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.
143 JH/32 Fix CHUNKING for a second message on a connection when the first was
144 rejected. Previously we did not reset the chunking-offered state, and
145 erroneously rejected the BDAT command. Investigation help from
148 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
149 an empty address. Previously the expansion returned an error.
155 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
156 related applications. Previously an "H" was used where available info
157 says that "M" should be, so change to match.
159 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
160 as arguments, so an implementation trying to copy these into a local
161 buffer was taking a taint-enforcement trap. Fix by using dynamically
162 created buffers. Similar fix for radius expansion condition.
164 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
165 reasonable, eg. to count headers. Fix by using dynamically created
166 buffers rather than a local. Do similar fixes for ACL actions "dcc",
167 "log_reject_target", "malware" and "spam"; the arguments are expanded
168 so could be handling tainted values.
170 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
171 broken the (no-op) support for this sendmail command. Restore it
172 to doing nothing, silently, and returning good status.
174 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
175 record path was given (or the default used) without a leading directory
176 path, an error occurred on trying to open it. Use the transport's working
179 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
180 transport. Previously it was the name on the DNS A-record; use instead
181 the head of the CNAME chain leading there (if there is one). This seems
182 to align better with RFC 6125.
184 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
185 smtp_accept_max_per_host allocated resources which were not released
186 when the limit was exceeded. This eventually crashed the daemon. Fix
187 by adding a release action in that path.
189 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
190 expanded; previously using tainted values was rejected. Fix by using
191 dynamically-created buffers.
193 JH/09 Relax restrictions on ACL verify condition needing access to message
194 headers. Previously they were only permitted in data and non-smtp ACLs;
195 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
196 not_blind, header_sender and header_names_ascii verification.
198 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
199 Previously a macro used one argument twice; when called with the
200 argument as an expression having side-effects, incorrect operation
201 resulted. Use an inlineable function.
203 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
204 held open for a verify callout. Previously this wan not accounted for
205 and a corrupt onward SMTP conversation resulted.
207 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
208 Previously they were embedded naively in the constructed address; when
209 needed, strip the quoting and quote the entire local_part.
210 Also make the inbound_srs expansion condition handle quoting.
212 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
213 excluded, not matching the documentation.
215 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
216 was given for the sqlite_dbfile a trap resulted.
218 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
219 "name" argument resulted in a trap. There is no reason to disallow such;
220 this was a coding error.
222 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
223 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
224 spent suspended, ignoring the POSIX definition. Previously we assumed
225 it did and a constant offset from real time could be used as a correction.
226 Change to using the same clock source for the start-of-message and the
227 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
228 exists, just to get a clock slightly more aligned to reality.
230 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
231 RFC says it is optional some validators care. The missing char was not
232 intended but triggered by a line-wrap alignment. Discovery and fix by
233 Guillaume Outters, hacked on by JH.
235 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
236 name being quoted was tainted a trap would be taken. Fix by using
237 dynamically created buffers. The routine could have been called by a
238 rewrite with the "h" flag, by using the "-F" command-line option, or
239 by using a "name=" option on a control=submission ACL modifier.
241 JH/19 SPF: change the Authentication-Results expansion component to give
242 smtp.helo when the sender domain is empty. Previously it gave
245 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
246 not in 127.0.0.0/8 to help in spotting list domains taken over by a
247 domain-parking registrar.
249 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
250 Previously when a whitespace character was specified it was not inserted
251 after removing the newline.
253 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
254 the domain part of the recipient address. This overrides any tls_sni
255 option set, which was previously used.
257 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
260 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
261 is_tainted() had an off-by-one error in the overenthusiastic direction.
262 Find and fix by Gavan. Although NetBSD is not a supported platform for
263 4.94 this bug could affect other platforms.
265 PP/01 Fix default prime selection to be consistent.
266 One path used ike23 still, instead of exim.dev.20160529.3; now both
267 execution flows will use the same DH primes (currently
268 exim.dev.20160529.3).
270 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
271 option in smtp transport, to match the documentation. Previously
272 verification was not being done in some cases where it should have been.
274 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
275 than one server was defined and depending on the platform memory layout
276 details, an internal consistency trap could be hit while walking the list
279 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
280 files. The value is used by the authresults expansion item. Previously
281 if this was used in a router or transport, a crash could result.
283 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
284 have a sufficient incidence of EINTR returns from select that an
285 interaction with other operations done by the main daemon loop exposed
286 a bug in the error-handling. This was benign apart from the log
289 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
290 for consistency with delivered "=>" and failed "**" lines. While we're
291 there, handle PRX and TFO.
293 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
294 applied. This resulted, if any header-line rewrite rules were configured,
295 in a panic-log triggerable by sending a message with a long address in
296 a header. Fix by increasing the arbitrary limit to larger than a single
297 (dewrapped) 5322 header line maximum size.
