processing the wait-<transport> messages could crash the delivery
1788 process if the message IDs didn't exist for some reason. When
1789 using 'split_spool_directory=yes' the construction of the spool
1790 file name failed already, exposing the same netto behaviour.
1792 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1793 mime_regex ACL conditions.
1795 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1796 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1797 to HELO, local diagnostic string.
1799 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1800 log line to debug. Even when configured with a tls authenticator many
1801 client connections are expected to not authenticate in this way, so
1802 an authenticate fail is not an error.
1804 HS/02 Add the Exim version string to the process info. This way exiwhat
1805 gives some more detail about the running daemon.
1807 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1808 matter for fast-change records such as DNSBLs.
1810 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1811 retry record, even if constant. There may be multiple transports with
1812 different interface settings and the retry behaviour needs to be kept
1815 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1817 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1819 JH/10 Bug 840: fix log_defer_output option of pipe transport
1821 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1822 hosts_randomize. This matters a lot when combined with mua_wrapper.
1824 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1825 ${quote_pgsql:<string>} operator.
1827 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1829 JH/14 Change header file name for freeradius-client. Relevant if compiling
1830 with Radius support; from the Gentoo tree and checked under Fedora.
1832 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1834 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1835 option transport_filter, meaning no filtering.
1837 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1839 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1840 defaults to "*" (all hosts). The variable is now available when not built
1841 with TLS, default unset, mainly to enable keeping the testsuite sane.
1842 If a server certificate is not supplied (via tls_certificate) an error is
1843 logged, and clients will find TLS connections fail on startup. Presumably
1844 they will retry in-clear.
1845 Packagers of Exim are strongly encouraged to create a server certificate
1846 at installation time.
1848 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1849 with the $config_file variable.
1851 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1852 in transport context, after the attempt, and per-recipient. The latter type
1853 is per host attempted. The event data is the error message, and the errno
1854 information encodes the lookup type (A vs. MX) used for the (first) host,
1855 and the trailing two digits of the smtp 4xx response.
1857 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1858 to write to mainlog (or rejectlog, paniclog) in the window between file
1859 creation and permissions/ownership being changed. Particularly affects
1860 installations where exicyclog is run as root, rather than exim user;
1861 result is that the running daemon panics and dies.
1863 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1865 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1866 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1867 "pri" and "weight". Note that the previous implicit priority given by the
1868 list order is no longer honoured.
1870 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1871 for DKIM processing.
1873 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1874 by defining SUPPORT_SOCKS.
1876 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1877 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1878 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1879 variables are renamed to proxy_{local,external}_{address,port}.
1881 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1882 for a build by defining SUPPORT_I18N
1884 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1885 of the query string, and make ${quote_redis:} do that quoting.
1887 JH/29 Move Events support from Experimental to mainline, enabled by default
1888 and removable for a build by defining DISABLE_EVENT.
1890 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1892 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1893 cached by the daemon.
1895 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1896 by defining LOOKUP_REDIS. The libhiredis library is required.
1898 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1899 keys are given for lookup.
1901 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1902 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1903 only supported when built with TLS support. The PolarSSL SHA routines
1904 are still used when the TLS library is too old for convenient support.
1906 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1907 openssl_options), for security. OpenSSL forces this from version 1.1.0
1908 server-side so match that on older versions.
1910 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1911 allocation for $value could be released as the expansion processing
1912 concluded, but leaving the global pointer active for it.
1914 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1915 and to use the domains and local_parts ACL conditions.
1917 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1918 incorrectly not doubled on cutthrough transmission, hence seen as a
1919 body-termination at the receiving system - resulting in truncated mails.
1920 Commonly the sender saw a TCP-level error, and retransmitted the message
1921 via the normal store-and-forward channel. This could result in duplicates
1922 received - but deduplicating mailstores were liable to retain only the
1923 initial truncated version.
1925 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1927 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1929 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1930 we're in there, support oversigning also; bug 1309.
1932 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1934 HS/04 Add support for keep_environment and add_environment options.
1936 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1937 either intentional arithmetic overflow during PRNG, or testing config-
1940 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1941 delivery resulted in actual delivery. Cancel cutthrough before DATA
1944 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1945 rejects a recipient: pass the reject to the originator.
1947 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1948 Many were false-positives and ignorable, but it's worth fixing the
1951 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1952 for the new environment-manipulation done at startup. Move the routines
1953 from being local to tls.c to being global via the os.c file.
1955 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1956 an extract embedded as result-arg for a map, the first arg for extract
1957 is unavailable so we cannot tell if this is a numbered or keyed
1958 extraction. Accept either.
1964 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1967 JH/02 The smtp transport option "multi_domain" is now expanded.
1969 JH/03 The smtp transport now requests PRDR by default, if the server offers
1972 JH/04 Certificate name checking on server certificates, when exim is a client,
1973 is now done by default. The transport option tls_verify_cert_hostnames
1974 can be used to disable this per-host. The build option
1975 EXPERIMENTAL_CERTNAMES is withdrawn.
1977 JH/05 The value of the tls_verify_certificates smtp transport and main options
1978 default to the word "system" to access the system default CA bundle.
1979 For GnuTLS, only version 3.0.20 or later.
1981 JH/06 Verification of the server certificate for a TLS connection is now tried
1982 (but not required) by default. The verification status is now logged by
1983 default, for both outbound TLS and client-certificate supplying inbound
1986 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1989 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1990 Status Notification (bounce) messages are now MIME format per RFC 3464.
1991 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1992 under the control of the dsn_advertise_hosts option, and routers may
1993 have a dsn_lasthop option.
1995 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1996 default, modifiable by a malware= option. The list separator for
1997 the options can now be changed in the usual way. Bug 68.
1999 JH/10 The smtp_receive_timeout main option is now expanded before use.
2001 JH/11 The incoming_interface log option now also enables logging of the
2002 local interface on delivery outgoing connections.
2004 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2005 if the interface and destination host and port all match.
2007 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2010 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2011 Patch from Andrew Lewis.
2013 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2014 now supports optional time-restrictions, weighting, and priority
2015 modifiers per server. Patch originally by <rommer@active.by>.
2017 JH/16 The spamd_address main option now supports a mixed list of local
2018 and remote servers. Remote servers can be IPv6 addresses, and
2019 specify a port-range.
2021 JH/17 Bug 68: The spamd_address main option now supports an optional
2022 timeout value per server.
2024 JH/18 Bug 1581: Router and transport options headers_add/remove can
2025 now have the list separator specified.
2027 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2030 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2033 JH/21 Support for the A6 type of dns record is withdrawn.
2035 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2036 rather than the verbs used.
2038 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2039 from 255 to 1024 chars.
2041 JH/24 Verification callouts now attempt to use TLS by default.
2043 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2044 are generic router options now. The defaults didn't change.
2046 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2047 Original patch from Alexander Shikoff, worked over by JH.
2049 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2050 files if ZCAT_COMMAND is not executable.
2052 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2054 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2056 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2057 Normally benign, it bites when the pair was led to by a CNAME;
2058 modern usage is to not canonicalize the domain to a CNAME target
2059 (and we were inconsistent anyway for A-only vs AAAA+A).
2061 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2063 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2064 when evaluating $sender_host_dnssec.
2066 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2067 $sender_helo_dnssec variable.
2069 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2071 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2073 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2075 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2076 documented as working, but never had. Support all but $spam_report.
2078 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2079 added for tls authenticator.
2081 HS/03 Add perl_taintmode main config option
2087 TL/01 When running the test suite, the README says that variables such as
2088 no_msglog_check are global and can be placed anywhere in a specific
2089 test's script, however it was observed that placement needed to be near
2090 the beginning for it to behave that way. Changed the runtest perl
2091 script to read through the entire script once to detect and set these
2092 variables, reset to the beginning of the script, and then run through
2093 the script parsing/test process like normal.
2095 TL/02 The BSD's have an arc4random API. One of the functions to induce
2096 adding randomness was arc4random_stir(), but it has been removed in
2097 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2098 function when detected.
2100 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2101 cause callback expansion.
2103 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2104 syntax errors in an expansion can be treated as a string instead of
2105 logging or causing an error, due to the internal use of bool_lax
2106 instead of bool when processing it.
2108 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2109 server certificates when making smtp deliveries.
2111 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2113 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2115 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2117 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2118 Merged patch from Sebastian Wiedenroth.
2120 JH/05 Fix results-pipe from transport process. Several recipients, combined
2121 with certificate use, exposed issues where response data items split
2122 over buffer boundaries were not parsed properly. This eventually
2123 resulted in duplicates being sent. This issue only became common enough
2124 to notice due to the introduction of connection certificate information,
2125 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2127 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2128 size buffer was used, resulting in syntax errors when an expansion
2131 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2132 version 3.3.6 or later.
2134 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2135 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2136 both become "event_action", the variables become $event_name, $event_data
2137 and $event_defer_errno. There is a new variable $verify_mode, usable in
2138 routers, transports and related events. The tls:cert event is now also
2139 raised for inbound connections, if the main configuration event_action
2142 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2143 early OCSP support, but no stapling (appears to be less than 1.0.0).
2145 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2146 server certificate names available under the smtp transport option
2147 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2150 JH/10 Time-related extraction expansions from certificates now use the main
2151 option "timezone" setting for output formatting, and are consistent
2152 between OpenSSL and GnuTLS compilations. Bug 1541.
2154 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2155 encoded parameter in the incoming message. Bug 1558.
2157 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2158 include certificate info, eximon was claiming there were spoolfile
2161 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2163 JH/14 Log delivery-related information more consistently, using the sequence
2164 "H=<name> [<ip>]" wherever possible.
2166 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2167 are problematic for Debian distribution, omit them from the release
2170 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2172 JH/16 Fix string representation of time values on 64bit time_t architectures.
2175 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2176 output list separator was used.
2181 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2182 checkers that were complaining about end of non-void function with no
2185 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2186 This was a regression introduced in 4.83 by another bugfix.
2188 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2190 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2191 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2197 TF/01 Correctly close the server side of TLS when forking for delivery.
2199 When a message was received over SMTP with TLS, Exim failed to clear up
2200 the incoming connection properly after forking off the child process to
2201 deliver the message. In some situations the subsequent outgoing
2202 delivery connection happened to have the same fd number as the incoming
2203 connection previously had. Exim would try to use TLS and fail, logging
2204 a "Bad file descriptor" error.
2206 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2207 utilities have not been installed.
2209 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2210 temporary space as the ACL may create new global variables.
2212 TL/01 LDAP support uses per connection or global context settings, depending
2213 upon the detected version of the libraries at build time.
2215 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2216 to extract and use the src ip:port in logging and expansions as if it
2217 were a direct connection from the outside internet. PPv2 support was
2218 updated based on HAProxy spec change in May 2014.
2220 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2222 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2223 Properly escape header and check for NULL return.
2225 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2228 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2230 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2231 characters in header names, implemented as a verify condition.
2232 Contributed by Michael Fischer v. Mollard.
2234 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2235 results permerror and temperror. Previous values are deprecated but
2236 still accepted. In a future release, err_perm and err_temp will be
2237 completely removed, which will be a backward incompatibility if the
2238 ACL tests for either of these two old results. Patch contributed by
2239 user bes-internal on the mailing list.
2241 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2243 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2244 selectors, in both main and reject logs.
2246 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2249 JH/07 Add malware type "sock" for talking to simple daemon.
2251 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2253 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2254 routers/transports under cutthrough routing.
2256 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2257 numbers. Touch up "bool" conditional to keep the same definition.
2259 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2261 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2263 JH/12 Expand items in router/transport headers_add or headers_remove lists
2264 individually rather than the list as a whole. Bug 1452.
2266 Required for reasonable handling of multiple headers_ options when
2267 they may be empty; requires that headers_remove items with embedded
2268 colons must have them doubled (or the list-separator changed).
2270 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2271 view the policy declared in the DMARC record. Currently, $dmarc_status
2272 is a combined value of both the record presence and the result of the
2275 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2277 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2278 dnslookup router and the smtp transport (applying to the forward
2281 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2282 of ldap servers used for a specific lookup. Patch provided by Heiko
2285 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2286 New variable $lookup_dnssec_authenticated for observability.
2288 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2289 Patch submitted by Lars Timman.
2291 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2293 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2294 Requires trusted mode and valid format message id, aborts otherwise.
2295 Patch contributed by Heiko Schlichting.
2297 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2298 certextract with support for various fields. Bug 1358.
2300 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2301 is requested by default, modifiable by smtp transport option
2304 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2305 operate on certificate variables to give certificate fingerprints
2306 Also new ${sha256:cert_variable}.
2308 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2310 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2311 Christian Aistleitner.
2313 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2315 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2316 file. Patch from Wolfgang Breyha.
2318 JH/25 Expand the coverage of the delivery $host and $host_address to
2319 client authenticators run in verify callout. Bug 1476.
2321 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2322 align with daemon_smtp_ports. Bug 72.
2324 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2325 support and error reporting did not work properly.
2327 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2328 and is readable. Patch from Andrew Colin Kissa.
2330 TL/14 Enhance documentation of ${run expansion and how it parses the
2331 commandline after expansion, particularly in the case when an
2332 unquoted variable expansion results in an empty value.
2334 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2336 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2337 by renaming away from it. Addresses GH issue 15, reported by
2340 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2342 TL/15 SECURITY: prevent double expansion in math comparison functions
2343 (can expand unsanitized data). Not remotely exploitable.
2350 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2352 PP/02 Make -n do something, by making it not do something.
2353 When combined with -bP, the name of an option is not output.
2355 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2358 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2359 $sender_host_name and config options to manage this, and basic check
2362 PP/05 DSCP support for outbound connections and control modifier for inbound.
2364 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2365 (Only plugin which currently uses this is kerberos4, which nobody should
2366 be using, but we should make it available and other future plugins might
2367 conceivably use it, even though it would break NAT; stuff *should* be
2368 using channel bindings instead).
2370 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2371 name; added for Sendmail compatibility; requires admin caller.
2372 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2373 just ignore it); requires trusted caller.
2374 Also parse but ignore: -Ac -Am -X<logfile>
2377 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2379 TL/02 Add +smtp_confirmation as a default logging option.
2381 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2382 Patch by Magnus Holmgren from 2007-02-20.
2384 TL/04 Bugzilla 1281 - Spec typo.
2385 Bugzilla 1283 - Spec typo.
2386 Bugzilla 1290 - Spec grammar fixes.
2388 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2390 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2392 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2393 mailing list by by Dmitry Isaikin.
2395 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2397 JH/02 Support "G" suffix to numbers in ${if comparisons.
2399 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2401 NM/01 Bugzilla 1197 - Spec typo
2402 Bugzilla 1196 - Spec examples corrections
2404 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2406 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2407 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2410 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2411 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2413 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2414 "acl {{name}{arg}...}", and optional args on acl condition
2417 JH/05 Permit multiple router/transport headers_add/remove lines.
2419 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2421 JH/07 Avoid using a waiting database for a single-message-only transport.
2422 Performance patch from Paul Fisher. Bugzilla 1262.
2424 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2427 JH/09 Add $headers_added variable, with content from use of ACL modifier
2428 add_header (but not yet added to the message). Bugzilla 199.
2430 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2431 Pulled from Bugzilla 817 by Wolfgang Breyha.
2433 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2435 (nb: this is the same fix as in Exim 4.80.1)
2437 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2440 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2443 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2445 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2446 advertise SMTP AUTH mechanism to us, instead of a generic
2447 protocol violation error. Also, make Exim more robust to bad
2448 data from the Dovecot auth socket.
2450 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2452 When a queue runner is handling a message, Exim first routes the
2453 recipient addresses, during which it prunes them based on the retry
2454 hints database. After that it attempts to deliver the message to
2455 any remaining recipients. It then updates the hints database using
2458 So if a recipient address works intermittently, it can get repeatedly
2459 deferred at routing time. The retry hints record remains fresh so the
2460 address never reaches the final cutoff time.
2462 This is a fairly common occurrence when a user is bumping up against
2463 their storage quota. Exim had some logic in its local delivery code
2464 to deal with this. However it did not apply to per-recipient defers
2465 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2467 This change adds a proper retry rule check during routing so that the
2468 final cutoff time is checked against the message's age. We only do
2469 this check if there is an address retry record and there is not a
2470 domain retry record; this implies that previous attempts to handle
2471 the address had the retry_use_local_parts option turned on. We use
2472 this as an approximation for the destination being like a local
2473 delivery, as in LMTP.
2475 I suspect this new check makes the old local delivery cutoff check
2476 redundant, but I have not verified this so I left the code in place.
2478 TF/02 Correct gecos expansion when From: is a prefix of the username.
2480 Test 0254 submits a message to Exim with the header
2484 When I ran the test suite under the user fanf2, Exim expanded
2485 the header to contain my full name, whereas it should have added
2486 a Resent-Sender: header. It erroneously treats any prefix of the
2487 username as equal to the username.
2489 This change corrects that bug.
2491 GF/01 DCC debug and logging tidyup
2492 Error conditions log to paniclog rather than rejectlog.
2493 Debug lines prefixed by "DCC: " to remove any ambiguity.
2495 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2497 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2498 Bug spotted by Jeremy Harris; was flawed since initial commit.
2499 Would have resulted in OCSP responses post-SNI triggering an Exim
2500 NULL dereference and crash.
2502 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2504 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2505 Bug detection, analysis and fix by Samuel Thibault.
2506 Bugzilla 1331, Debian bug #698092.
2508 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2510 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2511 Server implementation by Todd Lyons, client by JH.
2512 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2513 config variable "prdr_enable" controls whether the server
2514 advertises the facility. If the client requests PRDR a new
2515 acl_data_smtp_prdr ACL is called once for each recipient, after
2516 the body content is received and before the acl_smtp_data ACL.
2517 The client is controlled by both of: a hosts_try_prdr option
2518 on the smtp transport, and the server advertisement.
2519 Default client logging of deliveries and rejections involving
2520 PRDR are flagged with the string "PRDR".
2522 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2523 fclose(). Diagnosis by Todd Lyons.
2525 PP/17 Update configure.default to handle IPv6 localhost better.
2526 Patch by Alain Williams (plus minor tweaks).
2529 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2530 This is now consistent with GnuTLS, and is now documented: the
2531 previous undocumented portable approach to treating the option as
2532 unset was to force an expansion failure. That still works, and
2533 an empty string is now equivalent.
2535 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2536 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2537 not performing validation itself.
2539 PP/20 Added force_command boolean option to pipe transport.
2540 Patch from Nick Koston, of cPanel Inc.
2542 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2545 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2547 PP/21 Fix eximon continuous updating with timestamped log-files.
2548 Broken in a format-string cleanup in 4.80, missed when I repaired the
2549 other false fix of the same issue.
2550 Report and fix from Heiko Schlichting.
2553 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2554 Report from Prashanth Katuri.
2556 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2557 It's SecureTransport, so affects any MacOS clients which use the
2558 system-integrated TLS libraries, including email clients.
2560 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2561 using a MIME ACL for non-SMTP local injection.
2562 Report and assistance in diagnosis by Warren Baker.
2564 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2566 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2568 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2569 last id that failed so it may be referenced in subsequent ACL's.
2571 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2574 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2575 ldap library initialization, allowing self-signed CA's to be
2576 used. Also properly sets require_cert option later in code by
2577 using NULL (global ldap config) instead of ldap handle (per
2578 session). Bug diagnosis and testing by alxgomz.
2580 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2581 the src/util/ subdirectory.
2583 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2584 renamed to Transport Post Delivery Action by Jeremy Harris, as
2587 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2588 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2589 redis_servers = needs to be configured which will be used by the redis
2590 lookup. Patch from Warren Baker, of The Packet Hub.
2592 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2594 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2595 hostname or reverse DNS when processing a host list. Used suggestions
2596 from multiple comments on this bug.
2598 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2600 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2601 Missed a few lines, added it to make the runtest require no keyboard
2604 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2605 contains upper case chars. Make router use caseful_local_part.
2607 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2608 support when GnuTLS has been built with p11-kit.
2614 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2616 This, or similar/improved, will also be change PP/11 of 4.82.
2622 PP/01 Handle short writes when writing local log-files.
2623 In practice, only affects FreeBSD (8 onwards).
2624 Bugzilla 1053, with thanks to Dmitry Isaikin.
2626 NM/01 Bugzilla 949 - Documentation tweak
2628 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2631 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2633 PP/02 Implemented gsasl authenticator.
2635 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2637 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2638 `pkg-config foo` for cflags/libs.
2640 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2641 with rest of GSASL and with heimdal_gssapi.
2643 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2644 `pkg-config foo` for cflags/libs for the TLS implementation.
2646 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2647 properties get this fed in as external SSF. A number of robustness
2648 and debugging improvements to the cyrus_sasl authenticator.
2650 PP/08 cyrus_sasl server now expands the server_realm option.
2652 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2653 Patch by Jeremy Harris.
2655 PP/10 Added dbmjz lookup type.
2657 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2659 PP/12 MAIL args handles TAB as well as SP, for better interop with
2660 non-compliant senders.
2661 Analysis and variant patch by Todd Lyons.
2663 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2664 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2665 Patch from Dirk Mueller <dmueller@suse.com>
2667 PP/13 tls_peerdn now print-escaped for spool files.
2668 Observed some $tls_peerdn in wild which contained \n, which resulted
2669 in spool file corruption.
2671 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2672 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2673 or write after TLS renegotiation, which otherwise led to messages
2676 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2677 as a tracking header (ie: a signed header comes before the signature).
2678 Patch from Wolfgang Breyha.
2680 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2681 comma-sep list; embedded commas doubled.
2683 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2685 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2687 Report and patch from Dmitry Banschikov.
2689 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2690 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2691 failures. We appear to now support TLS1.1+ with Exim.
2693 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2694 lets Exim select keys and certificates based upon TLS SNI from client.
2695 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2696 before an outbound SMTP session. New log_selector, +tls_sni.
2698 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2699 NULL dereference. Report and patch from Alun Jones.
2701 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2702 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2703 Not seeing resolver debug output on NetBSD, but suspect this is a
2704 resolver implementation change.
2706 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2707 Left warnings. Added "eximon gdb" invocation mode.
2709 PP/21 Defaulting "accept_8bitmime" to true, not false.
2711 PP/22 Added -bw for inetd wait mode support.
2713 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2714 locate the relevant includes and libraries. Made this the default.
2716 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2717 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2719 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2720 This may cause build issues on older platforms.
2722 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2723 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2724 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2725 Added SNI support via GnuTLS too.
2726 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2728 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2730 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2731 Applied second patch from Janne, implementing suggestion to default
2732 multiple-strings-in-record handling to match SPF spec.
2734 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2736 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2737 read-only, out of scope).
2738 Patch from Wolfgang Breyha, report from Stuart Northfield.
2740 PP/29 Fix three issues highlighted by clang analyser static analysis.
2741 Only crash-plausible issue would require the Cambridge-specific
2742 iplookup router and a misconfiguration.
2743 Report from Marcin Mirosław.
2745 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2747 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2748 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2749 As part of this, removing so much warning spew let me fix some minor
2750 real issues in debug logging.
2752 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2753 assignment on my part. Fixed.
2755 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2756 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2757 Janne Snabb (who went above and beyond: thank you).
2759 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2760 string otherwise requires a connection and a bunch more work and it's
2761 relatively easy to get wrong. Should also expose TLS library linkage
2764 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2765 64-bit ${eval} (JH/03).
2767 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2768 GNU libc to support some of the 64-bit stuff, should not lead to
2769 conflicts. Defined before os.h is pulled in, so if a given platform
2770 needs to override this, it can.
2772 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2773 protection layer was required, which is not implemented.
2774 Bugzilla 1254, patch from Wolfgang Breyha.
2776 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2777 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2778 tls_dhparam take prime identifiers. Also unbreak combination of
2779 OpenSSL+DH_params+TLSSNI.
2781 PP/39 Disable SSLv2 by default in OpenSSL support.
2787 PP/01 Solaris build fix for Oracle's LDAP libraries.
2788 Bugzilla 1109, patch from Stephen Usher.
2790 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2792 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2795 TF/02 Fix a couple more cases where we did not log the error message
2796 when unlink() failed. See also change 4.74-TF/03.
2798 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2799 lock up or crash if it happened to be inside a call to libc when it
2800 got a SIGUSR1 from exiwhat.
2802 The SIGUSR1 handler appends the current process status to the process
2803 log which is later printed by exiwhat. It used to use the general
2804 purpose logging code to do this, but several functions it calls are
2805 not safe for signals.