299 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
300 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
301 RFC 6648 which deprecates X- options in protocols as a general practice.
302 Changeover between the implementations is handled by the mechanisms
305 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
306 a remote which uses callout/hold. Previously the local was queued.
308 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
311 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
312 log line. The asterisk is now consistently appended to the remote IP
313 (and port, if given), and will also be provided on defer and fail log
314 lines. Previously it could be placed on the local IP if that was being
315 logged, and was only provided on delivery lines.
317 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
319 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
320 authenticator client_send option. Previously the next char, after a pair
321 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
322 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
323 documentation. There is still no way to get a leading ^ immediately
324 after a NUL (ie. for the password of a PLAIN method authenticator.
326 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
327 files. For bad sizes read, delete the record and whine to paniclog.
329 JH/38 When logging an AUTH failure, as server, do not include sensitive
330 information. Previously, the credentials would be included if given
331 as part of the AUTH command line and an ACL denied authentication.
333 JH/39 Bug 2691: fix $local_part_data. When the matching list element
334 referred to a file, bad data was returned. This likely also affected
337 JH/40 The gsasl authenticator now supports caching of the salted password
338 generated by the client-side implementation. This required the addition
339 of a new variable: $auth4.
341 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
342 left undeleted; the attempt to re-create it then failed - resulting in
343 the usual "SIGHUP tp have daemon reload configuration" to not work.
344 This affected any platform not supporting "abstract" Unix-domain
345 sockets (i.e. not Linux).
347 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
348 recipients" response to RCPT in a later response, with a 250. The
349 previous coding assumed this would not happen, and under PIPELINING
350 would result in both lost and duplicate recipients for a message.
352 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
353 Previously the weighting was incorrectly applied. Similar fix for socks
354 proxies. Found and fixed by Heiko Schlichting.
356 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
357 not handle sub-lists included using the +namedlist syntax. While
358 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
359 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
362 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
363 Previously the permanent pool was used, so the sore could not be freed.
364 This meant a connection with many messages would use continually-growing
367 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
368 per-pool so as not to waste too much space. Previously a constant size
369 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
370 DOS attacks harder. The cost is wasted memory use in the larger blocks.
372 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
373 same space repeatedly, and a smaller process footprint.
375 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
376 for GnuTLS. Previously the name started "%s" which, while not a bug,
377 looked as if if might be one.
379 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
380 "continued connection") make the $proxy_* variables available. Previously
381 the information was not passed across the exec() call for subsequent
382 transport executions. This also mean that the log lines for the
383 messages can show the proxy information.
385 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
386 receive time. With modern systems the difference is significant.
387 The historical behaviour can be restored by disabling (a new) log_selector
388 "queue_time_exclusive".
390 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
391 content) were specifically tested for. Now, also cover expansions
392 resulting in ACL names and inline ACL content.
394 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
395 making it unusable in complex expressions.
397 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
398 fail with a taint-check on the spoolfile name, and leave the message
401 HS/01 Enforce absolute PID file path name.
403 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
405 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
407 PP/02 Bug 2643: Correct TLS DH constants.
408 A missing NUL termination in our code-generation tool had led to some
409 incorrect Diffie-Hellman constants in the Exim source.
410 Reported by kylon94, code-gen tool fix by Simon Arlott.
412 PP/03 Impose security length checks on various command-line options.
413 Fixes CVE-2020-SPRSS reported by Qualys.
415 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
416 better. Reported by Qualys.
418 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
419 providing a particularly obnoxious sender full name.
422 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
424 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
427 PP/08 Change default for recipients_max from unlimited to 50,000.
429 PP/09 Fix security issue with too many recipients on a message (to remove a
430 known security problem if someone does set recipients_max to unlimited,
431 or if local additions add to the recipient list).
432 Fixes CVE-2020-RCPTL reported by Qualys.
434 PP/10 Fix security issue in SMTP verb option parsing
435 Fixes CVE-2020-EXOPT reported by Qualys.
437 PP/11 Fix security issue in BDAT state confusion.
438 Ensure we reset known-good where we know we need to not be reading BDAT
439 data, as a general case fix, and move the places where we switch to BDAT
440 mode until after various protocol state checks.
441 Fixes CVE-2020-BDATA reported by Qualys.
443 HS/03 Die on "/../" in msglog file names
445 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
446 the Exim runtime user are allowed to create files.
448 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
451 QS/03 When reading the output from interpreted forward files we do not
452 pass the pipe between the parent and the interpreting process to
453 executed child processes (if any).
455 QS/04 Always die if requested from internal logging, even is logging is
458 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
459 the API; compilation noo longer completes with DMARC support included.
460 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
461 been reported on other platforms.
463 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
465 JH/56 Make the majority of info read from config files readonly, for defence-in-
466 depth against exploits. Suggestion by Qualys.
467 Not supported on Solaris 10.
469 JH/57 Fix control=fakreject for a custom message containing tainted data.
470 Previously this resulted in a log complaint, due to a re-expansion present
471 since fakereject was originally introduced.