2807 The new output code in the SIGUSR1 handler is specific to the process
2808 log, and simple enough that it's easy to inspect for signal safety.
2809 Removing some special cases also simplifies the general logging code.
2810 Removing the spurious timestamps from the process log simplifies
2813 TF/04 Improved ratelimit ACL condition.
2815 The /noupdate option has been deprecated in favour of /readonly which
2816 has clearer semantics. The /leaky, /strict, and /readonly update modes
2817 are mutually exclusive. The update mode is no longer included in the
2818 database key; it just determines when the database is updated. (This
2819 means that when you upgrade Exim will forget old rate measurements.)
2821 Exim now checks that the per_* options are used with an update mode that
2822 makes sense for the current ACL. For example, when Exim is processing a
2823 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2824 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2825 must specify per_mail/readonly. If you omit the update mode it defaults to
2826 /leaky where that makes sense (as before) or /readonly where required.
2828 The /noupdate option is now undocumented but still supported for
2829 backwards compatibility. It is equivalent to /readonly except that in
2830 ACLs where /readonly is required you may specify /leaky/noupdate or
2831 /strict/noupdate which are treated the same as /readonly.
2833 A useful new feature is the /count= option. This is a generalization
2834 of the per_byte option, so that you can measure the throughput of other
2835 aggregate values. For example, the per_byte option is now equivalent
2836 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2838 The per_rcpt option has been generalized using the /count= mechanism
2839 (though it's more complicated than the per_byte equivalence). When it is
2840 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2841 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2842 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2843 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2844 Note that using per_rcpt with a non-readonly update mode in more than
2845 one ACL will cause the recipients to be double-counted. (The per_mail
2846 and per_byte options don't have this problem.)
2848 The handling of very low rates has changed slightly. If the computed rate
2849 is less than the event's count (usually one) then this event is the first
2850 after a long gap. In this case the rate is set to the same as this event's
2851 count, so that the first message of a spam run is counted properly.
2853 The major new feature is a mechanism for counting the rate of unique
2854 events. The new per_addr option counts the number of different
2855 recipients that someone has sent messages to in the last time period. It
2856 behaves like per_rcpt if all the recipient addresses are different, but
2857 duplicate recipient addresses do not increase the measured rate. Like
2858 the /count= option this is a general mechanism, so the per_addr option
2859 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2860 example, measure the rate that a client uses different sender addresses
2861 with the options per_mail/unique=$sender_address. There are further
2862 details in the main documentation.
2864 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2866 TF/06 Removed a few PCRE remnants.
2868 TF/07 Automatically extract Exim's version number from tags in the git
2869 repository when doing development or release builds.
2871 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2872 Bugzilla 879. Patch from Paul Fisher.
2874 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2875 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2878 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2880 PP/05 Variable $av_failed, true if the AV scanner deferred.
2881 Bugzilla 1078. Patch from John Horne.
2883 PP/06 Stop make process more reliably on build failure.
2884 Bugzilla 1087. Patch from Heiko Schlittermann.
2886 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2887 Bugzilla 1089. Patch from Heiko Schlittermann.
2889 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2890 Bugzilla 1131. Patch from Holger Weiß.
2892 PP/09 Handle IPv6 addresses with SPF.
2893 Bugzilla 860. Patch from Wolfgang Breyha.
2895 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2897 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2900 PP/11 match_* no longer expand right-hand-side by default.
2901 New compile-time build option, EXPAND_LISTMATCH_RHS.
2902 New expansion conditions, "inlist", "inlisti".
2904 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2906 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2908 PP/14 fix log_write() format string regression from TF/03.
2909 Bugzilla 1152. Patch from Dmitry Isaikin.
2915 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2917 PP/02 Harmonised TLS library version reporting; only show if debugging.
2918 Layout now matches that introduced for other libraries in 4.74 PP/03.
2920 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2922 PP/04 New "dns_use_edns0" global option.
2924 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2927 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2928 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2930 TK/01 Updated PolarSSL code to 0.14.2.
2931 Bugzilla 1097. Patch from Andreas Metzler.
2933 PP/07 Catch divide-by-zero in ${eval:...}.
2934 Fixes bugzilla 1102.
2936 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2939 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2940 format-string attack -- SECURITY: remote arbitrary code execution.
2942 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2943 time unintentionally subject to list matching rules, letting the header
2944 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2945 arbitrary string expansion). This allowed for information disclosure.
2947 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2948 INT_MIN/-1 -- value coerced to INT_MAX.
2954 NM/01 Workaround for PCRE version dependency in version reporting
2957 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2958 This fixes portability to compilers other than gcc, notably
2959 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2961 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2962 makefiles for portability to HP-UX and POSIX correctness.
2964 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2965 Also via indented variable definition in the Makefile.
2966 (Debugging by Oliver Heesakkers).
2968 PP/02 Restore caching of spamd results with expanded spamd_address.
2969 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2971 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2972 Improves build reliability. Fix from: Frank Elsner
2974 NM/02 Fix wide character breakage in the rfc2047 coding
2975 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2977 NM/03 Allow underscore in dnslist lookups
2978 Fixes bug 1026. Patch from Graeme Fowler
2980 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2981 Code patches from Adam Ciarcinski of NetBSD.
2983 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2986 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2987 is logged, to avoid truncation. Patch from John Horne.
2989 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2990 Patch from Jakob Hirsch.
2992 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2993 SQL string expansion failure details.
2994 Patch from Andrey Oktyabrski.
2996 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2997 Patch from Simon Arlott.
2999 PP/09 New lookups functionality failed to compile on old gcc which rejects
3000 extern declarations in function scope.
3001 Patch from Oliver Fleischmann
3003 PP/10 Use sig_atomic_t for flags set from signal handlers.
3004 Check getgroups() return and improve debugging.
3005 Fixed developed for diagnosis in bug 927 (which turned out to be
3008 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3009 Patch from Mark Zealey.
3011 PP/12 Bugzilla 1056: Improved spamd server selection.
3012 Patch from Mark Zealey.
3014 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3015 Based on patch from Heiko Schlittermann.
3017 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3018 Patch from Uwe Doering, sign-off by Michael Haardt.
3020 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3021 variable declaration deep within a block. Bug and patch from
3024 PP/15 lookups-Makefile IRIX compatibility coercion.
3026 PP/16 Make DISABLE_DKIM build knob functional.
3028 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3029 Patch by Simon Arlott
3031 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3032 variable argument macros. Our copy now differs from upstream.
3038 TF/01 Failure to get a lock on a hints database can have serious
3039 consequences so log it to the panic log.
3041 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3042 controlled using the smtp_confirmation log selector.
3044 TF/03 Include the error message when we fail to unlink a spool file.
3046 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3047 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3048 for maintaining out-of-tree patches for some time.
3050 PP/01 Bugzilla 139: Documentation and portability issues.
3051 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3052 Handle per-OS dynamic-module compilation flags.
3054 PP/02 Let /dev/null have normal permissions.
3055 The 4.73 fixes were a little too stringent and complained about the
3056 permissions on /dev/null. Exempt it from some checks.
3057 Reported by Andreas M. Kirchwitz.
3059 PP/03 Report version information for many libraries, including
3060 Exim version information for dynamically loaded libraries. Created
3061 version.h, now support a version extension string for distributors
3062 who patch heavily. Dynamic module ABI change.
3064 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3065 privilege escalation vulnerability whereby the Exim run-time user
3066 can cause root to append content of the attacker's choosing to
3069 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3072 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3073 If dropping privileges for untrusted macros, we disabled normal logging
3074 on the basis that it would fail; for the Exim run-time user, this is not
3075 the case, and it resulted in successful deliveries going unlogged.
3076 Fixed. Reported by Andreas Metzler.
3082 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3083 only prepend for the Resent-* case. Fixes regression introduced in
3084 Exim 4.70 by NM/22 for Bugzilla 607.
3086 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3087 increasing numbers of administrators be bitten by this.
3089 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3091 PP/03 Bugzilla 994: added openssl_options main configuration option.
3093 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3095 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3097 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3099 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3100 without a peer certificate, leading to a segfault because of an
3101 assumption that peers always have certificates. Be a little more
3102 paranoid. Problem reported by Martin Tscholak.
3104 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3105 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3106 NB: ClamAV planning to remove STREAM in "middle of 2010".
3107 CL also introduces -bmalware, various -d+acl logging additions and
3108 more caution in buffer sizes.
3110 PP/09 Implemented reverse_ip expansion operator.
3112 PP/10 Bugzilla 937: provide a "debug" ACL control.
3114 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3116 PP/12 Bugzilla 973: Implement --version.
3118 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3120 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3122 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3124 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3125 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3126 ignore trailing whitespace.
3128 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3130 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3133 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3134 Notification from Dr Andrew Aitchison.
3136 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3137 ExtendedDetectionInfo response format.
3138 Notification from John Horne.
3140 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3143 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3144 XSL and documented dependency on system catalogs, with examples of how
3147 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3150 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3151 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3152 configuration file which is writeable by the Exim user or group.
3154 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3155 of configuration files to cover files specified with the -C option if
3156 they are going to be used with root privileges, not just the default
3159 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3160 option (effectively making it always true).
3162 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3163 files to be used while preserving root privileges.
3165 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3166 that rogue child processes cannot use them.
3168 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3169 run-time user, instead of root.
3171 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3172 Exim run-time user without dropping privileges.
3174 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3175 result string, instead of calling string_vformat() twice with the same
3178 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3179 for other users. Others should always drop root privileges if they use
3180 -C on the command line, even for a whitelisted configure file.
3182 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3184 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3190 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3191 $data_path, and $header_path variables; fixed documentation bugs and
3194 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3195 exipick to access non-standard spools, including the "frozen" queue
3198 NM/01 Bugzilla 965: Support mysql stored procedures.
3199 Patch from Alain Williams
3201 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3203 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3204 Patch from Andreas Metzler
3206 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3207 Patch from Kirill Miazine
3209 NM/05 Bugzilla 671: Added umask to procmail example.
3211 JJ/03 installed exipick 20100323.0, fixing doc bug
3213 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3214 directory. Notification and patch from Dan Rosenberg.
3216 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3218 TK/02 Improve log output when DKIM signing operation fails.
3220 MH/01 Treat the transport option dkim_domain as a colon separated
3221 list, not as a single string, and sign the message with each element,
3222 omitting multiple occurences of the same signer.
3224 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3225 Bugzilla 985, 986. Patch by Simon Arlott
3227 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3228 Patch by Simon Arlott
3230 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3231 MBX locking. Notification from Dan Rosenberg.
3237 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3239 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3241 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3243 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3245 NM/04 Bugzilla 915: Flag broken perl installation during build.
3251 TK/01 Added patch by Johannes Berg that expands the main option
3252 "spamd_address" if it starts with a dollar sign.
3254 TK/02 Write list of recipients to X-Envelope-Sender header when building
3255 the mbox-format spool file for content scanning (suggested by Jakob
3258 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3259 (http://www.dcc-servers.net/) support via dccifd. Activated by
3260 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3262 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3263 by Mark Daniel Reidel <mr@df.eu>.
3265 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3266 When building exim an external PCRE library is now needed -
3267 PCRE is a system library on the majority of modern systems.
3268 See entry on PCRE_LIBS in EDITME file.
3270 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3271 conversation. Added nologin parameter to request.
3272 Patch contributed by Kirill Miazine.
3274 TF/01 Do not log submission mode rewrites if they do not change the address.
3276 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3278 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3279 log files in place. Contributed by Roberto Lima.
3281 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3283 TF/03 Bugzilla 615: When checking the local_parts router precondition
3284 after a local_part_suffix or local_part_prefix option, Exim now
3285 does not use the address's named list lookup cache, since this
3286 contains cached lookups for the whole local part.
3288 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3289 Robert Millan. Documentation is in experimental-spec.txt.
3291 TF/04 Bugzilla 668: Fix parallel build (make -j).
3293 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3295 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3296 Patch provided by Jan Srzednicki.
3298 TF/05 Leading white space used to be stripped from $spam_report which
3299 wrecked the formatting. Now it is preserved.
3301 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3302 that they are available at delivery time.
3304 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3306 TF/08 TLS error reporting now respects the incoming_interface and
3307 incoming_port log selectors.
3309 TF/09 Produce a more useful error message if an SMTP transport's hosts
3310 setting expands to an empty string.
3312 NM/06 Bugzilla 744: EXPN did not work under TLS.
3313 Patch provided by Phil Pennock.
3315 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3316 Patch provided by Richard Godbee.
3318 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3319 acl_smtp_notquit, added index entry.
3321 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3322 Patch provided by Eugene Bujak.
3324 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3325 accept(). Patch provided by Maxim Dounin.
3327 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3328 Patch provided by Phil Pennock.
3330 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3332 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3333 Patch provided by Brad "anomie" Jorsch.
3335 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3336 Patch provided by Dean Brooks.
3338 TK/05 Add native DKIM support (does not depend on external libraries).
3340 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3341 Patch provided by Graeme Fowler.
3343 NM/16 Bugzilla 851: Documentation example syntax fix.
3345 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3347 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3350 NM/19 Bugzilla 745: TLS version reporting.
3351 Patch provided by Phil Pennock.
3353 NM/20 Bugzilla 167: bool: condition support.
3354 Patch provided by Phil Pennock.
3356 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3357 clients. Patch provided by Phil Pennock.
3359 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3360 Patch provided by Brad "anomie" Jorsch.
3362 NM/23 Bugzilla 687: Fix misparses in eximstats.
3363 Patch provided by Heiko Schlittermann.
3365 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3366 Patch provided by Heiko Schlittermann.
3368 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3369 plus update to original patch.
3371 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3373 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3374 Patch provided by David Brownlee.
3376 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3378 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3380 NM/30 Bugzilla 888: TLS documentation bugfixes.
3382 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3384 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3385 Unlike the original bugzilla I have changed all shell scripts in src tree.
3387 NM/33 Bugzilla 898: Transport filter timeout fix.
3388 Patch by Todd Rinaldo.
3390 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3391 Patch by Serge Demonchaux.
3393 NM/35 Bugzilla 39: Base64 decode bug fixes.
3394 Patch by Jakob Hirsch.
3396 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3398 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3400 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3402 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3408 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3409 ALT-N's libdkim that I have put here:
3410 http://duncanthrax.net/exim-experimental/
3412 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3413 were called 'true' and it seems that C99 defines that as a reserved
3414 keyword to be used with 'bool' variable types. That means you could
3415 not include C99-style headers which use bools without triggering
3416 build errors in sieve.c.
3418 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3419 as mailq or other aliases. Changed the --help handling significantly
3420 to do whats expected. exim_usage() emits usage/help information.
3422 SC/01 Added the -bylocaldomain option to eximstats.
3424 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3426 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3428 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3434 PH/01 Another patch from the Sieve maintainer.
3436 PH/02 When an IPv6 address is converted to a string for single-key lookup
3437 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3438 dots are used instead of colons so that keys in lsearch files need not
3439 contain colons. This was done some time before quoting was made available
3440 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3441 (notated using the quote facility) so as to distinguish them from IPv4
3442 keys. This meant that lookups for IP addresses in host lists did not work
3443 for iplsearch lookups.
3445 This has been fixed by arranging for IPv6 addresses to be expressed with
3446 colons if the lookup type is iplsearch. This is not incompatible, because
3447 previously such lookups could never work.
3449 The situation is now rather anomalous, since one *can* have colons in
3450 ordinary lsearch keys. However, making the change in all cases is
3451 incompatible and would probably break a number of configurations.
3453 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3456 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3457 conversion specification without a maximum field width, thereby enabling
3458 a rogue spamd server to cause a buffer overflow. While nobody in their
3459 right mind would setup Exim to query an untrusted spamd server, an
3460 attacker that gains access to a server running spamd could potentially
3461 exploit this vulnerability to run arbitrary code as the Exim user.
3463 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3464 $primary_hostname instead of what libspf2 thinks the hosts name is.
3466 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3467 a directory entry by the name of the lookup key. Previously, if a
3468 symlink pointed to a non-existing file or a file in a directory that
3469 Exim lacked permissions to read, a lookup for a key matching that
3470 symlink would fail. Now it is enough that a matching directory entry
3471 exists, symlink or not. (Bugzilla 503.)
3473 PH/03 The body_linecount and body_zerocount variables are now exported in the
3476 PH/04 Added the $dnslist_matched variable.
3478 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3479 This means they are set thereafter only if the connection becomes
3482 PH/06 Added the client_condition to authenticators so that some can be skipped
3483 by clients under certain conditions.
3485 PH/07 The error message for a badly-placed control=no_multiline_responses left
3486 "_responses" off the end of the name.
3488 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3490 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3491 (without spaces) instead of just copying the configuration text.
3493 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3495 PH/11 Added $max_received_linelength.
3497 PH/12 Added +ignore_defer and +include_defer to host lists.
3499 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3500 way in which PCRE > 7.0 is built.
3502 PH/14 Implemented queue_only_load_latch.
3504 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3505 MAIL command. The effect was to mangle the value on 64-bit systems.
3507 PH/16 Another patch from the Sieve maintainer.
3509 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3511 PH/18 If a system quota error occurred while trying to create the file for
3512 a maildir delivery, the message "Mailbox is full" was not appended to the
3513 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3514 only to a quota excession during the actual writing of the file.
3516 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3517 characters?) which causes problems in log lines. The DN values are now
3518 passed through string_printing() before being added to log lines.
3520 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3521 and InterBase are left for another time.)
3523 PH/21 Added message_body_newlines option.
3525 PH/22 Guard against possible overflow in moan_check_errorcopy().
3527 PH/23 POSIX allows open() to be a macro; guard against that.
3529 PH/24 If the recipient of an error message contained an @ in the local part
3530 (suitably quoted, of course), incorrect values were put in $domain and
3531 $local_part during the evaluation of errors_copy.
3537 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3538 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3541 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3542 issue a MAIL command.
3544 PH/02 In an ACL statement such as
3546 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3548 if a client was not listed at all, or was listed with a value other than
3549 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3550 the condition was not true (as it should be), so access was not denied.
3551 The bug was that the ! inversion was incorrectly passed on to the second
3552 item. This has been fixed.
3554 PH/03 Added additional dnslists conditions == and =& which are different from
3555 = and & when the dns lookup returns more than one IP address.
3557 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3558 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3560 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3561 FSYNC, which compiles an option called disable_fsync that allows for
3562 bypassing fsync(). The documentation is heavily laced with warnings.
3564 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3566 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3567 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3568 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3569 including adding "make clean"; (3) Added -fPIC when compiling the test
3570 dynamically loaded module, to get rid of a warning.
3572 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3573 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3574 = 0s. The bug is otherwise harmless.
3576 PH/07 There was a bug in the dovecot authenticator such that the value of
3577 $auth1 could be overwritten, and so not correctly preserved, after a
3578 successful authentication. This usually meant that the value preserved by
3579 the server_setid option was incorrect.
3581 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3583 PH/09 Installed PCRE release 7.0.
3585 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3586 run for batched SMTP input. It is now run at the start of every message
3587 in the batch. While fixing this I discovered that the process information
3588 (output by running exiwhat) was not always getting set for -bs and -bS
3589 input. This is fixed, and it now also says "batched" for BSMTP.
3591 PH/11 Added control=no_pipelining.
3593 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3594 patch, slightly modified), and move the expansion of helo_data till after
3595 the connection is made in the smtp transport (so it can use these
3598 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3600 PH/14 Added log_selector = +pid.
3602 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3604 PH/16 Add ${if forany and ${if forall.
3606 PH/17 Added dsn_from option to vary the From: line in DSNs.
3608 PH/18 Flush SMTP output before performing a callout, unless control =
3609 no_callout_flush is set.
3611 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3612 was true (the default) a successful delivery failed to delete the retry
3613 item, thus causing premature timeout of the address. The bug is now
3616 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3618 PH/21 Long custom messages for fakedefer and fakereject are now split up
3619 into multiline responses in the same way that messages for "deny" and
3620 other ACL rejections are.
3622 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3623 with slight modification.
3625 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3626 draft, changing the syntax and factoring some duplicate code".
3628 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3629 for deliveries of the second and subsequent messages over the same SMTP
3632 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3633 ${reduce, with only minor "tidies".
3635 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3637 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3638 expansion side effects.
3640 PH/27 When a message times out after an over-quota error from an Exim-imposed
3641 quota, the bounce message says "mailbox is full". This message was not
3642 being given when it was a system quota that was exceeded. It now should
3645 MH/03 Made $recipients available in local_scan(). local_scan() already has
3646 better access to the recipient list through recipients_list[], but
3647 $recipients can be useful in postmaster-provided expansion strings.
3649 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3650 in the case of a MAIL command with additional options following the
3651 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3652 were accidentally chopped off.
3654 PH/29 SMTP synchronization checks are implemented when a command is read -
3655 there is a check that no more input is waiting when there shouldn't be
3656 any. However, for some commands, a delay in an ACL can mean that it is
3657 some time before the response is written. In this time, more input might
3658 arrive, invalidly. So now there are extra checks after an ACL has run for
3659 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3660 pipelining has not been advertised.
3662 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3664 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3665 returned in the SMTP response when smtp_return_error_details was false.
3666 This has been fixed.
3668 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3669 instead of the C I/O that was originally supplied, because problems were
3670 reported on Solaris.
3672 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3673 Exim which did not show up earlier: it was assuming that a call to
3674 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3675 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3676 was a macro that became an assignment, so it seemed to work. This has
3677 changed to a proper function call with a void return, hence the compile
3678 error. Exim's code has been fixed.
3680 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3683 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3685 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3687 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3688 because it turns out to be too big or there is a timeout) there is no
3689 "Completed" line in the log. When some messages of this type were
3690 selected by exigrep, they were listed as "not completed". Others were
3691 picked up by some special patterns. I have improved the selection
3692 criteria to be more general.
3694 PH/38 The host_find_failed option in the manualroute router can now be set
3695 to "ignore", to completely ignore a host whose IP address cannot be
3696 found. If all hosts are ignored, the behaviour is controlled by the new
3697 host_all_ignored option.
3699 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3700 homing or because of multiple MX records with /mx) generated more than
3701 one IP address, and the following item turned out to be the local host,
3702 all the secondary addresses of the first item were incorrectly removed
3703 from the list, along with the local host and any following hosts (which
3704 is what is supposed to happen).
3706 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3707 whoever called Exim into the -H file. In the case of the daemon it was
3708 behaving confusingly. When first started, it used values for whoever
3709 started the daemon, but after a SIGHUP it used the Exim user (because it
3710 calls itself on a restart). I have changed the code so that it now always
3713 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3714 message are rejected with the same error (e.g. no authentication or bad
3715 sender address), and a DATA command is nevertheless sent (as can happen
3716 with PIPELINING or a stupid MUA), the error message that was given to the
3717 RCPT commands is included in the rejection of the DATA command. This is
3718 intended to be helpful for MUAs that show only the final error to their
3721 PH/42 Another patch from the Sieve maintainer.
3723 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3724 Eximstats - Fixed some broken HTML links and added missing column headers
3726 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3727 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3729 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3731 PH/43 Yet another patch from the Sieve maintainer.
3733 PH/44 I found a way to check for a TCP/IP connection going away before sending
3734 the response to the final '.' that terminates a message, but only in the
3735 case where the client has not sent further data following the '.'
3736 (unfortunately, this is allowed). However, in many cases there won't be
3737 any further data because there won't be any more messages to send. A call
3738 to select() can be used: if it shows that the input is "ready", there is
3739 either input waiting, or the socket has been closed. An attempt to read
3740 the next input character can distinguish the two cases. Previously, Exim
3741 would have sent an OK response which the client would never have see.
3742 This could lead to message repetition. This fix should cure that, at
3743 least in a lot of common cases.
3745 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3746 advertised in response to EHLO.
3752 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3753 fixed by 4.65/MH/01 (is this a record?) are fixed:
3755 (i) An empty string was always treated as zero by the numeric comparison
3756 operators. This behaviour has been restored.
3758 (ii) It is documented that the numeric comparison operators always treat
3759 their arguments as decimal numbers. This was broken in that numbers
3760 starting with 0 were being interpreted as octal.
3762 While fixing these problems I realized that there was another issue that
3763 hadn't been noticed. Values of message_size_limit (both the global option
3764 and the transport option) were treated as octal if they started with 0.
3765 The documentation was vague. These values are now always treated as
3766 decimal, and I will make that clear in the documentation.
3772 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3773 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3776 MH/01 Don't check that the operands of numeric comparison operators are
3777 integers when their expansion is in "skipping" mode (fixes bug
3778 introduced by 4.64-PH/07).