473 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
474 modifier was given, a loop resulted.
476 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
477 pipelined SMTP command followed the BDAT LAST then it would be
478 incorrectly treated as part of the message body, causing a verification
481 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
482 require looser validation rules than those for 821-level addresses,
483 which only permit IP addresses.
489 JH/01 Avoid costly startup code when not strictly needed. This reduces time
490 for some exim process initialisations. It does mean that the logging
491 of TLS configuration problems is only done for the daemon startup.
493 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
495 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
496 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
499 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
500 DKIM signing is being done. Previously a transport_filter would always
501 disable CHUNKING, falling back to traditional DATA.
503 JH/05 Regard command-line recipients as tainted.
505 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
507 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
508 PAM library frees one of the arguments given to it, despite the
509 documentation. Therefore a plain malloc must be used.
511 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
512 on-stack buffers were used, resulting in a taint trap when DSN information
513 copied from a received message was written into the buffer.
515 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
516 the ordering of its ARC headers. This caused a crash.
518 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
519 a new record was being constructed with information from the peer, a trap
522 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
523 installation would get error messages from DMARC verify, when it hit the
524 nonexistent file indicated by the default. Distros wanting DMARC enabled
525 should both provide the file and set the option.
526 Also enforce no DMARC verification for command-line sourced messages.
528 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
529 could, depending on the platform, hang at the STARTTLS response.
531 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
532 message on a connection. Previously if one message had ARC headers and
533 the following one did not, a crash could result when adding an
534 Authentication-Results: header.
536 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
537 Exim main code and Exim-related utities. The introduction of taint
538 tracking also did many adjustments to string handling. Since then, eximon
539 frequently terminated with an assert failure.
541 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
542 check for 452 responses. This slightly helps the inefficieny of doing
543 a large alias-expansion into a recipient-limited target. The max_rcpt
544 transport option still applies (and at the current default, will override
545 the new feature). The check is done for either cause of synch, and forces
546 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
547 connection. The new facility is not tunable at this time.
549 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
550 library live data was being used, so the results became garbage. Make
551 copies while it is still usable.
553 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
554 on delivery deferred (==) and failed (**) lines (if a delivery was
555 attemtped). Previously it was only on completion (=>) lines.
557 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
558 for the expansion of the server_scram_iter and server_scram_salt options.
560 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
561 are now specifically given a NO_DATA response without hitting the system
562 resolver. The library goes on to do the now-standard TXT lookup.
563 Use of dnsdb lookups is not affected.
565 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
566 only retrieve the errormessage once. Previously two calls to dlerror()
567 were used, and the second one (for mainlog/paniclog) retrieved null
570 JH/20 Taint checking: disallow use of tainted data for
571 - the appendfile transport file and directory options
572 - the pipe transport command
573 - the autoreply transport file, log and once options
574 - file names used by the redirect router (including filter files)
576 - paths used by single-key lookups
577 Previously this was permitted.
579 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
580 adjusted the size of a major service buffer; this failed because the
581 buffer was in use at the time. Change to a compile-time increase in the
582 buffer size, when this authenticator is compiled into exim.
584 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
585 previous fast-mode was untenable in the face of glibs using mmap to
586 support larger malloc requests.
588 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
589 New values supported, if defined on system where compiled:
590 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
591 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
593 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
594 running a limited number of proceses in parallel, a benefit is gained. The
595 amount varies with the platform hardware and load. The use of the option
596 queue_run_in_order means we cannot do this, as ordering becomes
599 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
600 had introduced a string-copy (for ensuring NUL-termination) which was not
601 appropriate for that case, which can include embedded NUL bytes in the
602 block of data. Investigation showed the copy to actually be needless, the
603 data being length-specified.
605 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
606 done during a receiving connection, and both used TLS, global info was
607 used rather than per-connection info for tracking the state of data
608 queued for transmission. This could result in a connection hang.
610 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
611 Previously, when delivering serveral messages down a single connection
612 only the first would provide a SIZE. This was due to the size information
613 not being properly tracked.
615 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
616 TAI (at 37 seconds currently), pretend to be in UTC for time-related
617 expansion and logging. Previously, spurious values such as a future
618 minute could be seen.
620 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
621 it could crash from a null-deref. This could also affect the
622 ${addresses: } operator and ${readsock } item.
624 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
625 message following a mime one, the variable was not reset.
627 JH/30 When an pipelined-connect fails at the first response, assume incorrect
628 cached capability (perhaps the peer reneged?) and immediately retry in
631 JH/31 Fix spurious detection of timeout while writing to transport filter.
633 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
634 an attempt to copy the string was made before checking it.
636 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
637 taint of the lookup key was maintained; we now regard the presence in the
638 filesystem as sufficient validation.
640 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
641 argument is supplied.
643 JH/35 The dsearch lookup now requires that the directory is an absolute path.
644 Previously this was not checked, and nonempty relative paths made an
645 access under Exim's