3780 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3781 child addresses, Exim now panics and dies. Previously, because the count
3782 is held in a short int, deliveries were likely to be lost. As such a
3783 large number of recipients for a single message is ridiculous
3784 (performance will be very, very poor), I have chosen to impose a limit
3785 rather than extend the field.
3791 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3792 leftover -K file (the existence of which was triggered by #402).
3793 While we were at it, introduced process PID as part of the -K
3794 filename. This should rule out race conditions when creating
3797 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3798 processing considerably. Previous code took too long for large mails,
3799 triggering a timeout which in turn triggers #401.
3801 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3802 in the DK code in transports.c. sendfile() is not really portable,
3803 hence the _LINUX specificness.
3805 TF/01 In the add_headers option to the mail command in an Exim filter,
3806 there was a bug that Exim would claim a syntax error in any
3807 header after the first one which had an odd number of characters
3810 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3811 callout verification, Exim cached a "reject" for the entire domain. This
3812 is correct for most verifications, but it is not correct for a recipient
3813 verification with use_sender or use_postmaster set, because in that case
3814 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3815 case of MAIL FROM:<> rejection from other early rejections (e.g.
3816 rejection of HELO). When verifying a recipient using a non-null MAIL
3817 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3818 Whatever the result of the callout, the value of the domain cache is
3819 left unchanged (for any other kind of callout, getting as far as trying
3820 RCPT means that the domain itself is ok).
3822 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3825 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3826 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3827 session, the daemon ignores ECONNECT errors and logs others; it now
3828 ignores EPIPE as well.
3830 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3831 (quoted-printable decoding).
3833 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3834 later the small subsequent patch to fix an introduced bug.
3836 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3838 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3840 PH/08 An error is now given if message_size_limit is specified negative.
3842 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3843 to be given (somewhat) arbitrary names.
3845 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3848 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3849 miscellaneous code fixes
3851 PH/10 Added the log_reject_target ACL modifier to specify where to log
3854 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3855 hostname. This is wrong, because it relates to the incoming message (and
3856 probably the interface on which it is arriving) and not to the outgoing
3857 callout (which could be using a different interface). This has been
3858 changed to use the value of the helo_data option from the smtp transport
3859 instead - this is what is used when a message is actually being sent. If
3860 there is no remote transport (possible with a router that sets up host
3861 addresses), $smtp_active_hostname is used.
3863 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3864 tweaks were necessary in order to get it to work (see also 21 below):
3865 (a) The code assumed that strncpy() returns a negative number on buffer
3866 overflow, which isn't the case. Replaced with Exim's string_format()
3868 (b) There were several signed/unsigned issues. I just did the minimum
3869 hacking in of casts. There is scope for a larger refactoring.
3870 (c) The code used strcasecmp() which is not a standard C function.
3871 Replaced with Exim's strcmpic() function.
3872 (d) The code set only $1; it now sets $auth1 as well.
3873 (e) A simple test gave the error "authentication client didn't specify
3874 service in request". It would seem that Dovecot has changed its
3875 interface. Fortunately there's a specification; I followed it and
3876 changed what the client sends and it appears to be working now.
3878 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3881 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3882 address A is aliased to B and C, where B exists and C does not. Without
3883 -v the output is "A verified" because verification stops after a
3884 successful redirection if more than one address is generated. However,
3885 with -v the child addresses are also verified. Exim was outputting "A
3886 failed to verify" and then showing the successful verification for C,
3887 with its parentage. It now outputs "B failed to verify", showing B's
3888 parentage before showing the successful verification of C.
3890 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3891 look up a TXT record in a specific list after matching in a combined
3894 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3895 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3896 they consult the DNS. I had assumed they would set it the way they
3897 wanted; and indeed my experiments on Linux seem to show that in some
3898 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3899 To be on the safe side, however, I have now made the interface to
3900 host_find_byname() similar to host_find_bydns(), with an argument
3901 containing the DNS resolver options. The host_find_byname() function now
3902 sets these options at its start, just as host_find_bydns() does. The smtp
3903 transport options dns_qualify_single and dns_search_parents are passed to
3904 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3905 of host_find_byname() use the default settings of RES_DEFNAMES
3906 (qualify_single) but not RES_DNSRCH (search_parents).
3908 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3909 spool_read_header() do less string testing, by means of a preliminary
3910 switch on the second character of optional "-foo" lines. (This is
3911 overdue, caused by the large number of possibilities that now exist.
3912 Originally there were few.) While I was there, I also converted the
3913 str(n)cmp tests so they don't re-test the leading "-" and the first
3914 character, in the hope this might squeeze out yet more improvement.
3916 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3917 flag allowing group syntax was set by the header_syntax check but not
3918 turned off, possible causing trouble later; (2) The flag was not being
3919 set at all for the header_verify test, causing "group"-style headers to
3920 be rejected. I have now set it in this case, and also caused header_
3921 verify to ignore an empty address taken from a group. While doing this, I
3922 came across some other cases where the code for allowing group syntax
3923 while scanning a header line wasn't quite right (mostly, not resetting
3924 the flag correctly in the right place). These bugs could have caused
3925 trouble for malformed header lines. I hope it is now all correct.
3927 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3928 with the "reply" argument non-NULL. The code, however (which originally
3929 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3930 but it didn't always do it. This confused somebody who was copying the
3931 code for some other use. I have removed all the tests.
3933 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3934 feature that was used to support insecure browsers during the U.S. crypto
3935 embargo. It requires special client support, and Exim is probably the
3936 only MTA that supported it -- and would never use it because real RSA is
3937 always available. This code has been removed, because it had the bad
3938 effect of slowing Exim down by computing (never used) parameters for the
3939 RSA_EXPORT functionality.
3941 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3942 authenticator to fail if there's a tab character in the incoming data
3943 (there should never be unless someone is messing about, as it's supposed
3944 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3945 if the connection is using TLS or if the remote IP is the same as the
3946 local IP, and the "valid-client-cert option" if a client certificate has
3949 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3950 authenticators. This can be used for authorization after authentication
3951 succeeds. (In the case of plaintext, it servers for both authentication
3954 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3955 if any retry times were supplied.
3957 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3958 connection, where there is no client IP address to check. In this
3959 situation, the verify now always succeeds.
3961 PH/25 Applied John Jetmore's -Mset patch.
3963 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3965 PH/27 In a string expansion for a processed (not raw) header when multiple
3966 headers of the same name were present, leading whitespace was being
3967 removed from all of them, but trailing whitespace was being removed only
3968 from the last one. Now trailing whitespace is removed from each header
3969 before concatenation. Completely empty headers in a concatenation (as
3970 before) are ignored.
3972 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3973 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3975 PH/29 [Removed. This was a change that I later backed out, and forgot to
3976 correct the ChangeLog entry (that I had efficiently created) before
3977 committing the later change.]
3979 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3980 address errors (4xx response to RCPT) over the same connection as other
3981 messages routed to the same hosts. Such deliveries are always "forced",
3982 so retry times are not inspected. This resulted in far too many retries
3983 for the affected addresses. The effect occurred only when there were more
3984 hosts than the hosts_max_try setting in the smtp transport when it had
3985 the 4xx errors. Those hosts that it had tried were not added to the list
3986 of hosts for which the message was waiting, so if all were tried, there
3987 was no problem. Two fixes have been applied:
3989 (i) If there are any address or message errors in an SMTP delivery, none
3990 of the hosts (tried or untried) are now added to the list of hosts
3991 for which the message is waiting, so the message should not be a
3992 candidate for sending over the same connection that was used for a
3993 successful delivery of some other message. This seems entirely
3994 reasonable: after all the message is NOT "waiting for some host".
3995 This is so "obvious" that I'm not sure why it wasn't done
3996 previously. Hope I haven't missed anything, but it can't do any
3997 harm, as the worst effect is to miss an optimization.
3999 (ii) If, despite (i), such a delivery is accidentally attempted, the
4000 routing retry time is respected, so at least it doesn't keep
4001 hammering the server.
4003 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4004 in ${readsocket because some servers need this prod.
4006 PH/32 Added some extra debug output when updating a wait-xxx database.
4008 PH/33 The hint "could be header name not terminated by colon", which has been
4009 given for certain expansion errors for a long time, was not being given
4010 for the ${if def:h_colon_omitted{... case.
4012 PH/34 The spec says: "With one important exception, whenever a domain list is
4013 being scanned, $domain contains the subject domain." There was at least
4014 one case where this was not true.
4016 PH/35 The error "getsockname() failed: connection reset by peer" was being
4017 written to the panic log as well as the main log, but it isn't really
4018 panic-worthy as it just means the connection died rather early on. I have
4019 removed the panic log writing for the ECONNRESET error when getsockname()
4022 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4023 runs only) independently of the message's sender address. This meant
4024 that, if the 4xx error was in fact related to the sender, a different
4025 message to the same recipient with a different sender could confuse
4026 things. In particular, this can happen when sending to a greylisting
4027 server, but other circumstances could also provoke similar problems.
4028 I have changed the default so that the retry time for these errors is now
4029 based a combination of the sender and recipient addresses. This change
4030 can be overridden by setting address_retry_include_sender=false in the
4033 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4034 remote server are returned as part of bounce messages. This was not
4035 happening for LMTP over a pipe (the lmtp transport), but now it is the
4036 same for both kinds of LMTP.
4038 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4039 in header lines that were in fact CNAMEs. This is no longer the case.
4041 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4042 and queue runs started by the daemon processed all messages. This has
4043 been fixed so that -R and -S can now usefully be given with -q<time>.
4045 PH/40 Import PCRE release 6.7 (fixes some bugs).
4047 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4049 PH/42 Give an error if -q is specified more than once.
4051 PH/43 Renamed the variables $interface_address and $interface_port as
4052 $received_ip_address and $received_port, to make it clear that these
4053 values apply to message reception, and not to the outgoing interface when
4054 a message is delivered. (The old names remain recognized, of course.)
4056 PH/44 There was no timeout on the connect() call when using a Unix domain
4057 socket in the ${readsocket expansion. There now is.
4059 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4060 be meaningful with "accept".
4062 SC/01 Eximstats V1.43
4063 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4065 SC/02 Eximstats V1.44
4066 Use a glob alias rather than an array ref in the generated
4067 parser. This improves both readability and performance.
4069 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4070 Collect SpamAssassin and rejection statistics.
4071 Don't display local sender or destination tables unless
4072 there is data to show.
4073 Added average volumes into the top table text output.
4075 SC/04 Eximstats V1.46
4076 Collect data on the number of addresses (recipients)
4077 as well as the number of messages.
4079 SC/05 Eximstats V1.47
4080 Added 'Message too big' to the list of mail rejection
4081 reasons (thanks to Marco Gaiarin).
4083 SC/06 Eximstats V1.48
4084 Mainlog lines which have GMT offsets and are too short to
4085 have a flag are now skipped.
4087 SC/07 Eximstats V1.49 (Alain Williams)
4088 Added the -emptyok flag.
4090 SC/08 Eximstats V1.50
4091 Fixes for obtaining the IP address from reject messages.
4093 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4094 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4095 whitespace changes from 4.64-PH/27
4097 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4100 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4101 are found, allow negative numbers in numeric criteria)
4103 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4105 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4106 to match changes made in 4.64-PH/43
4108 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4110 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4111 "domain" is actually an IP address, and force a failure. This locks out
4112 those revolvers/nameservers that support "A-for-A" lookups, in
4113 contravention of the specifications.
4115 PH/48 When a host name was looked up from an IP address, and the subsequent
4116 forward lookup of the name timed out, the host name was left in
4117 $sender_host_name, contrary to the specification.
4119 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4120 restricted to single-key lookups, Exim was not diagnosing an error if
4121 * or *@ was used with a query-style lookup.
4123 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4125 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4126 long ago, but noone interested enough thought of it. Let's just say that
4127 the "1.1" means that there are some new functions that weren't there at
4128 some point in the past.
4130 PH/51 Error processing for expansion failure of helo_data from an smtp
4131 transport during callout processing was broken.
4133 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4134 tested/used via the -bh/-bhc/-bs options.
4136 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4137 bug, fixed in subsequent PCRE releases).
4139 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4140 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4142 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4148 SC/01 Use a glob alias rather than an array ref in eximstats generated
4149 parser. This improves both readability and performance.
4151 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4152 Don't display local sender or destination tables in eximstats unless
4153 there is data to show.
4154 Added average volumes into the eximstats top table text output.
4156 SC/03 Collect data on the number of addresses (recipients) as well
4157 as the number of messages in eximstats.
4159 TF/01 Correct an error in the documentation for the redirect router. Exim
4160 does (usually) call initgroups() when daemonizing.
4162 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4163 with consistent privilege compared to when running as a daemon.
4165 TF/03 Note in the spec that $authenticated_id is not set for local
4166 submissions from trusted users.
4168 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4169 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4171 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4172 by adding some example configuration directives to the default
4173 configuration file. A little bit of work is required to uncomment the
4174 directives and define how usernames and passwords are checked, but
4175 there is now a framework to start from.
4177 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4178 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4179 without this. I don't know how relevant this is to other LDAP libraries.
4181 PH/02 Add the verb name to the "unknown ACL verb" error.
4183 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4185 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4187 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4188 directory not expanded when it should be if an expanded home directory
4189 was set for the address (which is overridden by the transport).
4191 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4194 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4195 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4196 because it is too late at that time, and has no effect.
4198 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4199 security issue with \' (bugzilla #107). I could not use the
4200 PQescapeStringConn() function, because it needs a PGconn value as one of
4203 PH/08 When testing addresses using -bt, indicate those final addresses that
4204 are duplicates that would not cause an additional delivery. At least one
4205 person was confused, thinking that -bt output corresponded to deliveries.
4206 (Suppressing duplicates isn't a good idea as you lose the information
4207 about possibly different redirections that led to the duplicates.)
4209 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4210 systems where poll() doesn't work, in particular OS X.
4212 PH/10 Added more information to debugging output for retry time not reached.
4214 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4215 operations in malware.c.
4217 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4220 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4221 syslog_duplication set false, log lines that would normally be written
4222 both the the main log and to the reject log were not written to syslog at
4225 PH/14 In the default configuration, change the use of "message" in ACL warn
4226 statements to "add_header".
4228 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4229 not followed by a command (e.g. "seen endif").
4231 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4232 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4235 PH/17 Added extra conditions to the default value of delay_warning_condition
4239 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4240 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4241 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4244 The Auto-Submitted: and various List- headers are standardised, whereas I
4245 don't think Precedence: ever was.
4247 PH/18 Refactored debugging code in route_finduser() to show more information,
4248 in particular, the error code if getpwnam() issues one.
4250 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4251 This is apparently needed in addition to the PH/07 change above to avoid
4252 any possible encoding problems.
4254 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4255 but not after initializing Perl.
4257 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4258 output them only if debugging. By default they are written stderr,
4259 apparently, which is not desirable.
4261 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4264 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4267 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4269 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4270 authenticated or an ident call has been made. Suppress the default
4271 values for $authenticated_id and $authenticated_sender (but permit -oMai
4272 and -oMas) when testing with -bh.
4274 PH/24 Re-jigged the order of the tests in the default configuration so that the
4275 tests for valid domains and recipients precede the DNS black list and CSA
4276 tests, on the grounds that those ones are more expensive.
4278 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4279 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4280 This bug exists in every version of Exim that I still have, right back to
4283 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4284 However, an attempt to turn on case-sensitivity in a regex key by
4285 including (?-i) didn't work because the subject string was already
4286 lowercased, and the effects were non-intuitive. It turns out that a
4287 one-line patch can be used to allow (?-i) to work as expected.
4293 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4294 other effects) broke the use of negated acl sub-conditions.
4296 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4299 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4300 "Deny" causes Exim to reject the incoming connection with a 554 error.
4301 Unfortunately, if there is a major crisis, such as a disk failure,
4302 tcp-wrappers gives "deny", whereas what one would like would be some
4303 kind of temporary error. A kludge has been added to help with this.
4304 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4305 554 error is used if errno is still zero or contains ENOENT (which occurs
4306 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4309 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4311 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4312 errors. Otherwise a message that provokes a temporary error (when other
4313 messages do not) can cause a whole host to time out.
4315 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4316 addresses were routed directly to files or pipes from a redirect router.
4317 File deliveries just didn't batch; pipe deliveries might have suffered
4320 PH/06 A failure to get a lock for a hints database would erroneously always say
4321 "Failed to get write lock", even when it was really a read lock.
4323 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4324 of 0600. This has been changed to use the value of the lockfile_mode
4325 option (which defaults to 0600).
4327 PH/08 Applied small patch from the Sieve maintainer.
4329 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4330 folder from quota calculations, a direct delivery into this folder messed
4331 up the contents of the maildirsize file. This was because the regex was
4332 used only to exclude .Trash (or whatever) when the size of the mailbox
4333 was calculated. There was no check that a delivery was happening into an
4334 excluded directory. This bug has been fixed by ignoring all quota
4335 processing for deliveries into excluded directories.
4337 PH/10 Added the maildirfolder_create_regex option to appendfile.
4343 PH/01 The code for finding all the local interface addresses on a FreeBSD
4344 system running IPv6 was broken. This may well have applied to all BSD
4345 systems, as well as to others that have similar system calls. The broken
4346 code found IPv4 interfaces correctly, but gave incorrect values for the
4347 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4348 that it would not match correctly against @[] and not recognize the IPv6
4351 PH/02 The ipliteral router was not recognizing addresses of the form user@
4352 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4354 PH/03 Added disable_ipv6.
4356 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4357 decoded form, because it is most often used to construct To: headers
4358 lines in autoreplies, and the decoded form may well be syntactically
4359 invalid. However, $reply_address has leading white space removed, and all
4360 newlines turned into spaces so that the autoreply transport does not
4363 PH/05 If group was specified without a user on a router, and no group or user
4364 was specified on a transport, the group from the router was ignored.
4366 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4367 for visible compile-time settings that can be used to change these
4368 numbers, for those that want even more. Backwards compatibility with old
4369 spool files has been maintained. However, going back to a previous Exim
4370 release will lost any variables that are in spool files.
4372 PH/07 Two small changes when running in the test harness: increase delay when
4373 passing a TCP/IP connection to a new process, in case the original
4374 process has to generate a bounce, and remove special handling of
4375 127.0.0.2 (sic), which is no longer necessary.
4377 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4378 be the same on different OS.
4380 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4383 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4384 whether --show-vars was specified or not
4386 JJ/02 exipick: Added support for new ACL variable spool format introduced
4389 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4390 syntactically invalid From: or Reply-to: line, and a filter used this to
4391 generate an autoreply, and therefore failed to obtain an address for the
4392 autoreply, Exim could try to deliver to a non-existent relative file
4393 name, causing unrelated and misleading errors. What now happens is that
4394 it logs this as a hard delivery error, but does not attempt to create a
4397 PH/11 The exinext utility has a -C option for testing purposes, but although
4398 the given file was scanned by exinext itself; it wasn't being passed on
4399 when Exim was called.
4401 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4402 an end-of-file indication when reading a command response.
4404 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4405 compiled. In many other places in Exim, IPv6 addresses are always
4406 recognized, so I have changed this. It also means that IPv4 domain
4407 literals of the form [IPV4:n.n.n.n] are now always recognized.
4409 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4410 used if the router is not running as root, for example, when verifying at
4411 ACL time, or when using -bh. The debugging output from this situation was
4412 non-existent - all you got was a failure to exec. I have made two
4415 (a) Failures to set uid/gid, the current directory, or a process leader
4416 in a subprocess such as that created by queryprogram now generate
4417 suitable debugging output when -d is set.
4419 (b) The queryprogram router detects when it is not running as root,
4420 outputs suitable debugging information if -d is set, and then runs
4421 the subprocess without attempting to change uid/gid.
4423 PH/15 Minor change to Makefile for building test_host (undocumented testing
4426 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4427 additional section of a DNS packet that returns MX or SRV records.
4428 Instead, it always explicitly searches for A/AAAA records. This avoids
4429 major problems that occur when a DNS server includes only records of one
4430 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4431 fixed another bug: if SRV records were looked up and the corresponding
4432 address records were *not* found in the additional section, the port
4433 values from the SRV records were lost.
4435 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4436 using the correct key (the original address) when searching the retry
4437 rules in order to find which one to use for generating the retry hint.
4439 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4440 adding the default one. Similarly, if it contains a Reply-To: header, the
4441 errors_reply_to option, if set, is not used.
4443 PH/19 When calculating a retry time, Exim used to measure the "time since
4444 failure" by looking at the "first failed" field in the retry record. Now
4445 it does not use this if it is later than than the arrival time of the
4446 message. Instead it uses the arrival time. This makes for better
4447 behaviour in cases where some deliveries succeed, thus re-setting the
4448 "first failed" field. An example is a quota failure for a huge message
4449 when small messages continue to be delivered. Without this change, the
4450 "time since failure" will always be short, possible causing more frequent
4451 delivery attempts for the huge message than are intended.
4452 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4454 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4455 $1, $2, $3) because the numerical variables can be reset during some
4456 expansion items (e.g. "match"), thereby losing the authentication data.
4458 PH/21 Make -bV show the size of off_t variables so that the test suite can
4459 decide whether to run tests for quotas > 2G.
4461 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4462 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4463 filecount value is greater than 2G or if a quota value is greater than 2G
4464 on a system where the size of off_t is not greater than 4, a panic error
4467 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4468 never match. The debug and -bh output now contains an explicit error
4469 message indicating a malformed IPv4 address or mask.
4471 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4472 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4473 PH/23 above applies.
4475 PH/25 Do not write to syslog when running in the test harness. The only
4476 occasion when this arises is a failure to open the main or panic logs
4477 (for which there is an explicit test).
4479 PH/26 Added the /no_tell option to "control=freeze".
4481 PH/27 If a host name lookup failed very early in a connection, for example, if
4482 the IP address matched host_lookup and the reverse lookup yielded a name
4483 that did not have a forward lookup, an error message of the form "no IP
4484 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4485 could be logged. Now it outputs the IP address instead of "NULL".
4487 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4488 allows the sender and the authenticated sender to be set when
4489 submitting a message from within Exim. Since child_open_exim() is
4490 documented for local_scan(), the new function should be too.
4492 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4493 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4494 results in an empty string is now treated as unset.
4496 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4498 PH/31 Added sender_verify_fail logging option.
4500 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4501 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4502 tidied the source and removed it altogether.
4504 PH/33 When a queue run was abandoned because the load average was too high, a
4505 log line was always written; now it is written only if the queue_run log
4506 selector is set. In addition, the log line for abandonment now contains
4507 information about the queue run such as the pid. This is always present
4508 in "start" and "stop" lines but was omitted from the "abandon" line.
4510 PH/34 Omit spaces between a header name and the colon in the error message that
4511 is given when verify = headers_syntax fails (if there are lots of them,
4512 the message gets confusing).
4514 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4515 names, as there are now some PTR records that contain slashes. This check
4516 is only to protect against broken name servers that fall over on strange
4517 characters, so the fact that it applies to all lookups doesn't matter.
4519 PH/36 Now that the new test suite is complete, we can remove some of the
4520 special code in Exim that was needed for the old test suite. For example,
4521 sorting DNS records because real resolvers return them in an arbitrary
4522 order. The new test suite's fake resolver always returns records in the
4525 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4526 bounces) except when queue_only is set, to avoid logging races between
4527 the different processes.
4529 PH/38 Panic-die if .include specifies a non-absolute path.
4531 PH/39 A tweak to the "H" retry rule from its user.
4533 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4534 a label. They prevented compilation on older perls.
4536 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4537 a warning to be raised on newish perls.
4539 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4540 on queue. Changes to match documented behaviour of showing count of
4541 messages matching specified criteria.
4543 PH/40 Changed the default ident timeout from 30s to 5s.
4545 PH/41 Added support for the use of login_cap features, on those BSD systems
4546 that have them, for controlling the resources used by pipe deliveries.
4548 PH/42 The content-scanning code uses fopen() to create files in which to put
4549 message data. Previously it was not paying any attention to the mode of
4550 the files. Exim runs with umask(0) because the rest of the code creates
4551 files with open(), and sets the required mode explicitly. Thus, these
4552 files were ending up world-writeable. This was not a big issue, because,
4553 being within the spool directory, they were not world-accessible. I have
4554 created a function called modefopen, which takes an additional mode
4555 argument. It sets umask(777), creates the file, chmods it to the required
4556 mode, then resets the umask. All the relevant calls to fopen() in the
4557 content scanning code have been changed to use this function.
4559 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4560 to 24 hours. This avoids potential overflow problems when processing G
4561 and H retry rules. I suspect nobody ever tinkers with this value.
4563 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4565 PH/45 When the plaintext authenticator is running as a client, the server's
4566 challenges are checked to ensure they are valid base64 strings. By
4567 default, the authentication attempt is cancelled if an invalid string is
4568 received. Setting client_ignore_invalid_base64 true ignores these errors.
4569 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4570 they are received. Thus, the responses can be made to depend on the
4571 challenges. If an invalid string is ignored, an empty string is placed in
4574 PH/46 Messages that are created by the autoreply transport now contains a
4575 References: header, in accordance with RFCs 2822 and 3834.
4577 PH/47 Added authenticated_sender_force to the smtp transport.
4579 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4581 PH/49 Installed latest patch from the Sieve maintainer.
4583 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4584 was also set, the appendfile transport was unnecessarily scanning a
4585 directory of message files (e.g. for maildir delivery) to find the count
4586 of files (along with the size), even though it did not need this
4587 information. It now does the scan only if it needs to find either the
4588 size of the count of files.
4590 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4592 PH/52 Two bugs concerned with error handling when the smtp transport is
4595 (i) Exim was not creating retry information for temporary errors given
4596 for individual recipients after the DATA command when the smtp transport
4597 was used in LMTP mode. This meant that they could be retried too
4598 frequently, and not timed out correctly.
4600 (ii) Exim was setting the flag that allows error details to be returned
4601 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4602 recipients that were returned after the DATA command.
4604 PH/53 This is related to PH/52, but is more general: for any failing address,
4605 when detailed error information was permitted to be returned to the
4606 sender, but the error was temporary, then after the final timeout, only
4607 "retry timeout exceeded" was returned. Now it returns the full error as
4608 well as "retry timeout exceeded".
4610 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4611 do this, and (what is worse) MTAs that accept it.
4613 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4614 will now be deprecated.
4616 PH/56 New os.c-cygwin from the Cygwin maintainer.
4618 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4619 formats (previously only available in exim formats via -bpr, -bpru,
4620 and -bpra. Now also available in native and exiqgrep formats)
4622 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4623 with very large, slow to parse queues
4625 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4627 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4629 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4630 responses to authentication challenges, though it was showing the
4631 challenges; (ii) I've removed the CR characters from the debug output for
4634 PH/58 Allow for the insertion of a newline as well as a space when a string
4635 is turned into more than one encoded-word during RFC 2047 encoding. The
4636 Sieve code now uses this.
4638 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4639 data_4xx, lost_connection, tls_required.
4641 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4642 message was being sent as an SMTP response.
4644 PH/61 Add -l and -k options to exicyclog.
4646 PH/62 When verifying, if an address was redirected to one new address, so that
4647 verification continued, and the new address failed or deferred after
4648 having set something in $address_data, the value of $address_data was not
4649 passed back to the ACL. This was different to the case when no
4650 redirection occurred. The value is now passed back in both cases.
4652 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4653 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4654 use login_cap.h, so on its own it isn't the distinguishing feature. The
4655 new name refers directly to the setclassresources() function.
4657 PH/65 Added configuration files for NetBSD3.
4659 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4661 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4662 is preferred over IPv4.
4664 PH/68 The bounce_return_message and bounce_return_body options were not being
4665 honoured for bounces generated during the reception of non-SMTP messages.
4666 In particular, this applied to messages rejected by the ACL. This bug has
4667 been fixed. However, if bounce_return_message is true and bounce_return_
4668 body is false, the headers that are returned for a non-SMTP message
4669 include only those that have been read before the error was detected.
4670 (In the case of an ACL rejection, they have all been read.)
4672 PH/69 The HTML version of the specification is now built in a directory called
4673 spec_html instead of spec.html, because the latter looks like a path with
4674 a MIME-type, and this confuses some software.
4676 PH/70 Catch two compiler warnings in sieve.c.
4678 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4679 function verify_get_ident() calls ip_connect() to connect a socket, but
4680 if the "connect()" function timed out, ip_connect() used to close the
4681 socket. However, verify_get_ident() also closes the socket later, and in
4682 between Exim writes to the log, which may get opened at this point. When
4683 the socket was closed in ip_connect(), the log could get the same file
4684 descriptor number as the socket. This naturally causes chaos. The fix is
4685 not to close the socket in ip_connect(); the socket should be closed by
4686 the function that creates it. There was only one place in the code where
4687 this was missing, in the iplookup router, which I don't think anybody now
4688 uses, but I've fixed it anyway.
4690 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4691 well as to direct DNS lookups. Otherwise the handling of names in host
4692 lists is inconsistent and therefore confusing.
4698 PH/01 Two changes to the default runtime configuration:
4700 (1) Move the checks for relay_from_hosts and authenticated clients from
4701 after to before the (commented out) DNS black list checks.
4703 (2) Add control=submission to the relay_from_hosts and authenticated
4704 clients checks, on the grounds that messages accepted by these
4705 statements are most likely to be submissions.
4707 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4709 (1) Generate an error if the third argument for the ${prvs expansion is
4712 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4715 (3) Reset the variables that are obtained from the first argument of
4716 ${prvscheck and used in the second argument before leaving the code,
4717 because their memory is reclaimed, so using them afterwards may do
4720 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4721 one (it's much easier than Tom thought :-).
4723 (5) Because of (4), we can now allow for the use of $prvscheck_result
4724 inside the third argument.
4726 PH/03 For some reason, the default setting of PATH when running a command from
4727 a pipe transport was just "/usr/bin". I have changed it to
4730 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4731 anything to be listed in the output from -bV.
4733 PH/05 When a filter generated an autoreply, the entire To: header line was
4734 quoted in the delivery log line, like this:
4736 => >A.N.Other <ano@some.domain> <original@ddress> ...
4738 This has been changed so that it extracts the operative address. There
4739 may be more than one such address. If so, they are comma-separated, like
4742 => >ano@some.domain,ona@other.domain <original@ddress> ...
4744 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4745 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4746 not being looked up in the rDNS to get a host name, Exim was showing the
4747 IP address twice in Received: lines, even though the IP addresses were
4748 identical. For example:
4750 Received: from [1.2.3.4] (helo=[1.2.3.4])
4752 However, if the real host name was known, it was omitting the HELO data
4753 if it matched the actual IP address. This has been tidied up so that it
4754 doesn't show the same IP address twice.
4756 PH/07 When both +timestamp and +memory debugging was on, the value given by
4757 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4758 called by the debug printing, thereby overwriting the timestamp buffer.
4759 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4761 PH/08 When the original message was included in an autoreply transport, it
4762 always said "this is a copy of the message, including all the headers",
4763 even if body_only or headers_only was set. It now gives an appropriate
4766 PH/09 Applied a patch from the Sieve maintainer which:
4768 o fixes some comments
4769 o adds the (disabled) notify extension core
4770 o adds some debug output for the result of if/elsif tests
4771 o points to the current vacation draft in the documentation
4772 and documents the missing references header update
4776 o fixes a bug in processing the envelope test (when testing
4777 multiple envelope elements, the last element determined the
4780 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4781 Electronic Mail") by including:
4783 Auto-submitted: auto-generated
4785 in the messages that it generates (bounce messages and others, such as
4786 warnings). In the case of bounce messages for non-SMTP messages, there was
4787 also a typo: it was using "Auto_submitted" (underscore instead of
4788 hyphen). Since every message generated by Exim is necessarily in response
4789 to another message, thes have all been changed to:
4791 Auto-Submitted: auto-replied
4793 in accordance with these statements in the RFC:
4795 The auto-replied keyword:
4797 - SHOULD be used on messages sent in direct response to another
4798 message by an automatic process,
4800 - MUST NOT be used on manually-generated messages,
4802 - MAY be used on Delivery Status Notifications (DSNs) and Message
4803 Disposition Notifications (MDNs),
4805 - MUST NOT be used on messages generated by automatic or periodic
4806 processes, except for messages which are automatic responses to
4809 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4810 to the default Received: header definition.
4812 PH/12 Added log selector acl_warn_skipped (default on).
4814 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4815 variables because (a) they are in the wrong storage pool and (b) even if
4816 they were copied, it wouldn't work properly because of the caching.
4818 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4819 checking when decoding. Apparently there are clients that generate
4820 overlong encoded strings. Why am I not surprised?
4822 PH/15 If the first argument of "${if match_address" was not empty, but did not
4823 contain an "@" character, Exim crashed. Now it writes a panic log message
4824 and treats the condition as false.
4826 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4828 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4829 extension "envelope-auth". The code is finished and in agreement with
4830 other implementations, but there is no documentation so far and in fact,
4831 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4832 not changing the active code.
4834 Print executed "if" and "elsif" statements when debugging is used. This
4835 helps a great deal to understand what a filter does.
4837 Document more things not specified clearly in RFC3028. I had all this
4838 sorted out, when out of a sudden new issues came to my mind. Oops."
4840 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4843 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4844 canonical form (as documented). However, after a host name lookup from
4845 the IP address, check_host() was doing a simple string comparison with
4846 addresses acquired from the DNS when checking that the found name did
4847 have the original IP as one of its addresses. Since any found IPv6
4848 addresses are likely to be in abbreviated form, the comparison could
4849 fail. Luckily, there already exists a function for doing the comparison
4850 by converting both addresses to binary, so now that is used instead of
4851 the text comparison.
4853 PH/20 There was another similar case to PH/19, when a complete host name was
4854 given in a host list; looking up its IP address could give an abbreviated
4855 form, whereas the current host's name might or might not be abbreviated.
4856 The same fix has been applied.
4862 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4863 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4866 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4867 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4869 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4871 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4872 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4873 Also, for "Auto-Submitted", treat anything other than "no" as
4874 non-personal, in accordance with RFC 3834. (Previously it treated
4875 anything starting "auto-" as non-personal.)
4877 TF/01 The control=submission/name=... option had a problem with syntax
4878 errors if the name included a slash character. The /name= option
4879 now slurps the rest of the string, so it can include any characters
4880 but it must come last in the list of options (after /sender_retain
4883 PH/05 Some modifications to the interface to the fake nameserver for the new
4891 TK/01 Added the "success_on_redirect" address verification option. See
4892 NewStuff for rationale and an example.
4894 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4896 PH/02 Patch to exigrep to allow it to work on syslog lines.
4898 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4899 fread() to skip over the body file's header line, because in Cygwin the
4900 header line is locked and is inaccessible.
4902 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4903 co-exist for some time) to make it clear that it is the Exim ID that is
4904 referenced, not the Message-ID: header line.
4906 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4907 string_format() function, because snprintf() does not exist on all
4910 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4911 ${stat: expansion item.
4913 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4914 protocol synchronization error", to keep the pedants happy.
4916 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4917 well as for IRIX systems, when gcc is being used. See the host.c source
4920 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4922 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4925 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4926 transport and to the smtp transport in LMTP mode.
4928 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4930 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4931 run a filter in a subprocess. This could lead to confusion in subsequent
4932 lookups in the parent process. There should also be a search_tidyup() at
4933 the end of the subprocess.
4935 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4936 only if the host matched helo_try_verify_hosts, which caused the
4937 verification to occur when the EHLO/HELO command was issued. The ACL just
4938 tested the remembered result. Now, if a previous verification attempt has
4939 not happened, "verify = helo" does it there and then.
4941 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4943 TK/03 Fix log output including CR from clamd.
4945 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4946 exist provoked a memory error which could cause a segfault.
4948 PH/15 Installed PCRE 6.2
4950 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4952 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4953 of the problem. Specifically, suggested +O2 rather than +O1 for the
4956 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4958 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4959 record, the port was not correctly logged when the outgoing_port log
4960 selector was set (it logged the transort's default port).
4962 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4963 fallback_hosts, and "hosts" in the smtp transport.
4965 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4966 host errors such as "Connection refused".
4968 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4969 authentication with radiusclient 0.4.9:
4971 - Error returned from rc_read_config was caught wrongly
4972 - Username/password not passed on to radius server due to wrong length.
4974 The presumption is that some radiusclient API changes for 4.51/PH/17
4975 were not taken care of correctly. The code is still untested by me (my
4976 Linux distribution still has 0.3.2 of radiusclient), but it was
4977 contributed by a Radius user.
4979 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4980 expanding the "port" option of the smtp transport.
4982 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4983 while reading a MIME header. Thanks to Tom Hughes for a patch.
4985 PH/24 Include config.h inside local_scan.h so that configuration settings are
4988 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4989 that in an ACL for RCPT (for example), you can examine exactly what was
4992 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4993 commands, but it was not correctly comparing the address with the actual
4994 client host address. Thus, it would show the EHLO address in Received:
4995 header lines when this was not necessary.
4997 PH/27 Added the % operator to ${eval:}.
4999 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5000 it should be ignoring failures (because with -C, for example, it has lost
5001 privilege). It wasn't ignoring creation failures other than "already
5004 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5007 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5008 because an input line was too long, either on its own, or by virtue of
5009 too many continuations, the temporary file was not being removed, and the
5010 return code was incorrect.
5012 PH/31 Missing "BOOL" in function definition in filtertest.c.
5014 PH/32 Applied Sieve patches from the maintainer.
5016 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5018 PH/33 Added "verify = not_blind".
5020 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5021 Local/Makefile (with some defaults set). These are used in built scripts
5022 such as exicyclog, but they have never been used in the exim_install
5023 script (though there are many overriding facilities there). I have
5024 arranged that the exim_install script now takes note of these two
5027 PH/35 Installed configuration files for Dragonfly.
5029 PH/36 When a locally submitted message by a trusted user did not contain a
5030 From: header, and the sender address was obtained from -f or from an SMTP
5031 MAIL command, and the trusted user did not use -F to supply a sender
5032 name, $originator_name was incorrectly used when constructing a From:
5033 header. Furthermore, $originator_name was used for submission mode
5034 messages from external hosts without From: headers in a similar way,
5035 which is clearly wrong.
5037 PH/37 Added control=suppress_local_fixups.
5039 PH/38 When log_selector = +received_sender was set, and the addition of the
5040 sender made the log line's construction buffer exactly full, or one byte
5041 less than full, an overflow happened when the terminating "\n" was
5044 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5045 when the result of a list match is failure because a DNS lookup failed.
5047 PH/40 RM_COMMAND is now used in the building process.
5049 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5050 the "build-* directories that it finds.
5052 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5053 address in a domain literal was a prefix of an interface address.
5055 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5056 when verifying a sender address, unless rewrite_headers is false.
5058 PH/44 Wrote a long comment about why errors_to addresses are verified as
5059 recipients, not senders.
5061 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5062 the ratelimit ACL was added.
5064 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5066 PH/46 Added extra information about PostgreSQL errors to the error string.
5068 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5069 avoiding the need to install special zones in a real server. This is
5070 backwards compatible; if it can't find the fake resolver, it drops back.
5071 Thus, both old and new test suites can be run.
5073 TF/02 Added util/ratelimit.pl
5075 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5076 clock is set back in time.
5078 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5079 Candler <B.Candler@pobox.com>.
5081 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5082 OK for addresses that are the result of redirections.
5084 PH/48 A number of further additions for the benefit of the new test suite,
5085 including a fake gethostbyname() that interfaces to the fake DNS resolver
5088 TF/06 The fix for widen_domains has also been applied to qualify_single and
5089 search_parents which are the other dnslookup options that can cause
5092 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5095 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5097 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5098 service" if it encountered an MX record with an empty target hostname.
5099 The message is now "an MX or SRV record indicated no SMTP service".
5101 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5102 if the DNS of the sending site is misconfigured. This is quite a
5103 common situation. This change restores the behaviour of treating a
5104 helo verification defer as a failure.
5106 PH/51 If self=fail was set on a router, the bounce message did not include the
5107 actual error message.
5113 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5115 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5116 command itself ended in error, the underlying message about the transport
5117 filter timeout was being overwritten with the pipe command error. Now the
5118 underlying error message should be appended to the second error message.
5120 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5122 PH/02 Reduce the amount of output that "make" produces by default. Full output
5123 can still be requested.
5125 PH/03 The warning log line about a condition test deferring for a "warn" verb
5126 was being output only once per connection, rather than after each
5127 occurrence (because it was using the same function as for successful
5128 "warn" verbs). This seems wrong, so I have changed it.
5130 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5131 it should not have, which might have caused a crash in the right
5132 circumstances, but probably never did.
5134 PH/04 Installed a modified version of Tony Finch's patch to make submission
5135 mode fix the return path as well as the Sender: header line, and to
5136 add a /name= option so that you can make the user's friendly name appear
5139 TF/03 Added the control = fakedefer ACL modifier.
5141 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5142 Mark Lowes for thorough testing.
5144 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5146 TK/03 Merged latest SRS patch from Miles Wilton.
5148 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5149 with the definition in sysexits.h (which is #included earlier).
5150 Fortunately, Exim does not actually use EX_OK. The code used to try to
5151 preserve the sysexits.h value, by assuming that macro definitions were
5152 scanned for macro replacements. I have been disabused of this notion,
5153 so now the code just undefines EX_OK before #including unistd.h.
5155 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5156 in the smtp transport. When a block could not be written in a single
5157 write() function, the timeout was being re-applied to each part-write.
5158 This seems wrong - if the receiver was accepting one byte at a time it
5159 would take for ever. The timeout is now adjusted when this happens. It
5160 doesn't have to be particularly precise.
5162 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5163 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5165 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5166 without a domain if the check to <postmaster@domain> fails.
5168 SC/01 Eximstats: added -xls and the ability to specify output files
5169 (patch written by Frank Heydlauf).
5171 SC/02 Eximstats: use FileHandles for outputting results.
5173 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5175 SC/04 Eximstats: fixed display of large numbers with -nvr option
5177 SC/05 Eximstats: fixed merging of reports with empty tables.
5179 SC/06 Eximstats: added the -include_original_destination flag
5181 SC/07 Eximstats: removed tabs and trailing whitespace.
5183 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5185 TK/06 MBOX spool code: Add real "From " MBOX separator line
5186 so the .eml file is really in mbox format (even though
5187 most programs do not really care). Patch from Alex Miller.
5189 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5190 The latter is generated from $received_to and is only set if the
5191 message has one envelope recipient. SA can use these headers,
5192 obviously out-of-the-box. Patch from Alex Miller.
5194 PH/08 The ${def test on a variable was returning false if the variable's
5195 value was "0", contrary to what the specification has always said!
5196 The result should be true unless the variable is empty.
5198 PH/09 The syntax error of a character other than { following "${if
5199 def:variable_name" (after optional whitespace) was not being diagnosed.
5200 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5201 accidental colon was present, for example, could give incorrect results.
5203 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5204 result is used (not including appendfile, where other changes are about
5207 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5208 This involved changing a lot of size variables from int to off_t. It
5209 should work with maildirs and everything.
5211 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5212 spamd dying while we are connected to it.
5214 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5217 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5218 function for BDB 4.3.
5220 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5222 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5223 the sources for PCRE have been re-arranged and more files are now
5226 PH/15 The code I had for printing potentially long long variables in PH/11
5227 above was not the best (it lost precision). The length of off_t variables
5228 is now inspected at build time, and an appropriate printing format (%ld
5229 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5230 to be "long long int" or "long int". This is needed for the internal
5231 formatting function string_vformat().
5233 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5234 the configuration file to be ":syslog", then the script "guesses" where
5235 the logs files are, rather than using the compiled in default. In our
5236 case the guess is not the same as the compiled default, so the script
5237 suddenly stopped working when I started to use syslog. The patch checks
5238 to see if log_file_path is "". If so, it attempts to read it from exim
5239 with no configuration file to get the compiled in version, before it
5240 falls back to the previous guessing code."
5242 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5243 implementing BATV in an Exim configuration. See NewStuff for the gory
5246 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5247 Makefile that are specific to HP-UX.
5249 PH/18 If the "use_postmaster" option was set for a recipient callout together
5250 with the "random" option, the postmaster address was used as the MAIL
5251 FROM address for the random test, but not for the subsequent recipient
5252 test. It is now used for both.
5254 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5255 patch removes a few documentation additions to RFC 3028, because the
5256 latest draft now contains them. It adds the new en;ascii-case comparator
5257 and a new error check for 8bit text in MIME parts. Comparator and
5258 require names are now matched exactly. I enabled the subaddress
5259 extension, but it is not well tested yet (read: it works for me)."
5261 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5262 rework some of the code of TK/09 above to avoid the hardwired use of
5263 "%lld" and "long long". Replaced the call to snprintf() with a call to
5266 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5267 records point to non-existent hosts", "retry timeout exceeded", and
5268 "retry time not reached for any host after a long failure period".
5270 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5271 experimental DomainKeys support:
5273 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5274 (2) On an error such as an illegally used "control", the wrong name for
5275 the control was given.
5277 These problems did NOT occur unless DomainKeys support was compiled.
5279 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5281 PH/24 Added ${if match_ip condition.
5283 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5284 databases so that it will be absolutely obvious if a crash occurs in the
5285 DB library. This is a regular occurrence (often caused by mis-matched
5288 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5289 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5290 picked up on a user's system that detects such things. There doesn't seem
5291 to be a gcc warning option for this - only an attribute that has to be
5292 put on the function's prototype. It seems that in Fedora Core 4 they have
5293 set this on a number of new functions. No doubt there will be more in due
5296 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5297 specify a transport. However, if an address that was verified by such a
5298 router was the subject of a callout, Exim crashed because it tried to
5299 read the rcpt_include_affixes from the non-existent transport. Now it
5300 just assumes that the setting of that option is false. This bug was
5301 introduced by 4.51/PH/31.
5303 PH/28 Changed -d+all to exclude +memory, because that information is very
5304 rarely of interest, but it makes the output a lot bigger. People tend to
5305 do -d+all out of habit.
5307 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5308 code in os-type was giving problems when libc.so lives in lib64, like on
5311 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5312 aren't the modern standard, and it seems that some systems' include files
5313 don't always have them. Exim was already checking for some of the newer
5314 ones like T_AAAA, and defining it itself. I've added checks for all the
5315 record types that Exim uses.
5317 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5318 not automatically generating a new one, as it is supposed to. This
5319 prevented TLS from working. If the file did exist, but contained invalid
5320 data, a new version was generated, as expected. It was only the case of a
5321 non-existent file that was broken.
5323 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5324 with a change in libdomainkeys > 0.64.
5326 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5327 from DNS. If the selector record carries the flag, it now has
5328 precedence over the domain-wide flag.
5330 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5332 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5333 the use of a port name that isn't defined in /etc/services) occurred, the
5334 message was deferred as in a normal delivery, and thus remained on the
5335 spool, instead of being failed because of the mua_wrapper setting. This
5336 is now fixed, and I tidied up some of the mua_wrapper messages at the
5339 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5340 the messages in a hash of arrays rather than using individual hashes.
5341 This is a bit cleaner and results in dramatic memory savings, albeit
5342 at a slight CPU cost.
5344 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5345 as requested by Marc Sherman.
5347 SC/10 Eximstats: added histograms for user specified patterns as requested
5350 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5352 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5353 fopen() in the content-scanning modules that did not already have it.
5359 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5360 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5362 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5364 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5366 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5367 to test Sieve filters that use "vacation".
5369 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5370 that changes the way the GnuTLS parameters are stored in the cache file.
5371 The new format can be generated externally. For backward compatibility,
5372 if the data in the cache doesn't make sense, Exim assumes it has read an
5373 old-format file, and it generates new data and writes a new file. This
5374 means that you can't go back to an older release without removing the
5377 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5378 work if there are any delivery delays because "one_time" forces the
5379 parent to be marked "delivered", so its unseen clone is never tried
5380 again. For this reason, Exim now forbids the simultaneous setting of
5383 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5384 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5385 entry if you want to know the details. The fix, however, neglected to
5386 consider the case when local delivery batching is involved. The test for
5387 "previously delivered" was not happening when checking to see if an
5388 address could be batched with a previous (undelivered) one; under
5389 certain circumstances this could lead to multiple deliveries to the same
5392 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5393 in its include files, and this causes problems building Exim.
5395 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5396 header_syntax) but Exim was just ignoring anything given after a slash.
5397 In particular, this caused confusion with an attempt to use "verify =
5398 reverse_host_lookup/defer_ok". An error is now given when options are
5399 supplied for verify items that do not have them. (Maybe reverse_host_
5400 lookup should have a defer_ok option, but that's a different point.)
5402 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5403 defined by RFC 821) to 2048, because there were problems with some AUTH
5404 commands, and RFC 1869 says the size should be increased for extended
5405 SMTP commands that take arguments.
5407 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5410 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5411 "unknown" error; now it says that the functionality isn't in the binary.
5413 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5414 an address' error message when a string expansion fails (syntax or
5415 whatever). Otherwise the password may appear in the log. Following change
5416 PH/42 below, there is no longer a chance of it appearing in a bounce
5419 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5421 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5422 of its addresses was ever tried. (Bugzilla bug #2.)
5424 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5425 the result incorrectly in the debug output. (It correctly added a newline
5426 to what was transported.)
5428 TF/01 Added $received_time.
5430 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5431 commented out examples of how to interface to a virus scanner and to
5432 SpamAssassin. Also added commented examples of av_scanner and
5433 spamd_address settings.
5435 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5436 and controls are allowed in which ACLs. There were a couple of minor
5437 errors. Some of the entries in the conditions table (which is a table of
5438 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5439 negation of where the condition IS allowed.
5441 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5443 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5444 header file does not have a version number, so I've had to invent a new
5445 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5446 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5447 radiusclient), but it was contributed by a Radius user.
5449 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5450 files or directories, for OpenSSL.
5452 PH/19 When an Exim process that is running as root has to create an Exim log
5453 file, it does so in a subprocess that runs as exim:exim so as to get the
5454 ownership right at creation (otherwise, other Exim processes might see
5455 the file with the wrong ownership). There was no test for failure of this
5456 fork() call, which would lead to the process getting stuck as it waited
5457 for a non-existent subprocess. Forks do occasionally fail when resources
5458 run out. I reviewed all the other calls to fork(); they all seem to check
5461 PH/20 When checking for unexpected SMTP input at connect time (before writing
5462 the banner), Exim was not dealing correctly with a non-positive return
5463 from the read() function. If the client had disconnected by this time,
5464 the result was a log entry for a synchronization error with an empty
5465 string after "input=" when read() returned zero. If read() returned -1
5466 (an event I could not check), uninitialized data bytes were printed.
5467 There were reports of junk text (parts of files, etc) appearing after
5470 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5472 PH/22 Added support for macro redefinition, and (re)definition in between
5473 driver and ACL definitions.
5475 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5476 forgetting to use the resulting value; it was using the unexpanded value.
5478 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5479 hadn't been configured. The fix is from Juergen Kreileder, who
5480 understands it better than I do:
5482 "Here's what I see happening with three configured cyrus_sasl
5483 authenticators configured (plain, login, cram-md5):
5485 On startup auth_cyrus_sasl_init() gets called for each of these.
5486 This means three calls to sasl_listmech() without a specified mech_list.
5487 => SASL tests which mechs of all available mechs actually work
5488 => three warnings about OTP not working
5489 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5491 With the patch, sasl_listmech() also gets called three times. But now
5492 SASL's mech_list option is set to the server_mech specified in the the
5493 authenticator. Or in other words, the answer from sasl_listmech()
5494 gets limited to just the mech you're testing for (which is different
5496 => the return list contains just 'plain' or 'login', 'cram-md5' or
5497 nothing depending on the value of ob->server_mech.
5499 I've just tested the patch: Authentication still works fine,
5500 unavailable mechs specified in the exim configuration are still
5501 caught, and the auth.log warnings about OTP are gone."
5503 PH/25 When debugging is enabled, the contents of the command line are added
5504 to the debugging output, even when log_selector=+arguments is not
5507 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5508 answer is "GNU", and only if the return is "GNU/something" is the answer
5511 PH/27 $acl_verify_message is now set immediately after the failure of a
5512 verification in an ACL, and so is available in subsequent modifiers. In
5513 particular, the message can be preserved by coding like this:
5515 warn !verify = sender
5516 set acl_m0 = $acl_verify_message
5518 Previously, $acl_verify_message was set only while expanding "message"
5519 and "log_message" when a very denied access.
5521 PH/28 Modified OS/os.c-Linux with
5523 -#ifndef OS_LOAD_AVERAGE
5524 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5526 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5527 nomenclature these days.)
5529 PH/29 Installed patch from the Sieve maintainer that adds the options
5530 sieve_useraddress and sieve_subaddress to the redirect router.
5532 PH/30 In these circumstances:
5533 . Two addresses routed to the same list of hosts;
5534 . First host does not offer TLS;
5535 . First host accepts first address;
5536 . First host gives temporary error to second address;
5537 . Second host offers TLS and a TLS session is established;
5538 . Second host accepts second address.
5539 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5540 and peerdn, if requested) that were in fact used only for the second
5543 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5544 attention to any local part prefix or suffix that was matched by the
5545 router that accepted the address. It now behaves in the same way as it
5546 does for delivery: the affixes are removed from the local part unless
5547 rcpt_include_affixes is set on the transport.
5549 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5550 timeout during the DATA phase of an incoming message.
5552 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5553 applied a patch sanctioned by the Sieve maintainer.
5555 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5556 the uid or gid is negative. A case of a negative gid caused this to be
5557 noticed. The fix allows for either to be negative.
5559 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5560 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5563 PH/36 The change PH/12 above was broken. Fixed it.
5565 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5566 the grounds that routing the same address twice would always produce the
5567 same answer. This might have been true once, but it is certainly no
5568 longer true now. Routing a child address may depend on the previous
5569 routing that produced that child. Some complicated redirection strategies
5570 went wrong when messages had multiple recipients, and made Exim's
5571 behaviour dependent on the order in which the addresses were given.
5573 I have moved the duplicate checking until after the routing is complete.
5574 Exim scans the addresses that are assigned to local and remote
5575 transports, and removes any duplicates. This means that more work will be
5576 done, as duplicates will always all be routed, but duplicates are
5577 presumably rare, so I don't expect this is of any significance.
5579 For deliveries to pipes, files, and autoreplies, the duplicate checking
5580 still happens during the routing process, since they are not going to be
5583 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5584 It corrects a timeout issue with spamd. This is Ian's comment: "The
5585 background is that sometimes spamd either never reads data from a
5586 connection it has accepted, or it never writes response data. The exiscan
5587 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5588 blindly assumes that writes won't block so it may never time out."
5590 PH/39 Allow G after quota size as well as K and M.
5592 PH/40 The value set for $authenticated_id in an authenticator may not contain
5593 binary zeroes or newlines because the value is written to log lines and
5594 to spool files. There was no check on this. Now the value is run through
5595 the string_printing() function so that such characters are converted to
5596 printable escape sequences.
5598 PH/41 $message_linecount is a new variable that contains the total number of
5599 lines in the message. Compare $body_linecount, which is the count for the
5602 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5603 bounce and delay warning messages, except in certain special cases, which
5606 (a) An SMTP error message from a remote host;
5607 (b) A message specified in a :fail: redirection;
5608 (c) A message specified in a "fail" command in a system filter;
5609 (d) A message specified in a FAIL return from the queryprogram router;
5610 (e) A message specified by the cannot_route_message router option.
5612 In these cases only, Exim does include the error details in bounce and
5613 warning messages. There are also a few cases where bland messages such
5614 as "unrouteable address" or "local delivery error" are given.
5616 PH/43 $value is now also set for the "else" part of a ${run expansion.
5618 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5619 being worked on, but at least Exim now implements the latest version to
5622 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5623 process to complete was treated as a delivery failure, a timeout while
5624 writing the message to the pipe was logged, but erroneously treated as a
5625 successful delivery. Such timeouts include transport filter timeouts. For
5626 consistency with the overall process timeout, these timeouts are now
5627 treated as errors, giving rise to delivery failures by default. However,
5628 there is now a new Boolean option for the pipe transport called
5629 timeout_defer, which, if set TRUE, converts the failures into defers for
5630 both kinds of timeout. A transport filter timeout is now identified in
5633 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5634 systems where "make" and "gmake" are different, calling "gmake" at top
5635 level broke things. I've arranged for the value of $(MAKE) to be passed
5636 from the Makefile to this script so that it can call the same version of
5640 A note about Exim versions 4.44 and 4.50
5641 ----------------------------------------
5643 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5644 changes of various kinds. As a consequence, a big documentation update was
5645 needed. This delayed the release for rather longer than seemed good, especially
5646 in the light of a couple of (minor) security issues. Therefore, the changes
5647 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5648 release. So 4.44 and 4.50 are in effect two different branches that both start
5651 I have left the 4.50 change log unchanged; it contains all the changes since
5652 4.43. The change log for 4.44 is below; many of its items are identical to
5653 those for 4.50. This seems to be the most sensible way to preserve the
5654 historical information.
5660 1. Minor wording change to the doc/README.SIEVE file.
5662 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5663 computation of the current number of files was incorrect.
5665 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5666 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5669 4. Give more explanation in the error message when the command for a transport
5670 filter fails to execute.
5672 5. There are several places where Exim runs a non-Exim command in a
5673 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5674 was being done only for the command run by the queryprogram router. It is
5675 now done for all such subprocesses. The other cases are: ${run, transport
5676 filters, and the commands run by the lmtp and pipe transports.
5678 6. Added CONFIGURE_GROUP build-time option.
5680 7. Some older OS have a limit of 256 on the maximum number of file
5681 descriptors. Exim was using setrlimit() to set 1000 as a large value
5682 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5683 systems. I've change it so that if it can't get 1000, it tries for 256.
5685 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5686 was an oversight, and furthermore, ever since the addition of extra
5687 controls (e.g. 4.43/32), the checks on when to allow different forms of
5688 "control" were broken. There should now be diagnostics for all cases when a
5689 control that does not make sense is encountered.
5691 9. Added the /retain_sender option to "control=submission".
5693 10. $recipients is now available in the predata ACL (oversight).
5695 11. Tidy the search cache before the fork to do a delivery from a message
5696 received from the command line. Otherwise the child will trigger a lookup
5697 failure and thereby defer the delivery if it tries to use (for example) a
5698 cached ldap connection that the parent has called unbind on.
5700 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5701 of $address_data from the recipient verification was clobbered by the
5702 sender verification.
5704 13. The value of address_data from a sender verification is now available in
5705 $sender_address_data in subsequent conditions in the ACL statement.
5707 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5709 15. Added a new option "connect=<time>" to callout options, to set a different
5712 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5713 was its contents. (It was OK if the option was not defined at all.)
5715 17. A "Completed" log line is now written for messages that are removed from
5716 the spool by the -Mrm option.
5718 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5719 information about exactly what failed.
5721 19. Added -dd to debug only the daemon process.
5723 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5724 handles timeouts, both on the server side and network timeouts. Renamed the
5725 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5727 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5728 It is now set to "smtps".
5730 22. $host_address is now set to the target address during the checking of
5731 ignore_target_hosts.
5733 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5734 being passed; this would have caused $sender_host_name to have been used if
5735 matching the list had actually called for a host name (not very likely,
5736 since this list is usually IP addresses). A host name is now passed as
5739 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5740 code that specifies a non-restarting handler (typically sigaction() in
5741 modern systems) in an attempt to fix a rare and obscure crash bug.
5743 25. Narrowed the window for a race in the daemon that could cause it to ignore
5744 SIGCHLD signals. This is not a major problem, because they are used only to
5745 wake it up if nothing else does.
5747 26. A malformed maildirsize file could cause Exim to calculate negative values
5748 for the mailbox size or file count. Odd effects could occur as a result.
5749 The maildirsize information is now recalculated if the size or filecount
5752 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5753 support for a long time. Removed HAVE_SYS_VFS_H.
5755 28. Installed the latest version of exipick from John Jetmore.
5757 29. In an address list, if the pattern was not a regular expression, an empty
5758 subject address (from a bounce message) matched only if the pattern was an
5759 empty string. Non-empty patterns were not even tested. This was the wrong
5760 because it is perfectly reasonable to use an empty address as part of a
5761 database query. An empty address is now tested by patterns that are
5762 lookups. However, all the other forms of pattern expect the subject to
5763 contain a local part and a domain, and therefore, for them, an empty
5764 address still always fails if the pattern is not itself empty.
5766 30. Exim went into a mad DNS loop when attempting to do a callout where the
5767 host was specified on an smtp transport, and looking it up yielded more
5768 than one IP address.
5770 31. Re-factored the code for checking spool and log partition space into a
5771 function that finds that data and another that does the check. The former
5772 is then used to implement four new variables: $spool_space, $log_space,
5773 $spool_inodes, and $log_inodes.
5775 32. The RFC2047 encoding function was originally intended for short strings
5776 such as real names; it was not keeping to the 75-character limit for
5777 encoded words that the RFC imposes. It now respects the limit, and
5778 generates multiple encoded words if necessary. To be on the safe side, I
5779 have increased the buffer size for the ${rfc2047: expansion operator from
5782 33. It is now permitted to omit both strings after an "if" condition; if the
5783 condition is true, the result is "true". As before, when the second string
5784 is omitted, a false condition yields an empty string. This makes it less
5785 cumbersome to write custom ACL and router conditions.
5787 34. Failure to deliver a bounce message always caused it to be frozen, even if
5788 there was an errors_to setting on the router. The errors_to setting is now
5791 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5792 canonical form (fully expanded) before being placed in
5793 $sender_host_address.
5795 36. The table in the code that translates DNS record types into text (T_A to
5796 "A" for instance) was missing entries for NS and CNAME. It is just possible
5797 that this could have caused confusion if both these types were looked up
5798 for the same domain, because the text type is used as part of Exim's
5799 per-process caching. But the chance of anyone hitting this buglet seems
5802 37. The dnsdb lookup has been extended in a number of ways.
5804 (1) There is a new type, "zns", which walks up the domain tree until it
5805 finds some nameserver records. It should be used with care.
5807 (2) There is a new type, "mxh", which is like "mx" except that it returns
5808 just the host names, not the priorities.
5810 (3) It is now possible to give a list of domains (or IP addresses) to be
5811 looked up. The behaviour when one of the lookups defers can be
5812 controlled by a keyword.
5814 (4) It is now possible to specify the separator character for use when
5815 multiple records are returned.
5817 38. The dnslists ACL condition has been extended: it is now possible to supply
5818 a list of IP addresses and/or domains to be looked up in a particular DNS
5821 39. Added log_selector=+queue_time_overall.
5823 40. When running the queue in the test harness, wait just a tad after forking a
5824 delivery process, to get repeatability of debugging output.
5826 41. Include certificate and key file names in error message when GnuTLS fails
5827 to set them up, because the GnuTLS error message doesn't include the name
5828 of the failing file when there is a problem reading it.
5830 42. Allow both -bf and -bF in the same test run.
5832 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5834 44. The "Exiscan patch" is now merged into the mainline Exim source.
5836 45. Sometimes the final signoff response after QUIT could fail to get
5837 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5838 < 0 before doing a fflush(). This bug looks as though it goes back to the
5839 introduction of TLS in release 3.20, but "sometimes" must have been rare
5840 because the tests only now provoked it.
5842 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5843 (this can affect the format of dates).
5845 47. exim_tidydb, when checking for the continued existence of a message for
5846 which it has found a message-specific retry record, was not finding
5847 messages that were in split spool directories. Consequently, it was
5848 deleting retry records that should have stayed in existence.
5850 48. Steve fixed some bugs in eximstats.
5852 49. The SPA authentication driver was not abandoning authentication and moving
5853 on to the next authenticator when an expansion was forced to fail,
5854 contradicting the general specification for all authenticators. Instead it
5855 was generating a temporary error. It now behaves as specified.
5857 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5858 (the order specifies the preference for clients). The order is now AES256,
5859 AES128, 3DES, ARCFOUR128.
5861 51. Small patch to Sieve code - explicitly set From: when generating an
5864 52. Exim crashed if a remote delivery caused a very long error message to be
5865 recorded - for instance if somebody sent an entire SpamAssassin report back
5866 as a large number of 550 error lines. This bug was coincidentally fixed by
5867 increasing the size of one of Exim's internal buffers (big_buffer) that
5868 happened as part of the Exiscan merge. However, to be on the safe side, I
5869 have made the code more robust (and fixed the comments that describe what
5872 53. Now that there can be additional text after "Completed" in log lines (if
5873 the queue_time_overall log selector is set), a one-byte patch to exigrep
5874 was needed to allow it to recognize "Completed" as not the last thing in
5877 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5878 patch that reportedly fixes this has been added. I am not expert enough to
5879 create a test for it. This is what the patch creator wrote:
5881 "I found a little strange behaviour of ldap code when working with
5882 Windows 2003 AD Domain, where users was placed in more than one
5883 Organization Units. When I tried to give exim partial DN, the exit code
5884 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5885 But simultaneously result of request was absolutely normal ldap result,
5886 so I produce this patch..."
5888 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5889 REFERENCE, so I have modified the code to exclude the patch when that macro
5892 55. Some experimental protocols are using DNS PTR records for new purposes. The
5893 keys for these records are domain names, not reversed IP addresses. The
5894 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5895 leaves it alone. Component reversal etc. now happens only for IP addresses.
5898 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5900 57. Double the size of the debug message buffer (to 2048) so that more of very
5901 long debug lines gets shown.
5903 58. The exicyclog utility now does better if the number of log files to keep
5904 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5906 59. Two changes related to the smtp_active_hostname option:
5908 (1) $smtp_active_hostname is now available as a variable.
5909 (2) The default for smtp_banner uses $smtp_active_hostname instead
5910 of $primary_hostname.
5912 60. The host_aton() function is supposed to be passed a string that is known
5913 to be a valid IP address. However, in the case of IPv6 addresses, it was
5914 not checking this. This is a hostage to fortune. Exim now panics and dies
5915 if the condition is not met. A case was found where this could be provoked
5916 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5917 components; fortuitously, this particular loophole had already been fixed
5918 by change 4.50/55 above.
5920 If there are any other similar loopholes, the new check in host_aton()
5921 itself should stop them being exploited. The report I received stated that
5922 data on the command line could provoke the exploit when Exim was running as
5923 exim, but did not say which command line option was involved. All I could
5924 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5925 running as the user.
5928 61. There was a buffer overflow vulnerability in the SPA authentication code
5929 (which came originally from the Samba project). I have added a test to the
5930 spa_base64_to_bits() function which I hope fixes it.
5933 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5934 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5936 63. The daemon start-up calls getloadavg() while still root for those OS that
5937 need the first call to be done as root, but it missed one case: when
5938 deliver_queue_load_max is set with deliver_drop_privilege. This is
5939 necessary for the benefit of the queue runner, because there is no re-exec
5940 when deliver_drop_privilege is set.
5942 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5943 This has been fixed.
5945 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5946 was in use, was not putting the data itself into the right store pool;
5947 consequently, it could be overwritten for a subsequent message in the same
5948 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5951 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5953 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5954 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5955 the function were treating the return as a boolean value, which happened to
5956 work because 0=false and not-0=true, but is not correct code.
5958 68. The host_aton() function was not handling scoped IPv6 addresses (those
5959 with, for example, "%eth0" on the end) correctly.
5961 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5962 negated items (that is, ~something) in unsigned ints. Some compilers
5963 apparently mutter when there is no cast.
5965 70. If an address verification called from an ACL failed, and did not produce a
5966 user-specific message (i.e. there was only a "system" message), nothing was
5967 put in $acl_verify_message. In this situation, it now puts the system
5970 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5971 session; change 4.31/43 added the unwanted input to the log line - except
5972 that it did not do this in the start of session case. It now does.
5974 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5975 This is wrong and can cause the other end to generate a synchronization
5976 error if it is another Exim or anything else that does the synchronization
5977 check. A QUIT command is no longer sent after a timeout.
5979 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5980 during host lookups.
5982 74. The defer_ok option of callout verification was not working if it was used
5983 when verifying addresses in header lines, that is, for this case:
5985 verify = header_sender/callout=defer_ok
5987 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5988 those file descriptors could be used for SMTP connections. If anything
5989 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5990 could be sent to the SMTP client, causing chaos. The daemon now opens
5991 stdin, stdout, and stderr to /dev/null when it puts itself into the
5994 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5995 log by default. The user can override this with suitable Perl magic.
5997 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5998 the log message when discard triggers, was not working for the DATA ACL or
5999 for the non-SMTP ACL.
6001 78. Error message wording change in sieve.c.
6003 79. If smtp_accept_max_per_host was set, the number of connections could be
6004 restricted to fewer than expected, because the daemon was trying to set up
6005 a new connection before checking whether the processes handling previous
6006 connections had finished. The check for completed processes is now done
6007 earlier. On busy systems, this bug wouldn't be noticed because something
6008 else would have woken the daemon, and it would have reaped the completed
6011 80. If a message was submitted locally by a user whose login name contained one
6012 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6013 It caused a spool format error. I have fixed the spool reading code. A
6014 related problem was that the "from" clause in the Received: line became
6015 illegal because of the space(s). It is now covered by ${quote_local_part.
6017 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6020 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6022 83. Patch from Sieve maintainer to fix unterminated string problem in
6023 "vacation" handling.
6025 84. Some minor changes to the Linux configuration files to help with other
6026 OS variants using glibc.
6028 85. One more patch for Sieve to update vacation handling to latest spec.
6031 ----------------------------------------------------
6032 See the note above about the 4.44 and 4.50 releases.
6033 ----------------------------------------------------
6039 1. Change 4.43/35 introduced a bug that caused file counts to be
6040 incorrectly computed when quota_filecount was set in an appendfile
6043 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6044 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6047 3. Give more explanation in the error message when the command for a transport
6048 filter fails to execute.
6050 4. There are several places where Exim runs a non-Exim command in a
6051 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6052 was being done only for the command run by the queryprogram router. It is
6053 now done for all such subprocesses. The other cases are: ${run, transport
6054 filters, and the commands run by the lmtp and pipe transports.
6056 5. Some older OS have a limit of 256 on the maximum number of file
6057 descriptors. Exim was using setrlimit() to set 1000 as a large value
6058 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6059 systems. I've change it so that if it can't get 1000, it tries for 256.
6061 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6062 was an oversight, and furthermore, ever since the addition of extra
6063 controls (e.g. 4.43/32), the checks on when to allow different forms of
6064 "control" were broken. There should now be diagnostics for all cases when a
6065 control that does not make sense is encountered.
6067 7. $recipients is now available in the predata ACL (oversight).
6069 8. Tidy the search cache before the fork to do a delivery from a message
6070 received from the command line. Otherwise the child will trigger a lookup
6071 failure and thereby defer the delivery if it tries to use (for example) a
6072 cached ldap connection that the parent has called unbind on.
6074 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6075 of $address_data from the recipient verification was clobbered by the
6076 sender verification.
6078 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6079 was its contents. (It was OK if the option was not defined at all.)
6081 11. A "Completed" log line is now written for messages that are removed from
6082 the spool by the -Mrm option.
6084 12. $host_address is now set to the target address during the checking of
6085 ignore_target_hosts.
6087 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6088 being passed; this would have caused $sender_host_name to have been used if
6089 matching the list had actually called for a host name (not very likely,
6090 since this list is usually IP addresses). A host name is now passed as
6093 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6094 code that specifies a non-restarting handler (typically sigaction() in
6095 modern systems) in an attempt to fix a rare and obscure crash bug.
6097 15. Narrowed the window for a race in the daemon that could cause it to ignore
6098 SIGCHLD signals. This is not a major problem, because they are used only to
6099 wake it up if nothing else does.
6101 16. A malformed maildirsize file could cause Exim to calculate negative values
6102 for the mailbox size or file count. Odd effects could occur as a result.
6103 The maildirsize information is now recalculated if the size or filecount
6106 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6107 support for a long time. Removed HAVE_SYS_VFS_H.
6109 18. Updated exipick to current release from John Jetmore.
6111 19. Allow an empty sender to be matched against a lookup in an address list.
6112 Previously the only cases considered were a regular expression, or an
6115 20. Exim went into a mad DNS lookup loop when doing a callout where the
6116 host was specified on the transport, if the DNS lookup yielded more than
6119 21. The RFC2047 encoding function was originally intended for short strings
6120 such as real names; it was not keeping to the 75-character limit for
6121 encoded words that the RFC imposes. It now respects the limit, and
6122 generates multiple encoded words if necessary. To be on the safe side, I
6123 have increased the buffer size for the ${rfc2047: expansion operator from
6126 22. Failure to deliver a bounce message always caused it to be frozen, even if
6127 there was an errors_to setting on the router. The errors_to setting is now
6130 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6131 canonical form (fully expanded) before being placed in
6132 $sender_host_address.
6134 24. Updated eximstats to version 1.33
6136 25. Include certificate and key file names in error message when GnuTLS fails
6137 to set them up, because the GnuTLS error message doesn't include the name
6138 of the failing file when there is a problem reading it.
6140 26. Expand error message when OpenSSL has problems setting up cert/key files.
6143 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6144 (this can affect the format of dates).
6146 28. exim_tidydb, when checking for the continued existence of a message for
6147 which it has found a message-specific retry record, was not finding
6148 messages that were in split spool directories. Consequently, it was
6149 deleting retry records that should have stayed in existence.
6151 29. eximstats updated to version 1.35
6152 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6153 1.35 - bugfix such that pie charts by volume are generated correctly
6155 30. The SPA authentication driver was not abandoning authentication and moving
6156 on to the next authenticator when an expansion was forced to fail,
6157 contradicting the general specification for all authenticators. Instead it
6158 was generating a temporary error. It now behaves as specified.
6160 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6161 (the order specifies the preference for clients). The order is now AES256,
6162 AES128, 3DES, ARCFOUR128.
6164 31. Small patch to Sieve code - explicitly set From: when generating an
6167 32. Exim crashed if a remote delivery caused a very long error message to be
6168 recorded - for instance if somebody sent an entire SpamAssassin report back
6169 as a large number of 550 error lines. This bug was coincidentally fixed by
6170 increasing the size of one of Exim's internal buffers (big_buffer) that
6171 happened as part of the Exiscan merge. However, to be on the safe side, I
6172 have made the code more robust (and fixed the comments that describe what
6175 33. Some experimental protocols are using DNS PTR records for new purposes. The
6176 keys for these records are domain names, not reversed IP addresses. The
6177 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6178 leaves it alone. Component reversal etc. now happens only for IP addresses.
6181 34. The host_aton() function is supposed to be passed a string that is known
6182 to be a valid IP address. However, in the case of IPv6 addresses, it was
6183 not checking this. This is a hostage to fortune. Exim now panics and dies
6184 if the condition is not met. A case was found where this could be provoked
6185 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6186 components; fortuitously, this particular loophole had already been fixed
6187 by change 4.50/55 or 4.44/33 above.
6189 If there are any other similar loopholes, the new check in host_aton()
6190 itself should stop them being exploited. The report I received stated that
6191 data on the command line could provoke the exploit when Exim was running as
6192 exim, but did not say which command line option was involved. All I could
6193 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6194 running as the user.
6197 35. There was a buffer overflow vulnerability in the SPA authentication code
6198 (which came originally from the Samba project). I have added a test to the
6199 spa_base64_to_bits() function which I hope fixes it.
6202 36. The daemon start-up calls getloadavg() while still root for those OS that
6203 need the first call to be done as root, but it missed one case: when
6204 deliver_queue_load_max is set with deliver_drop_privilege. This is
6205 necessary for the benefit of the queue runner, because there is no re-exec
6206 when deliver_drop_privilege is set.
6208 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6209 was in use, was not putting the data itself into the right store pool;
6210 consequently, it could be overwritten for a subsequent message in the same
6211 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6214 38. Sometimes the final signoff response after QUIT could fail to get
6215 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6216 < 0 before doing a fflush(). This bug looks as though it goes back to the
6217 introduction of TLS in release 3.20, but "sometimes" must have been rare
6218 because the tests only now provoked it.
6224 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6225 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6226 Now it returns an integer. A number of calls were still expecting a T/F
6227 return. Fortuitously, in all cases, the tests worked in OK situations,
6228 which is the norm. However, things would have gone wrong on any write
6229 failures on the smtp file descriptor. This function is used when sending
6230 messages over SMTP and also when doing verify callouts.
6232 2. When Exim is called to do synchronous delivery of a locally submitted
6233 message (the -odf or -odi options), it no longer closes stderr before doing
6236 3. Implemented the mua_wrapper option.
6238 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6240 5. Implemented the functions header_remove(), header_testname(),
6241 header_add_at_position(), and receive_remove_recipient(), and exported them
6244 6. If an ACL "warn" statement specified the addition of headers, Exim already
6245 inserted X-ACL-Warn: at the start if there was no header name. However, it
6246 was not making this test for the second and subsequent header lines if
6247 there were newlines in the string. This meant that an invalid header could
6248 be inserted if Exim was badly configured.
6250 7. Allow an ACL "warn" statement to add header lines at the start or after all
6251 the Received: headers, as well as at the end.
6253 8. Added the rcpt_4xx retry error code.
6255 9. Added postmaster_mailfrom=xxx to callout verification option.
6257 10. Added mailfrom=xxxx to the callout verification option, for verify=
6260 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6261 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6263 12. Inserted some casts to stop certain compilers warning when using pointer
6264 differences as field lengths or precisions in printf-type calls (mostly
6265 affecting debugging statements).
6267 13. Added optional readline() support for -be (dynamically loaded).
6269 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6270 same clock tick as a message's arrival, so that its received time was the
6271 same as the "first fail" time on the retry record, and that message
6272 remained on the queue past the ultimate address timeout, every queue runner
6273 would try a delivery (because it was past the ultimate address timeout) but
6274 after another failure, the ultimate address timeout, which should have then
6275 bounced the address, did not kick in. This was a "< instead of <=" error;
6276 in most cases the first failure would have been in the next clock tick
6277 after the received time, and all would be well.
6279 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6280 being recognized when the domain list was tested by the match_domain
6281 condition in an expansion string.
6283 16. Added the ${str2b64: operator.
6285 17. Exim was always calling setrlimit() to set a large limit for the number of
6286 processes, without checking whether the existing limit was already
6287 adequate. (It did check for the limit on file descriptors.) Furthermore,
6288 errors from getrlimit() and setrlimit() were being ignored. Now they are
6289 logged to the main and panic logs, but Exim does carry on, to try to do its
6290 job under whatever limits there are.
6292 18. Imported PCRE 5.0.
6294 19. Trivial typo in log message " temporarily refused connection" (the leading
6297 20. If the log selector return_path_on_delivery was set and an address was
6298 redirected to /dev/null, the delivery process crashed because it assumed
6299 that a return path would always be set for a "successful" delivery. In this
6300 case, the whole delivery is bypassed as an optimization, and therefore no
6303 21. Internal re-arrangement: the function for sending a challenge and reading
6304 a response while authentication was assuming a zero-terminated challenge
6305 string. It's now changed to take a pointer and a length, to allow for
6306 binary data in such strings.
6308 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6310 23. Exim was not respecting finduser_retries when seeking the login of the
6311 uid under which it was called; it was always trying 10 times. (The default
6312 setting of finduser_retries is zero.) Also, it was sleeping after the final
6313 failure, which is pointless.
6315 24. Implemented tls_on_connect_ports.
6317 25. Implemented acl_smtp_predata.
6319 26. If the domain in control=submission is set empty, Exim assumes that the
6320 authenticated id is a complete email address when it generates From: or
6321 Sender: header lines.
6323 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6324 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6325 chown and chgrp in /bin and hostname in /usr/bin.
6327 28. Exim was keeping the "process log" file open after each use, just as it
6328 does for the main log. This opens the possibility of it remaining open for
6329 long periods when the USR1 signal hits a daemon. Occasional processlog
6330 errors were reported, that could have been caused by this. Anyway, it seems
6331 much more sensible not to leave this file open at all, so that is what now
6334 29. The long-running daemon process does not normally write to the log once it
6335 has entered its main loop, and it closes the log before doing so. This is
6336 so that log files can straightforwardly be renamed and moved. However,
6337 there are a couple of unusual error situations where the daemon does write
6338 log entries, and I had neglected to close the log afterwards.
6340 30. The text of an SMTP error response that was received during a remote
6341 delivery was being truncated at 512 bytes. This is too short for some of
6342 the long messages that one sometimes sees. I've increased the limit to
6345 31. It is now possible to make retry rules that apply only when a message has a
6346 specific sender, in particular, an empty sender.
6348 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6349 it possible to be selective about when SMTP synchronization is enforced.
6351 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6353 32. Implemented hosts_connection_nolog.
6355 33. Added an ACL for QUIT.
6357 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6360 35. Added mailbox_size and mailbox_filecount to appendfile.
6362 36. Added control = no_multiline_responses to ACLs.
6364 37. There was a bug in the logic of the code that waits for the clock to tick
6365 in the case where the clock went backwards by a substantial amount such
6366 that the microsecond fraction of "now" was more than the microsecond
6367 fraction of "then" (but the whole seconds number was less).
6369 38. Added support for the libradius Radius client library this is found on
6370 FreeBSD (previously only the radiusclient library was supported).
6376 1. When certain lookups returned multiple values in the form name=value, the
6377 quoting of the values was not always being done properly. Specifically:
6378 (a) If the value started with a double quote, but contained no whitespace,
6380 (b) If the value contained whitespace other than a space character (i.e.
6381 tabs or newlines or carriage returns) it was not quoted.
6382 This fix has been applied to the mysql and pgsql lookups by writing a
6383 separate quoting function and calling it from the lookup code. The fix
6384 should probably also be applied to nisplus, ibase and oracle lookups, but
6385 since I cannot test any of those, I have not disturbed their existing code.
6387 2. A hit in the callout cache for a specific address caused a log line with no
6388 reason for rejecting RCPT. Now it says "Previous (cached) callout
6389 verification failure".
6391 3. There was an off-by-one bug in the queryprogram router. An over-long
6392 return line was truncated at 256 instead of 255 characters, thereby
6393 overflowing its buffer with the terminating zero. As well as fixing this, I
6394 have increased the buffer size to 1024 (and made a note to document this).
6396 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6397 when Exim is waiting for an SMTP response from a remote server, Exim
6398 restarts its select() call on the socket, thereby resetting its timeout.
6399 This is not a problem when such interrupts are rare. Somebody set up a cron
6400 job to run exiwhat every 2 minutes, which is less than the normal select()
6401 timeout (5 or 10 minutes). This meant that the select() timeout never
6402 kicked in because it was always reset. I have fixed this by comparing the
6403 time when an interrupt arrives with the time at the start of the first call
6404 to select(). If more time than the timeout has elapsed, the interrupt is
6405 treated as a timeout.
6407 5. Some internal re-factoring in preparation for the addition of Sieve
6408 extensions (by MH). In particular, the "personal" test is moved to a
6409 separate function, and given an option for scanning Cc: and Bcc: (which is
6410 not set for Exim filters).
6412 6. When Exim created an email address using the login of the caller as the
6413 local part (e.g. when creating a From: or Sender: header line), it was not
6414 quoting the local part when it contained special characters such as @.
6416 7. Installed new OpenBSD configuration files.
6418 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6419 try to make them clearer.
6421 9. Callout options, other than the timeout value, were being ignored when
6422 verifying sender addresses in header lines. For example, when using
6424 verify = header_sender/callout=no_cache
6426 the cache was (incorrectly) being used.
6428 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6429 only the Cygwin environment.
6431 11. When return_path_on_delivery was set as a log selector, if different remote
6432 addresses in the same message used different return paths and parallel
6433 remote delivery occurred, the wrong values would sometimes be logged.
6434 (Whenever a remote delivery process finished, the return path value from
6435 the most recently started remote delivery process was logged.)
6437 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6438 lines when AUTH and/or TLS are in use. This is the "received protocol"
6439 field. Exim used to use "asmtp" for authenticated SMTP, without any
6440 indication (in the protocol name) for TLS use. Now it follows the RFC and
6441 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6442 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6443 names appear in log lines as well as in Received: header lines.
6445 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6446 extensions, and comparison tests, and to fix some bugs.
6448 14. Changes to the "personal" filter test:
6450 (1) The test was buggy in that it was just doing the equivalent of
6451 "contains" tests on header lines. For example, if a user's address was
6452 anne@some.where, the "personal" test would incorrectly be true for
6454 To: susanne@some.where
6456 This test is now done by extracting each address from the header in turn,
6457 and checking the entire address. Other tests that are part of "personal"
6458 are now done using regular expressions (for example, to check local parts
6459 of addresses in From: header lines).
6461 (2) The list of non-personal local parts in From: addresses has been
6462 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6463 taken from the Sieve specification recommendations.
6465 (3) If the message contains any header line starting with "List-" it is
6466 treated as non-personal.
6468 (4) The test for "circular" in the Subject: header line has been removed
6469 because it now seems ill-conceived.
6471 15. Minor typos in src/EDITME comments corrected.
6473 16. Installed latest exipick from John Jetmore.
6475 17. If headers_add on a router specified a text string that was too long for
6476 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6477 of string_sprintf() is now avoided.
6479 18. $message_body_size was not set (it was always zero) when running the DATA
6480 ACL and the local_scan() function.
6482 19. For the "mail" command in an Exim filter, no default was being set for
6483 the once_repeat time, causing a random time value to be used if "once" was
6484 specified. (If the value happened to be <= 0, no repeat happened.) The
6485 default is now 0s, meaning "never repeat". The "vacation" command was OK
6486 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6487 (I found it when inspecting the code).
6489 20. There is now an overall timeout for performing a callout verification. It
6490 defaults to 4 times the callout timeout, which applies to individual SMTP
6491 commands during the callout. The overall timeout applies when there is more
6492 than one host that can be tried. The timeout is checked before trying the
6493 next host. This prevents very long delays if there are a large number of
6494 hosts and all are timing out (e.g. when the network connections are timing
6495 out). The value of the overall timeout can be changed by specifying an
6496 additional sub-option for "callout", called "maxwait". For example:
6498 verify = sender/callout=5s,maxwait=20s
6500 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6501 to the end before writing, but this should make it even safer).
6503 22. Exim was forgetting that it had advertised PIPELINING for the second and
6504 subsequent messages on an SMTP connection. It was also not resetting its
6505 memory on STARTTLS and an internal HELO.
6507 23. When Exim logs an SMTP synchronization error within a session, it now
6508 records whether PIPELINING has been advertised or not.
6510 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6511 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6512 rather than long int.
6514 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6516 26. Added the never_mail option to autoreply.
6522 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6523 crash if the getsockname() call failed; this can happen if a connection is
6524 closed very soon after it is established. The problem was simply in the
6525 order in which certain operations were done, causing Exim to try to write
6526 to the SMTP stream before it had set up the file descriptor. The bug has
6527 been fixed by making things happen in the correct order.
6533 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6534 before the connection was closed, thus losing the rejection response.
6536 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6537 some early Solaris releases, but causes trouble in current releases where
6538 socklen_t is defined.
6540 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6543 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6546 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6547 that releasing the top of it at the end releases what was used for sub-
6548 expansions (unless the block got too big). However, discard this block if
6549 the first thing is a variable or header, so that we can use its block when
6550 it is dynamic (useful for very large $message_headers, for example).
6552 6. Lookups now cache *every* query, not just the most recent. A new, separate
6553 store pool is used for this. It can be recovered when all lookup caches are
6554 flushed. Lookups now release memory at the end of their result strings.
6555 This has involved some general refactoring of the lookup sources.
6557 7. Some code has been added to the store_xxx() functions to reduce the amount
6558 of flapping under certain conditions.
6560 8. log_incoming_interface used to affect only the <= reception log lines. Now
6561 it causes the local interface and port to be added to several more SMTP log
6562 lines, for example "SMTP connection from", and rejection lines.
6564 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6566 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6568 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6569 could be overwritten at the end of the current message (or the start of a
6570 new message if it was set in a HELO ACL). The value is now preserved for
6571 the duration of the SMTP connection.
6573 12. If a transport had a headers_rewrite setting, and a matching header line
6574 contained an unqualified address, that address was qualified, even if it
6575 did not match any rewriting rules. The underlying bug was that the values
6576 of the flags that permit the existence of unqualified sender and recipient
6577 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6578 non-local messages, and by -bnq for local messages) were not being
6579 preserved with the message after it was received.
6581 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6582 "next input=" as part of the text comprising the host identity instead of
6583 the correct text. The code was using the same buffer for two different
6584 strings. However, depending on which order the printing function evaluated
6585 its arguments, the bug did not always show up. Under Linux, for example, my
6586 test suite worked just fine.
6588 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6589 doesn't work with some older versions of Perl. It has been changed to "my",
6590 which in any case is probably the better facility to use.
6592 15. A really picky compiler found some instances of statements for creating
6593 error messages that either had too many or two few arguments for the format
6596 16. The size of the buffer for calls to the DNS resolver has been increased
6597 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6598 for addresses that have a lot of PTR records. This alleviates a problem; it
6599 does not fully solve it.
6601 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6602 buffer now truncates the list and logs the incident, which is the same
6603 action as happens when Exim is looking up a host name and its aliases.
6604 Previously in this situation something unpredictable would happen;
6605 sometimes it was "internal error: store_reset failed".
6607 18. If a server dropped the connection unexpectedly when an Exim client was
6608 using GnuTLS and trying to read a response, the client delivery process
6609 crashed while trying to generate an error log message.
6611 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6612 string, for example:
6614 warn message = H1: something\nH2: something
6616 the text was added as a single header line from Exim's point of view
6617 though it ended up OK in the delivered message. However, searching for the
6618 second and subsequent header lines using $h_h2: did not work. This has been
6619 fixed. Similarly, if a system filter added multiple headers in this way,
6620 the routers could not see them.
6622 20. Expanded the error message when iplsearch is called with an invalid key to
6623 suggest using net-iplsearch in a host list.
6625 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6626 ACLs are no longer actually imposed (and a message to that effect is
6629 22. If a "gecos" field in a passwd entry contained escaped characters, in
6630 particular, if it contained a \" sequence, Exim got it wrong when building
6631 a From: or a Sender: header from that name. A second bug also caused
6632 incorrect handling when an unquoted " was present following a character
6633 that needed quoting.
6635 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6636 was not being matched caselessly.
6638 24. Arranged for all hyphens in the exim.8 source to be escaped with
6641 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6642 properly. Recipient callout cache records were still being keyed to include
6643 the sender, even when use_sender was set false. This led to far more
6644 callouts that were necessary. The sender is no longer included in the key
6645 when use_sender is false.
6647 26. Added "control = submission" modifier to ACLs.
6649 27. Added the ${base62d: operator to decode base 62 numbers.
6651 28. dnsdb lookups can now access SRV records.
6653 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6654 the configuration file.
6656 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6657 (-v) mode. This makes the output for a verbose queue run more intelligible.
6659 31. Added a use_postmaster feature to recipient callouts.
6661 32. Added the $body_zerocount variable, containing the number of binary zero
6662 bytes in the message body.
6664 33. The time of last modification of the "new" subdirectory is now used as the
6665 "mailbox time last read" when there is a quota error for a maildir
6668 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6670 35. Added +ignore_unknown as a special item in host lists.
6672 36. Code for decoding IPv6 addresses in host lists is now included, even if
6673 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6674 address was recognized as an IP address, but was then not correctly decoded
6675 into binary, causing unexpected and incorrect effects when compared with
6682 1. Very minor rewording of debugging text in manualroute to say "list of
6683 hosts" instead of "hostlist".
6685 2. If verify=header_syntax was set, and a header line with an unqualified
6686 address (no domain) and a large number of spaces between the end of the
6687 name and the colon was received, the reception process suffered a buffer
6688 overflow, and (when I tested it) crashed. This was caused by some obsolete
6689 code that should have been removed. The fix is to remove it!
6691 3. When running in the test harness, delay a bit after writing a bounce
6692 message to get a bit more predictability in the log output.
6694 4. Added a call to search_tidyup() just before forking a reception process. In
6695 theory, someone could use a lookup in the expansion of smtp_accept_max_
6696 per_host which, without the tidyup, could leave open a database connection.
6698 5. Added the variables $recipient_data and $sender_data which get set from a
6699 lookup success in an ACL "recipients" or "senders" condition, or a router
6700 "senders" option, similar to $domain_data and $local_part_data.
6702 6. Moved the writing of debug_print from before to after the "senders" test
6705 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6706 problems for message scanning, either using a data ACL, or using
6707 local_scan() because the Received: header was not generated till after they
6708 were called (in order to set the time as the time of reception completion).
6709 I have revised the way this works. The header is now generated after the
6710 body is received, but before the ACL or local_scan() are called. After they
6711 are run, the timestamp in the header is updated.
6717 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6718 before starting a queue runner without re-exec. This happened only when
6719 deliver_drop_privilege was set or when the Exim user was set to root. The
6720 effect of the bug was that timeouts during subsequent deliveries caused
6721 crashes instead of being properly handled. The handler is now left at its
6722 default (and expected) setting.
6724 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6725 message, again when deliver_drop_privilege is set or Exim is run as root.
6726 The bug described in (1) was not present in this case, but the tidying up
6727 of the other signals was missing. I have made the two cases consistent.
6729 3. The ignore_target_hosts setting on a manualroute router was being ignored
6730 for hosts that were looked up using the /MX notation.
6732 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6735 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6736 operated on the sender address. After changing the $sender_address to <>
6737 for the sender address verify, Exim was re-instated it as the original
6738 (before rewriting) address, but remembering that it had rewritten it, so it
6739 wasn't rewriting it again. This bug also had the effect of breaking the
6740 sender address verification caching when the sender address was rewritten.
6742 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6743 This has been changed so that if the ip literal address matches
6744 ignore_target_hosts, the router declines.
6746 7. Added expansion conditions match_domain, match_address, and match_local_
6747 part (NOT match_host).
6749 8. The placeholder for the Received: header didn't have a length field set.
6751 9. Added code to Exim itself and to exim_lock to test for a specific race
6752 condition that could lead to file corruption when using MBX delivery. The
6753 issue is with the lockfile that is created in /tmp. If this file is removed
6754 after a process has opened it but before that process has acquired a lock,
6755 there is the potential for a second process to recreate the file and also
6756 acquire a lock. This could lead to two Exim processes writing to the file
6757 at the same time. The added code performs the same test as UW imapd; it
6758 checks after acquiring the lock that its file descriptor still refers to
6759 the same named file.
6761 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6762 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6765 11. Added the smtp_active_hostname option. If used, this will typically be made
6766 to depend on the incoming interface address. Because $interface_address is
6767 not set up until the daemon has forked a reception process, error responses
6768 that can happen earlier (such as "too many connections") no longer contain
6771 12. If an expansion in a condition on a "warn" statement fails because a lookup
6772 defers, the "warn" statement is abandoned, and the next ACL statement is
6773 processed. Previously this caused the whole ACL to be aborted.
6775 13. Added the iplsearch lookup type.
6777 14. Added ident_timeout as a log selector.
6779 15. Added tls_certificate_verified as a log selector.
6781 16. Added a global option tls_require_ciphers (compare the smtp transport
6782 option of the same name). This controls incoming TLS connections.
6784 17. I finally figured out how to make tls_require_ciphers do a similar thing
6785 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6786 before starting the TLS session.
6788 18. Tabs are now shown as \t in -bP output.
6790 19. If the log selector return_path_on_delivery was set, Exim crashed when
6791 bouncing a message because it had too many Received: header lines.
6793 20. If two routers both had headers_remove settings, and the first one included
6794 a superfluous trailing colon, the final name in the first list and the
6795 first name in the second list were incorrectly joined into one item (with a
6796 colon in the middle).
6802 1. Added -C and -D options to the exinext utility, mainly to make it easier
6803 to include in the automated testing, but these could be helpful when
6804 multiple configurations are in use.
6806 2. The exinext utility was not formatting the output nicely when there was
6807 an alternate port involved in the retry record key, nor when there was a
6808 message id as well (for retries that were specific to a specific message
6809 and a specific host). It was also confused by IPv6 addresses, because of
6810 the additional colons they contain. I have fixed the IPv4 problem, and
6811 patched it up to do a reasonable job for IPv6.
6813 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6814 delivery, the log line now contains "pipelined" if PIPELINING was used.
6816 4. An SMTP transport process used to panic and die if the bind() call to set
6817 an explicit outgoing interface failed. This has been changed; it is now
6818 treated in the same way as a connect() failure.
6820 5. A reference to $sender_host_name in the part of a conditional expansion
6821 that was being skipped was still causing a DNS lookup. This no longer
6824 6. The def: expansion condition was not recognizing references to header lines
6825 that used bh_ and bheader_.
6827 7. Added the _cache feature to named lists.
6829 8. The code for checking quota_filecount in the appendfile transport was
6830 allowing one more file than it should have been.
6832 9. For compatibility with Sendmail, the command line option
6840 and sets the incoming protocol and host name (for trusted callers). The
6841 host name and its colon can be omitted when only the protocol is to be set.
6842 Note the Exim already has two private options, -pd and -ps, that refer to
6843 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6844 "s", but I don't think that's a major issue.
6846 10. A number of refactoring changes to the code, none of which should affect
6849 (a) The number of logging options was getting close to filling up the
6850 32-bit word that was used as a bit map. I have split them into two classes:
6851 those that are passed in the argument to log_write(), and those that are
6852 only ever tested independently outside of that function. These are now in
6853 separate 32-bit words, so there is plenty of room for expansion again.
6854 There is no change in the user interface or the logging behaviour.
6856 (b) When building, for example, log lines, the code previously used a
6857 macro that called string_cat() twice, in order to add two strings. This is
6858 not really sufficiently general. Furthermore, there was one instance where
6859 it was actually wrong because one of the argument was used twice, and in
6860 one call a function was used. (As it happened, calling the function twice
6861 did not affect the overall behaviour.) The macro has been replaced by a
6862 function that can join an arbitrary number of extra strings onto a growing
6865 (c) The code for expansion conditions now uses a table and a binary chop
6866 instead of a serial search (which was left over from when there were very
6867 few conditions). Also, it now recognizes conditions like "pam" even when
6868 the relevant support is not compiled in: a suitably worded error message is
6869 given if an attempt is made to use such a condition.
6871 11. Added ${time_interval:xxxxx}.
6873 12. A bug was causing one of the ddress fields not to be passed back correctly
6874 from remote delivery subprocesses. The field in question was not being
6875 subsequently used, so this caused to problems in practice.
6877 13. Added new log selectors queue_time and deliver_time.
6879 14. Might have fixed a bug in maildirsizefile handling that threw up
6880 "unexpected character" debug warnings, and recalculated the data
6881 unnecessarily. In any case, I expanded the warning message to give more
6884 15. Added the message "Restricted characters in address" to the statements in
6885 the default ACL that block characters like @ and % in local parts.
6887 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6888 Three changes have been made:
6890 (a) There was a serious bug; a negative response to MAIL caused the whole
6891 recipient domain to be cached as invalid, thereby blocking all messages
6892 to all local parts at the same domain, from all senders. This bug has
6893 been fixed. The domain is no longer cached after a negative response to
6894 MAIL if the sender used is not empty.
6896 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6899 (c) A new callout option, "use_sender" has been added for people who want
6900 the modified behaviour.
6906 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6909 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6910 indeed breaks things for older releases.
6912 3. Added additional logging to the case where there is a problem reading data
6913 from a filter that is running in a subprocess using a pipe, in order to
6914 try to track down a specific problem.
6916 4. Testing facility fudge: when running in the test harness and attempting
6917 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6918 getting "No route to host". Convert this to a timeout.
6920 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6923 6. Some OS don't have socklen_t but use size_t instead. This affects the
6924 fifth argument of getsockopt() amongst other things. This is now
6925 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6926 can be set for individual OS. I have set it for SunOS5, OSF1, and
6927 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6928 some earlier ones do not.
6930 7. Change 4.30/15 was not doing the test caselessly.
6932 8. The standard form for an IPv6 address literal was being rejected by address
6933 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6934 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6935 this, as well as the form without the "IPv6" on the front (but only when
6936 address literals are enabled, of course).
6938 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6940 10. Exim crashed if a message with an empty sender address specified by -f
6941 encountered a router with an errors_to setting. This could be provoked only
6942 by a command such as
6946 where an empty string was supplied; "<>" did not hit this bug.
6948 11. Installed PCRE release 4.5.
6950 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6951 remained set. It is now erased.
6953 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6954 times from message ids (which are base 36 rather than the normal 62).
6956 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6957 were being counted as actual protocol errors, and logged if the log
6958 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6959 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6960 and DATA following a set of rejected RCPTs do not count as protocol errors.
6961 In other words, Exim assumes they were pipelined, though this may not
6962 actually be the case. Of course, in all cases the client gets an
6963 appropriate error code.
6965 15. If a lookup fails in an ACL condition, a message about the failure may
6966 be available; it is used if testing the ACL cannot continue, because most
6967 such messages specify what the cause of the deferral is. However, some
6968 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6969 that caused an old message to be retained and used if a later statement
6970 caused a defer, replacing the real cause of the deferral.
6972 16. If an IP address had so many PTR records that the DNS lookup buffer
6973 was not large enough to hold them, Exim could crash while trying to process
6974 the truncated data. It now detects and logs this case.
6976 17. Further to 4.21/58, another change has been made: if (and only if) the
6977 first line of a message (the first header line) ends with CRLF, a bare LF
6978 in a subsequent header line has a space inserted after it, so as not to
6979 terminate the header.
6981 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6982 unnecessarily, used atoi() instead of strtol(), and didn't check the
6983 termination when getting file sizes from file names by regex.
6985 19. Completely re-implemented the support for maildirsize files, in the light
6986 of a number of problems with the previous contributed implementation
6987 (4.30/29). In particular:
6989 . If the quota is zero, the maildirsize file is maintained, but no quota is
6992 . If the maildir directory does not exist, it is created before any attempt
6993 to write a maildirsize file.
6995 . The quota value in the file is just a cache; if the quota is changed in
6996 the transport, the new value overrides.
6998 . A regular expression is available for excluding directories from the
7001 20. The autoreply transport checks the characters in options that define the
7002 message's headers; it allows continued headers, but it was checking with
7003 isspace() after an embedded newline instead of explicitly looking for a
7006 21. If all the "regular" hosts to which an address was routed had passed their
7007 expiry times, and had not reached their retry times, the address was
7008 bounced, even if fallback hosts were defined. Now Exim should go on to try
7011 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7012 equivalent code in the SMTP transport. Some hosts send humungous responses
7013 to HELO/EHLO, more than 1024 it seems.
7015 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7016 gives compiler warnings in some environments. I've now done it "properly",
7019 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7020 (because of problems with the built-in one) was declared to return uschar *
7021 instead of char *, causing compiler failure.
7023 25. Fixed a file descriptor leak when processing alias/forward files.
7025 26. Fixed a minor format string issue in dbfn.c.
7027 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7029 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7030 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7033 29. When a sender address is verified, it is cached, to save repeating the test
7034 when there is more than one recipient in a message. However, when the
7035 verification involves a callout, it is possible for different callout
7036 options to be set for different recipients. It is too complicated to keep
7037 track of this in the cache, so now Exim always runs a verification when a
7038 callout is required, relying on the callout cache for the optimization.
7039 The overhead is duplication of the address routing, but this should not be
7042 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7043 to be verified with callout=postmaster, and the main callout worked but the
7044 postmaster check failed, the verification correctly failed. However, if a
7045 subsequent RCPT command asked for sender verification *without* the
7046 postmaster check, incorrect caching caused this verification also to fail,
7049 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7050 it was not caching the DNS options (qualify_single, search_parents) that
7051 were used when the lookup failed. A subsequent lookup with different
7052 options therefore always gave the same answer, though there were cases
7053 where it should not have. (Example: a "domains = !$mx_any" option on a
7054 dnslookup router: the "domains" option is always processed without any
7055 widening, but the router might have qualify_single set.) Now Exim uses the
7056 cached value only when the same options are set.
7058 32. Added John Jetmore's "exipick" utility to the distribution.
7060 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7061 than a timeout (e.g. a certificate is required, and is not provided), an
7062 Exim server now closes the connection immediately. Previously it waited for
7063 the client to close - but if the client is SSL, it seems that they each
7064 wait for each other, leading to a delay before one of them times out.
7066 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7067 maintained 0.8.x compatibility because I don't think many are using it, and
7068 it is clearly obsolete.
7070 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7073 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7074 client certificate was expired. A simple patch fixes this, though I don't
7075 understand the full logic of why the verify callback is called multiple
7078 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7079 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7080 which causes problems with some clients (such as the Certicom SSL Plus
7081 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7082 disables the coutermeasure allowing Eudora to connect."
7084 38. Exim was not checking that a write() to a log file succeeded. This could
7085 lead to Bad Things if a log got too big, in particular if it hit a file
7086 size limit. Exim now panics and dies if it cannot write to a log file, just
7087 as it does if it cannot open a log file.
7089 39. Modified OS/Makefile-Linux so that it now contains
7091 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7093 The two -D definitions ensure that Exim is compiled with large file
7094 support, which makes it possible to handle log files that are bigger than
7097 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7098 instance) a domain was checked against a named list that involved a lookup,
7099 causing $domain_data to be set, then another domain was checked against the
7100 same list, then the first domain was re-checked, the value of $domain_data
7101 after the final check could be wrong. In particular, if the second check
7102 failed, it could be set empty. This bug probably also applied to
7105 41. The strip_trailing_dot option was not being applied to the address given
7106 with the -f command-line option.
7108 42. The code for reading a message's header from the spool was incrementing
7109 $received_count, but never initializing it. This meant that the value was
7110 incorrect (doubled) while delivering a message in the same process in which
7111 it was received. In the most common configuration of Exim, this never
7112 happens - a fresh exec is done - but it can happen when
7113 deliver_drop_privilege is set.
7115 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7116 it now includes up to 150 characters of the unexpected data in the log
7119 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7120 and building data strings. The size of both of these buffers was 10 000
7121 bytes - far larger than anybody would *ever* want, thought I. Needless to
7122 say, somebody hit the limit. I have increased the maximum line length to
7123 20 000 and the maximum data length of concatenated lines to 100 000. I have
7124 also fixed two bugs, because there was no checking on these buffers. Tsk,
7125 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7126 buffer is too small.
7128 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7129 lsearch lookups. Now it does.
7131 46. When parsing a route_list item in a manualroute router, a fixed-length
7132 buffer was used for the list of hosts. I made this 1024 bytes long,
7133 thinking that nobody would ever have a list of hosts that long. Wrong.
7134 Somebody had a whole pile of complicated expansion conditions, and the
7135 string was silently truncated, leading to an expansion error. It turns out
7136 that it is easier to change to an unlimited length (owing to other changes
7137 that have happened since this code was originally written) than to build
7138 structure for giving a limitation error. The length of the item that
7139 expands into the list of hosts is now unlimited.
7141 47. The lsearch lookup could not handle data where the length of text line was
7142 more than 4095 characters. Such lines were truncated, leading to shortened
7143 data being returned. It should now handle lines of any length.
7145 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7146 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7149 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7150 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7151 no longer happens in comment lines. A list of replacements is now placed
7152 at the head of all of the source files, except those whose only change is
7153 to replace PERL_COMMAND in the very first #! line.
7155 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7156 messages on the queue, with a bottom-up merge sort, using code contributed
7157 by Michael Haardt. This should make operations like -bp somewhat faster on
7158 large queues. It won't affect queue runners, except when queue_run_in_order
7161 51. Installed eximstats 1.31 in the distribution.
7163 52. Added support for SRV lookups to the dnslookup router.
7165 53. If an ACL referred to $message_body or $message_body_end, the value was not
7166 reset for any messages that followed in the same SMTP session.
7168 54. The store-handling optimization for building very long strings was not
7169 differentiating between the different store pools. I don't think this
7170 actually made any difference in practice, but I've tidied it.
7172 55. While running the routers to verify a sender address, $sender_address
7173 was still set to the sender address. This is wrong, because when routing to
7174 send a bounce to the sender, it would be empty. Therefore, I have changed
7175 it so that, while verifying a sender address, $sender_address is set to <>.
7176 (There is no change to what happens when verifying a recipient address.)
7178 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7179 target A or AAAA records (if not already returned) without resetting the
7180 qualify_single or search_parents options of the DNS resolver. These are
7181 inappropriate in this case because the targets of MX and SRV records must
7182 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7183 target that, when qualified, matched something in the local domain. These
7184 two options are now turned off when doing these lookups.
7186 57. It seems that at least some releases of Reiserfs (which does not have the
7187 concept of a fixed number of inodes) returns zero and not -1 for the
7188 number of available inodes. This interacted badly with check_spool_inodes,
7189 which assumed that -1 was the "no such thing" setting. What I have done is
7190 to check that the total number of inodes is greater than zero before doing
7191 the test of how many are available.
7193 58. When a "warn" ACL statement has a log_message modifier, the message is
7194 remembered, and not repeated. This is to avoid a lot of repetition when a
7195 message has many recipients that cause the same warning to be written.
7196 However, Exim was preserving the list of already written lines for an
7197 entire SMTP session, which doesn't seem right. The memory is now reset if a
7198 new message is started.
7200 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7201 debugging output unless log_rewrite was also set.
7203 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7204 of (int)(handle) when we know that handle contains (void *)(-1).
7206 61. The Exim daemon panic-logs an error return when it closes the incoming
7207 connection. However "connection reset by peer" seems to be common, and
7208 isn't really an error worthy of noting specially, so that particular error
7211 62. When Exim is trying to find all the local interfaces, it used to panic and
7212 die if the ioctl to get the interface flags failed. However, it seems that
7213 on at least one OS (Solaris 9) it is possible to have an interface that is
7214 included in the list of interfaces, but for which you get a failure error
7215 for this call. This happens when the interface is not "plumbed" into a
7216 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7217 failure of the "get flags" call assumes that the interface is down.
7219 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7220 makes life easier for people who are doing arithmetic on fields extracted
7221 from dates, where you often get leading zeros that should not be
7222 interpreted as octal.
7224 64. Added qualify_domain to the redirect router, to override the global
7227 65. If a pathologically long header line contained very many addresses (the
7228 report of this problem mentioned 10 000) and each of them was rewritten,
7229 Exim could use up a very large amount of memory. (It kept on making new
7230 copies of the header line as it rewrote, and never released the old ones.)
7231 At the expense of a bit more processing, the header rewriting function has
7232 been changed so that it no longer eats memory in this way.
7234 66. The generation of the Received: header has been moved from the time that a
7235 message starts to be received, to the time that it finishes. The timestamp
7236 in the Received: header should now be very close to that of the <= log
7237 line. There are two side-effects of this change:
7239 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7240 logged header lines no longer include the local Received: line, because
7241 it has not yet been created. The same applies to a copy of the message
7242 that is returned to a non-SMTP sender when a message is rejected.
7244 (b) When a filter file is tested using -bf, no additional Received: header
7245 is added to the test message. After some thought, I decided that this
7248 This change does not affect the value of $received_for. It is still set
7249 after address rewriting, but before local_scan() is called.
7251 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7253 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7254 gave an unhelpful panic error message, and a defer error. I have managed to
7255 change this behaviour so that it now rejects any supplied certificate,
7256 which seems right, as the list of acceptable certificates is empty.
7258 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7259 gave an unhelpful defer error. I have not managed to make this reject any
7260 supplied certificates, but the error message it gives is "no certificate
7261 supplied", which is not helpful.
7263 70. exigrep's output now also includes lines that are not associated with any
7264 message, but which match the given pattern. Implemented by a patch from
7265 Martin Sluka, which also tidied up the Perl a bit.
7267 71. Recipient callout verification, like sender verification, was using <> in
7268 the MAIL FROM command. This isn't really the right thing, since the actual
7269 sender may affect whether the remote host accepts the recipient or not. I
7270 have changed it to use the actual sender in the callout; this means that
7271 the cache record is now keyed on a recipient/sender pair, not just the
7272 recipient address. There doesn't seem to be a real danger of callout loops,
7273 since a callout by the remote host to check the sender would use <>.
7274 [SEE ABOVE: changed after hitting problems.]
7276 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7277 temporary errors. However, in the case of such a code being given after
7278 the end of a data transmission (i.e. after ".") Exim was failing to write
7279 a retry record for the message. (Yes, there was some broken host that was
7280 actually sending 8xx at this point.)
7282 73. An unknown lookup type in a host list could cause Exim to panic-die when
7283 the list was checked. (An example that provoked this was putting <; in the
7284 middle of a list instead of at the start.) If this happened during a DATA
7285 ACL check, a -D file could be left lying around. This kind of configuration
7286 error no longer causes Exim to die; instead it causes a defer error. The
7287 incident is still logged to the main and panic logs.
7289 74. Buglet left over from Exim 3 conversion. The message "too many messages
7290 in one connection" was written to the rejectlog but not the mainlog, except
7291 when address rewriting (yes!) was being logged.
7293 75. Added write_rejectlog option.
7295 76. When a system filter was run not as root (that is, when system_filter_user
7296 was set), the values of the $n variables were not being returned to the
7297 main process; thus, they were not subsequently available in the $sn
7300 77. Added +return_path_on_delivery log selector.
7302 78. A connection timeout was being treated differently from recipients deferred
7303 when testing hosts_max_try with a message that was older than the host's
7304 retry timeout. (The host should not be counted, thus allowing all hosts to
7305 be tried at least once before bouncing.) This may have been the cause of an
7306 occasionally reported bug whereby a message would remain on the queue
7307 longer than the retry timeout, but would be bounced if a delivery was
7308 forced. I say "may" because I never totally pinned down the problem;
7309 setting up timeout/retry tests is difficult. See also the next item.
7311 79. The ultimate address timeout was not being applied to errors that involved
7312 a combination of host plus message (for example, a timeout on a MAIL
7313 command). When an address resolved to a number of possible hosts, and they
7314 were not all tried for each delivery (e.g. because of hosts_max_try), a
7315 message could remain on the queue longer than the retry timeout.
7317 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7320 81. Fixed an obscure SMTP outgoing bug which required at least the following
7321 conditions: (a) there was another message waiting for the same server;
7322 (b) the server returned 5xx to all RCPT commands in the first message so
7323 that the message was not completed; (c) the server dropped the connection
7324 or gave a negative response to the RSET that Exim sends to abort the
7325 transaction. The observed case was a dropped connection after DATA that had
7326 been sent in pipelining mode. That is, the server had advertised PIPELINING
7327 but was not implementing it correctly. The effect of the bug was incorrect
7328 behaviour, such as trying another host, and this could lead to a crash.
7334 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7335 and daemon.c were passed as pointers to ints; they should have been
7336 pointers to socklen_t variables (which are typically unsigned ints).
7338 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7341 3. Fixed a really odd bug that affected only the testing scheme; patching a
7342 certain fixed string in the binary changed the value of another string that
7343 happened to be identical to the end of the original first string.
7345 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7346 name", it returns that address as the IP address. On some operating
7347 systems (e.g. Solaris), it also passes back the IP address string as the
7348 "host name". However, on others (e.g. Linux), it passes back an empty
7349 string. Exim wasn't checking for this, and was changing the host name to an
7350 empty string, assuming it had been canonicalized.
7352 5. Although rare, it is permitted to have more than one PTR record for a given
7353 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7354 all the names associated with an address, because they do in Solaris.
7355 However, it seems that they do not in Linux for data that comes from the
7356 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7357 I found this out when I moved to a new Linux workstation and tried to run
7358 the Exim test suite.
7360 To get round this problem I have changed the code so that it now does its
7361 own call to the DNS to look up PTR records when searching for a host name.
7362 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7363 addresses that are only in /etc/hosts are still found.
7365 This behaviour is, however, controlled by an option called host_lookup_
7366 order, which defaults to "bydns:byaddr". If people want to use the other
7367 order, or indeed, just use one or the other means of lookup, they can
7368 specify it in this variable.
7370 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7371 some operating systems, this comes back from gethostbyaddr() as an empty
7372 string, and this is what Exim used to test for. However, it seems that in
7373 other systems, "." is yielded. Exim now tests for this case too.
7375 7. The values of check_spool_space and check_log_space are now held internally
7376 as a number of kilobytes instead of an absolute number of bytes. If a
7377 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7378 kilobyte. This means that much larger values can be stored.
7380 8. Exim monitor: an attempt to get the action menu when not actually pointing
7381 at a message produces an empty menu entitled "No message selected". This
7382 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7383 no entries in it ("Shell widget menu has zero width and/or height"). So I
7384 have added a single, blank menu entry in this case.
7386 9. Added ${quote_local_part.
7388 10. MIME decoding is now applied to the contents of Subject: header lines when
7391 11. Now that a reference to $sender_host_address automatically causes a reverse
7392 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7393 host lookup before query-style lookups in lists that might use this
7394 variable. This has therefore been abolished, and the "net-" prefix is no
7395 longer necessary for query-style lookups.
7397 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7398 have been a typo for LFLAGS, so it has been changed.
7400 13. The install script calls Exim with "-C /dev/null" in order to find the
7401 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7402 to be output. However, since Exim outputs its version number before the
7403 error, it didn't break the script. It just looked ugly. I fixed this by
7404 always allowing "-C /dev/null" if the caller is root.
7406 14. Ignore overlarge ACL variable number when reading spool file - insurance
7407 against a later release with more variables having written the file.
7409 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7410 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7411 this, as well as the form without the "IPv6" on the front.
7413 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7414 OS/Makefile-Darwin file.
7416 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7417 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7419 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7420 variable. However, some OS use ints for inodes. I've added cast to long int
7421 to get rid of the compiler warning.
7423 19. I had forgotten to lock out "/../" in configuration file names when
7424 ALT_CONFIG_PREFIX was set.
7426 20. Routers used for verification do not need to specify transports. However,
7427 if such a router generated a host list, and callout was configured, Exim
7428 crashed, because it could not find a port number from the (non-existent)
7429 transport. It now assumes port 25 in this circumstance.
7431 21. Added the -t option to exigrep.
7433 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7434 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7435 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7436 them is set, LOOKUP_LSEARCH is forced.
7438 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7440 24. Added sender and host information to the "rejected by local_scan()" log
7441 line; previously there was no indication of these.
7443 25. Added .include_if_exists.
7445 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7446 Linux. It turns out that not all file systems support this. Apparently some
7447 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7448 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7449 not supported on the file descriptor, is now ignored when Exim is trying to
7450 sync a directory. This applies only to Linux.
7452 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7454 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7455 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7458 29. Added support for maildirsize files from supplied patch (modified a bit).
7460 30. The use of :fail: followed by an empty string could lead Exim to respond to
7461 sender verification failures with (e.g.):
7463 550 Verification failed for <xxx>
7464 550 Sender verify failed
7466 where the first response line was missing the '-' that indicates it is not
7467 the final line of the response.
7469 31. The loop for finding the name of the user that called Exim had a hardwired
7470 limit of 10; it now uses the value of finduser_retries, which is used for
7471 all other user lookups.
7473 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7476 33. Exim was neglecting to zero errno before one call of strtol() when
7477 expanding a string and expecting an integer value. On some systems this
7478 resulted in spurious "integer overflow" errors. Also, it was casting the
7479 result into an int without checking.
7481 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7482 did not work. The code looks as if it has *never* worked, though it appears
7483 to have been documented since at least release 1.62. I have made it work.
7485 35. The "timeout_DNS" error in retry rules, also documented since at least
7486 1.62, also never worked. As it isn't clear exactly what this means, and
7487 clearly it isn't a major issue, I have abolished the feature by treating it
7488 as "timeout", and writing a warning to the main and panic logs.
7490 36. The display of retry rules for -brt wasn't always showing the error code
7493 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7494 timeout_connect_A, timeout_connect_MX.
7496 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7497 to the empty sender.
7499 39. The daemon was not analysing the content of -oX till after it had closed
7500 stderr and disconnected from the controlling terminal. This meant that any
7501 syntax errors were only noted on the panic log, and the return code from
7502 the command was 0. By re-arranging the code a little, I've made the
7503 decoding happen first, so such errors now appear on stderr, and the return
7504 code is 1. However, the actual setting up of the sockets still happens in
7505 the disconnected process, so errors there are still only recorded on the
7508 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7509 connections (as happens on some IP stacks) was logged at start up time as
7510 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7511 it from "IPv6 and IPv4", which means that two separate sockets are being
7514 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7515 says whether AF_INET or AF_INET6 was passed as an argument.
7517 42. Exiwhat output was messed up when time zones were included in log
7520 43. Exiwhat now gives more information about the daemon's listening ports,
7521 and whether -tls-on-connect was used.
7523 44. The "port" option of the smtp transport is now expanded.
7525 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7526 silently ignored. Now an error message is written to the main and panic
7529 46. There's a new ACL modifier called "logwrite" which writes to a log file
7530 as soon as it is encountered.
7532 47. Added $local_user_uid and $local_user_gid at routing time.
7534 48. Exim crashed when trying to verify a sender address that was being
7537 49. Exim was recognizing only a space character after ".include". It now also
7538 recognizes a tab character.
7540 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7541 extracting the relevant information from the specification. The man page no
7542 longer contains scrambled data for the -d option, and I've added a section
7543 at the front about calling Exim under different names.
7545 51. Added "extra_headers" argument to the "mail" command in filter files.
7547 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7550 53. Installed eximstats 1.29.
7552 54. Added transport_filter_timeout as a generic transport option.
7554 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7555 Cc: header lines. This was required by RFC 822, but it not required by RFC
7558 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7559 incoming messages that did not have them. Now it does so only if the
7560 message originates locally, that is, if there is no associated remote host
7561 address. When Resent- header lines are present, this applies to the Resent-
7562 lines rather than the non-Resent- lines.
7564 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7565 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7567 58. Messages for configuration errors now include the name of the main
7568 configuration file - useful now that there may be more than one file in a
7569 list (.included file names were always shown).
7571 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7572 for those rare installations that do not start the daemon as root or run it
7573 setuid root. I've cut out the call to initgroups() if the daemon is not
7576 60. The Exim user and group can now be bound into the binary as text strings
7577 that are looked up at the start of Exim's processing.
7579 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7581 62. Added $mailstore_basename variable.
7583 63. Installed patch to sieve.c from Michael Haardt.
7585 64. When Exim failed to open the panic log after failing to open the main log,
7586 the original message it was trying to log was written to stderr and debug
7587 output, but if they were not available (the usual case in production), it
7588 was lost. Now it is written to syslog before the two lines that record the
7589 failures to open the logs.
7591 65. Users' Exim filters run in subprocesses under the user's uid. It is
7592 possible for a "deliver" command or an alias in a "personal" command to
7593 provoke an address rewrite. If logging of address rewriting is configured,
7594 this fails because the process is not running as root or exim. There may be
7595 a better way of dealing with this, but for the moment (because 4.30 needs
7596 to be released), I have disabled address rewrite logging when running a
7597 filter in a non-root, non-exim process.
7603 1. The buildconfig auxiliary program wasn't quoting the value set for
7604 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7605 not defined. This bug was masked in 4.22 by the effect that was fixed in
7608 2. Some messages that were rejected after a message id was allocated were
7609 shown as "incomplete" by exigrep. It no longer does this for messages that
7610 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7612 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7613 have allow_domain_literals set, the ID did not get logged in the <= line.
7614 Domain literals are now always recognized in Message-ID: header lines.
7616 4. The first argument for a ${extract expansion item is the key name or field
7617 number. Leading and trailing spaces in this item were not being ignored,
7618 causing some misleading effects.
7620 5. When deliver_drop_privilege was set, single queue runner processes started
7621 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7622 same command in the process it spins off) were not dropping privilege.
7624 6. When the daemon running as "exim" started a queue runner, it always
7625 re-executed Exim in the spun-off process. This is a waste of effort when
7626 deliver_drop_privilege is set. The new process now just calls the
7627 queue-runner function directly.
7633 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7636 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7637 ignored. Though the use of -oP was forcing the writing of a pid file, it
7638 was always written to the default place.
7640 3. If the message "no IP address found for host xxxx" is generated during
7641 incoming verification, it is now followed by identification of the incoming
7642 connection (so you can more easily find what provoked it).
7644 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7646 5. Added some features to "harden" Exim a bit more against certain attacks:
7648 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7649 be put in Local/Makefile. This is like the never_users runtime option,
7650 but it cannot be overridden. The default setting is "root".
7652 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7653 prefix string with which any file named in a -C command line option
7656 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7657 is retained for -C and -D only if the caller of Exim is root. Without
7658 it, the exim user may also use -C and -D and retain privilege.
7660 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7661 command line option is disabled.
7663 6. Macro names set by the -D option must start with an upper case letter, just
7664 like macro names defined in the configuration file.
7666 7. Added "dereference=" facility to LDAP.
7668 8. Two instances of the typo "uknown" in the source files are fixed.
7670 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7671 the Configure-Makefile script screwed up while processing it.
7673 10. Incorporated PCRE 4.4.
7675 11. The SMTP synchronization check was not operating right at the start of an
7676 SMTP session. For example, it could not catch a HELO sent before the client
7677 waited for the greeting. There is now a check for outstanding input at the
7678 point when the greeting is written. Because of the duplex, asynchronous
7679 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7680 way, but not yet received, when the check is performed.
7682 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7683 on TCP/IP sockets, because this apparently causes some broken clients to
7686 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7687 unchanged) from the Cygwin maintainer.
7689 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7690 was supported instead of testing for mbx. Effectively a typo.
7692 15. The spa authenticator server code was not checking that the input it
7693 received was valid base64.
7695 16. The debug output line for the "set" modifier in ACLs was not showing the
7696 name of the variable that was being set.
7698 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7700 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7701 lookup on its own. Something else was needed to trigger the lookup. For
7702 example, a match in host_lookup or the need for a host name in a host list.
7703 Now, if $sender_host_name is referenced and the host name has not yet been
7704 looked up, a lookup is performed. If the lookup fails, the variable remains
7705 empty, and $host_lookup_failed is set to "1".
7707 19. Added "eqi" as a case-independent comparison operator.
7709 20. The saslauthd authentication condition could segfault if neither service
7710 nor realm was specified.
7712 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7713 error message that was logged was misleading, and incoming SMTP
7714 connections were dropped. The message is now more accurate, and temporary
7715 errors are given to SMTP connections.
7717 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7718 (except RSET) with a 421 error, until QUIT is received. However, it was
7719 failing to send a response to QUIT.
7721 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7722 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7723 ACL is now run after the helo_try_verify_hosts code.
7725 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7726 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7727 case-independent, so other case variants are also recognized). Apparently
7728 some systems use these upper case variants.
7730 25. If more than two messages were waiting for the same host, and a transport
7731 filter was specified for the transport, Exim sent two messages over the
7732 same TCP/IP connection, and then failed with "socket operation on non-
7733 socket" when it tried to send the third.
7735 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7737 27. The extern definition of crypt16() in expand.c was not being excluded when
7738 the OS had its own crypt16() function.
7740 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7741 as a preferred synonym for return_size_limit, both as an option and as an
7744 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7746 30. Changed the default configuration ACL to relax the local part checking rule
7747 for addresses that are not in any local domains. For these addresses,
7748 slashes and pipe symbols are allowed within local parts, but the sequence
7749 /../ is explicitly forbidden.
7751 31. SPA server authentication was not clearing the challenge buffer before
7754 32. log_message in a "warn" ACL statement was writing to the reject log as
7755 well as to the main log, which contradicts the documentation and doesn't
7756 seem right (because no rejection is happening). So I have stopped it.
7758 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7759 However, I am unable to do any testing of this.
7761 34. Fixed an infelicity in the appendfile transport. When checking directories
7762 for a mailbox, to see if any needed to be created, it was accidentally
7763 using path names with one or more superfluous leading slashes; tracing
7764 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7766 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7767 contents are added to the log line that is written for every discarded
7768 recipient. (Previously a log_message setting was ignored.)
7770 36. The ${quote: operator now quotes the string if it is empty.
7772 37. The install script runs exim in order to find its version number. If for
7773 some reason other than non-existence or emptiness, which it checks, it
7774 could not run './exim', it was installing it with an empty version number,
7775 i.e. as "exim-". This error state is now caught, and the installation is
7778 38. An argument was missing from the function that creates an error message
7779 when Exim fails to connect to the socket for saslauthd authentication.
7780 This could cause Exim to crash, or give a corrupted message.
7782 39. Added isip, isip4, and isip6 to ${if conditions.
7784 40. The ACL variables $acl_xx are now saved with the message, and can be
7785 accessed later in routers, transports, and filters.
7787 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7788 strings in the file are not string-expanded.
7790 42. If a MAIL command specified a SIZE value that was too large to fit into an
7791 int variable, the check against message_size_limit failed. Such values are
7792 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7793 day this will have to be increased, but I don't think I want to be around
7794 when emails are that large.
7801 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7802 iconv() is not standard in FreeBSD.
7804 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7805 IPv6 enabled. The observed symptom was a segmentation fault on return from
7806 the function os_common_find_running_interfaces() in src/os.c.
7808 3. In the check_special_case() function in daemon.c I had used "errno" as an
7809 argument name, which causes warnings on some systems. This was basically a
7810 typo, since it was named "eno" in the comments!
7812 4. The code that waits for the clock to tick (at a resolution of some fraction
7813 of a second) so as to ensure message-id uniqueness was always waiting for
7814 at least one whole tick, when it could have waited for less. [This is
7815 almost certainly not relevant at current processor speeds, where it is
7816 unlikely to ever wait at all. But we try to future-proof.]
7818 5. The function that sleeps for a time interval that includes fractions of a
7819 second contained a race. It did not block SIGALRM between setting the
7820 timer, and suspending (a couple of lines later). If the interval was short
7821 and the sigsuspend() was delayed until after it had expired, the suspension
7822 never ended. On busy systems this could lead to processes getting stuck for
7825 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7826 process, before it forks any delivery processes. The open lookup caching
7827 mechanism meant that the open file or database connection was passed into
7828 the delivery process. The problem was that delivery processes always tidy
7829 up cached lookup data. This could cause a problem for the next delivery
7830 process started by the queue runner, because the external queue runner
7831 process does not know about the closure. So the next delivery process
7832 still has data in the lookup cache. In the case of a file lookup, there was
7833 no problem because closing a file descriptor in a subprocess doesn't affect
7834 the parent. However, if the lookup was caching a connection to a database,
7835 the connection was closed, and the second delivery process was likely to
7836 see errors such as "PGSQL: query failed: server closed the connection
7837 unexpectedly". The problem has been fixed by closing all cached lookups
7838 in a queue runner before running a delivery process.
7840 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7841 seem to have the "const" qualifier which it has on other OS. I've
7844 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7845 *of the same type* (client or server) with the same public name that an
7846 error should be diagnosed.
7848 9. When Exim looked up a host name for an IP address, but failed to find the
7849 original IP address when looking up the host name (a safety check), it
7850 output the message "<ip address> does not match any IP for NULL", which was
7851 confusing, to say the least. The bug was that the host name should have
7852 appeared instead of "NULL".
7854 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7855 user that is built into the binary, and the -C or -D options is used, root
7856 privilege is dropped before the configuration file is read. In addition,
7857 logging is switched to stderr instead of the normal log files. If the
7858 configuration then re-defines the Exim user, the unprivileged environment
7859 is probably not what is expected, so Exim logs a panic warning message (but
7862 However, if deliver_drop_privilege is set, the unprivileged state may well
7863 be exactly what is intended, so the warning has been cut out in that case,
7864 and Exim is allowed to try to write to its normal log files.
7870 1. smtp_return_error_details was not giving details for temporary sender
7871 or receiver verification errors.
7873 2. Diagnose a configuration error if two authenticators have the same public
7876 3. Exim used not to create the message log file for a message until the first
7877 delivery attempt. This could be confusing when incoming messages were held
7878 for policy or load reasons. The message log file is now created at the time
7879 the message is received, and an initial "Received" line is written to it.
7881 4. The automatically generated man page for command line options had a minor
7882 bug that caused no ill effects; however, a more serious problem was that
7883 the procedure for building the man page automatically didn't always
7884 operate. Consequently, release 4.20 contains an out-of-date version. This
7885 shouldn't happen again.
7887 5. When building Exim with embedded Perl support, the script that builds the
7888 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7889 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7891 6. The freeze_tell option was not being used for messages that were frozen on
7892 arrival, either by an ACL or by local_scan().
7894 7. Added the smtp_incomplete_transaction log selector.
7896 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7897 was accepting AUTH without a new EHLO.
7899 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7900 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7903 10. Logging of TCP/IP connections (when configured) now happens in the main
7904 daemon process instead of the child process, so that the TCP/IP connection
7905 count is more accurate (but it can never be perfect).
7907 11. The use of "drop" in a nested ACL was not being handled correctly in the
7908 outer ACL. Now, if condition failure induced by the nested "drop" causes
7909 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7910 or "require"), the connection is dropped.
7912 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7913 that yield "discard" can now be used with an "accept" or a "discard" verb,
7914 but an error is generated for any others (because I can't see a useful way
7915 to define what should happen).
7917 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7918 are now processed in the same way as lines in the Exim configuration file.
7919 In particular, continuation lines are supported.
7921 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7923 15. Added -ti meaning -t -i.
7925 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7926 domains, and warn by logging if others are found.
7928 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7929 returned from GIFCONF when Exim is trying to find the list of interfaces on
7930 a host. The code in os.c has been modified to copy each ifreq to an aligned
7931 structure in all cases.
7933 Also, in some cases, the returned ifreq's were being copied to a 'struct
7934 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7935 means the last couple of bytes of an IPv6 address could be chopped if the
7936 ifreq contained only a normal sockaddr (14 bytes storage).
7938 18. Named domain lists were not supported in the hosts_treat_as_local option.
7939 An entry such as +xxxx was not recognized, and was treated as a literal
7942 19. Ensure that header lines added by a DATA ACL are included in the reject log
7943 if the ACL subsequently rejects the message.
7945 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7946 MD5 (which is deprecated).
7948 21. When testing a filter file using -bf, Exim was writing a message when it
7949 took the sender from a "From " line in the message, but it was not doing so
7950 when it took $return_path from a Return-Path: header line. It now does.
7952 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7953 with a valid header line field name (a series of printing characters
7954 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7956 23. Changed "disc" in the source to "disk" to conform to the documentation and
7957 the book and for uniformity.
7959 24. Ignore Sendmail's -Ooption=value command line item.
7961 25. When execve() failed while trying to run a command in a pipe transport,
7962 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7963 could be confused with a return value of 69 from the command itself. This
7964 has been changed to 127, the value the shell returns if it is asked to run
7965 a non-existent command. The wording for the related log line suggests a
7966 non-existent command as the problem.
7968 26. If received_header_text expands to an empty string, do not add a Received:
7969 header line to the message. (Well, it adds a token one on the spool, but
7970 marks it "old" so that it doesn't get used or transmitted.)
7972 27. Installed eximstats 1.28 (addition of -nt option).
7974 28. There was no check for failure on the call to getsockname() in the daemon
7975 code. This can fail if there is a shortage of resources on the system, with
7976 ENOMEM, for example. A temporary error is now given on failure.
7978 29. Contrary to the C standard, it seems that in some environments, the
7979 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7980 program. Exim now does this explicitly; it affects the formatting of
7981 timestamps using strftime().
7983 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7984 complaints. I've now initialized all the variables, to avoid this.
7986 32. Header lines added by a system filter were not being "seen" during
7987 transport-time rewrites.
7989 33. The info_callback() function passed to OpenSSL is set up with type void
7990 (*)(SSL *, int, int), as described somewhere. However, when calling the
7991 function (actually a macro) that sets it up, the type void(*)() is
7992 expected. I've put in a cast to prevent warnings from picky compilers.
7994 34. If a DNS black list lookup found a CNAME record, but there were no A
7995 records associated with the domain it pointed at, Exim crashed.
7997 35. If a DNS black list lookup returned more than one A record, Exim ignored
7998 all but the first. It now scans all returned addresses if a particular IP
7999 value is being sought. In this situation, the contents of the
8000 $dnslist_value variable are a list of all the addresses, separated by a
8003 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8004 to accept a host name and all its aliases if the forward lookup for any of
8005 them yielded the IP address of the incoming connection. Now it accepts only
8006 those names whose forward lookup yields the correct IP address. Any other
8007 names are discarded. This closes a loophole whereby a rogue DNS
8008 administrator could create reverse DNS records to break through a
8009 wildcarded host restriction in an ACL.
8011 37. If a user filter or a system filter that ran in a subprocess used any of
8012 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8013 the wrong values were passed to the pipe command ($thisaddress had the
8014 value of $0, $0 had the value of $1, etc). This bug was introduced by
8015 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8017 38. Improved the line breaking for long SMTP error messages from ACLs.
8018 Previously, if there was no break point between 40 and 75 characters, Exim
8019 left the rest of the message alone. Two changes have been made: (a) I've
8020 reduced the minimum length to 35 characters; (b) if it can't find a break
8021 point between 35 and 75 characters, it looks ahead and uses the first one
8022 that it finds. This may give the occasional overlong line, but at least the
8023 remaining text gets split now.
8025 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8026 file descriptors might be low, and that setting 1000 would always raise it.
8027 It turns out that in some environments, the limit is already over 1000 and
8028 that lowering it causes trouble. So now Exim takes care not to decrease it.
8030 40. When delivering a message, the value of $return_path is set to $sender_
8031 address at the start of routing (routers may change the value). By an
8032 oversight, this default was not being set up when an address was tested by
8033 -bt or -bv, which affected the outcome if any router or filter referred to
8036 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8037 log only when "message" or "log_message" are set. However, if one of the
8038 conditions was an address verification, or a call to a nested ACL, the
8039 messages generated by the underlying test were being passed through. This
8040 no longer happens. The underlying message is available in $acl_verify_
8041 message for both "message" and "log_message" expansions, so it can be
8042 passed through if needed.
8044 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8045 new expansion $bh_ to give the encoded byte string without charset
8046 translation. Translation happens only if iconv() is available; HAVE_ICONV
8047 indicates this at build time. HEADERS_CHARSET gives the charset to
8048 translate to; headers_charset can change it in the configuration, and
8049 "headers charset" can change it in an individual filter file.
8051 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8052 that creates RFC 2047 encoded "words" labels them as that charset instead
8053 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8054 expansion operator; (ii) when Exim creates a From: line for a local
8055 message; (iii) when a header line is rewritten to include a "phrase" part.
8057 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8058 buggy, causing it to skip the first lines of messages whose message ID
8059 ended in 'D'. This would not have bitten before Exim release 4.14, because
8060 message IDs were unlikely to end in 'D' before then. The effect was to have
8061 incorrect size information for certain domains.
8063 45. #include "config.h" was missing at the start of the crypt16.c module. This
8064 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8067 46. If there was a timeout during a "random" callout check, Exim treated it as
8068 a failure of the random address, and carried on sending RSET and the real
8069 address. If the delay was just some slowness somewhere, the response to the
8070 original RCPT would be taken as a response to RSET and so on, causing
8071 mayhem of various kinds.
8073 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8074 when I implemented it. It didn't allow for the fact that some option values
8075 may legitimately be negative (e.g. size_addition), and it didn't even do
8076 the right test for positive values.
8078 48. Domain names in DNS records are case-independent. Exim always looks them up
8079 in lower case. Some resolvers return domain names in exactly the case they
8080 appear in the zone file, that is, they may contain uppercase letters. Not
8081 all resolvers do this - some return always lower case. Exim was treating a
8082 change of case by a resolver as a change of domain, similar to a widening
8083 of a domain abbreviation. This triggered its re-routing code and so it was
8084 trying to route what was effectively the same domain again. This normally
8085 caused routing to fail (because the router wouldn't handle the domain
8086 twice). Now Exim checks for this case specially, and just changes the
8087 casing of the domain that it ultimately uses when it transmits the message
8090 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8093 50. If a filter generated a file delivery with a non-absolute name (possible if
8094 no home directory exists for the router), the forbid_file option was not
8097 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8098 the existing equality matching.
8100 52. Exim was using ints instead of ino_t variables in some places where it was
8101 dealing with inode numbers.
8103 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8104 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8105 called TMPDIR, and if it finds it is different, it changes its value.
8107 54. The smtp_printf() function is now made available to local_scan() so
8108 additional output lines can be written before returning. There is also an
8109 smtp_fflush() function to enable the detection of a dropped connection.
8110 The variables smtp_input and smtp_batched_input are exported to
8113 55. Changed the default runtime configuration: the message "Unknown user"
8114 has been removed from the ACL, and instead placed on the localuser router,
8115 using the cannot_route_message feature. This means that any verification
8116 failures that generate their own messages won't get overridden. Similarly,
8117 the "Unrouteable address" message that was in the ACL for unverifiable
8118 relay addresses has also been removed.
8120 56. Added hosts_avoid_esmtp to the smtp transport.
8122 57. The exicyclog script was not checking for the esoteric option
8123 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8124 will work only if exicyclog is run under the appropriate euid.
8126 58. Following a discussion on the list, the rules by which Exim recognises line
8127 endings on incoming messages have been changed. The -dropcr and drop_cr
8128 options are now no-ops, retained only for backwards compatibility. The
8129 following line terminators are recognized: LF CRLF CR. However, special
8130 processing applies to CR:
8132 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8133 nor a local message in the state where . is a terminator.
8135 (ii) If a bare CR is encountered in a header line, an extra space is added
8136 after the line terminator so as not to end the header. The reasoning
8137 behind this is that bare CRs in header lines are most likely either
8138 to be mistakes, or people trying to play silly games.
8140 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8141 was being incorrectly given as 18 bytes larger than it should have been.
8142 This is a VOB (very old bug).
8144 60. This may never have affected anything current, but just in case it has:
8145 When the local host is found other than at the start of a list of hosts,
8146 the local host, those with the same MX, and any that follow, are discarded.
8147 When the list in question was part of a longer list of hosts, the following
8148 hosts (not currently being processed) were also being discarded. This no
8149 longer happens. I'm not sure if this situation could ever has previously
8152 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8155 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8156 header. This is something that is recommended in a new Internet Draft, and
8157 is something that is documented as being done by Sendmail. There are two
8158 possible values. For messages generated by the autoreply transport, Exim
8161 Auto-Submitted: auto-replied
8163 whereas for all other generated messages (e.g. bounces) it adds
8165 Auto-Submitted: auto-generated
8167 63. The "personal" condition in filters now includes a test for the
8168 Auto-Submitted: header. If it contains the string "auto-" the message it
8169 not considered personal.
8171 64. Added rcpt_include_affixes as a generic transport option.
8173 65. Added queue_only_override (default true).
8175 66. Added the syslog_duplication option.
8177 67. If what should have been the first header line of a message consisted of
8178 a space followed by a colon, Exim was mis-interpreting it as a header line.
8179 It isn't of course - it is syntactically invalid and should therefore be
8180 treated as the start of the message body. The misbehaviour could have
8181 caused a number of strange effects, including loss of data in subsequent
8182 header lines, and spool format errors.
8184 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8185 client host had authenticated. This control can now be exercised by an ACL
8186 for more flexibility.
8188 69. By default, callouts do not happen when testing with -bh. There is now a
8189 variant, -bhc, which does actually run the callout code, including
8190 consulting and updating the callout cache.
8192 70. Added support for saslauthd authentication, courtesy of Alexander
8195 71. If statvfs() failed on the spool or log directories while checking their
8196 size for availability, Exim confusingly gave the error "space shortage".
8197 Furthermore, in debugging mode it crashed with a floating point exception.
8198 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8199 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8200 serious problem, Exim now writes to the main and panic logs when this
8201 happens, with details of the failure. It then refuses to accept the
8202 incoming message, giving the message "spool directory problem" or "log
8203 directory problem" with a 421 code for SMTP messages.
8205 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8206 0, 1, and 2 exist, because some OS complain for execs without them (see
8207 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8208 descriptors. However, the code omitted to check that the open succeeded,
8209 causing mysterious errors if for some reason the permissions on /dev/null
8210 got screwed. Now Exim writes a message to the main and panic logs, and
8211 bombs out if it can't open /dev/null.
8213 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8214 interact so that it is all more flexible. It is supposed to remain
8215 backwards compatible. Also added extra_local_interfaces.
8217 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8218 to bomb out with an assertion failure - to the client this appears as a
8219 connection drop. This problem occurs in the part of the code that was taken
8220 from the Samba project. Fortunately, the assertion is in a very simple
8221 function, so I have fixed this by reproducing the function inline in the
8222 one place where it is called, and arranging for authentication to fail
8223 instead of killing the process with assert().
8225 75. The SPA client code was not working when the server requested OEM rather
8226 than Unicode encoding.
8228 76. Added code to make require_files with a specific uid setting more usable in
8229 the case where statting the file as root fails - usually a non-root-mounted
8230 NFS file system. When this happens and the failure is EACCES, Exim now
8231 forks a subprocess and does the per-uid checking as the relevant uid.
8233 77. Added process_log_path.
8235 78. If log_file_path was not explicitly set, a setting of check_log_space or
8236 check_log_inodes was ignored.
8238 79. If a space check for the spool or log partitions fails, the incident is now
8239 logged. Of course, in the latter case the data may get lost...
8241 80. Added the %p formatting code to string_format() so that it can be used to
8242 print addresses in debug_print(). Adjusted all the address printing in the
8243 debugging in store.c to use %p rather than %d.
8245 81. There was a concern that a line of code in smtp_in.c could overflow a
8246 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8247 initially expressed, the concern was not well-founded, because trailing
8248 spaces are removed early. However, if the trailing spaces were followed by
8249 a NULL, they did not get removed, so the overflow was possible. Two fixes
8252 (a) I re-wrote the offending code in a cleaner fashion.
8253 (b) If an incoming SMTP command contains a NULL character, it is rejected
8256 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8257 runs initgroups(), so that if the Exim user is in any additional groups,
8258 they will be used during message reception.
8264 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob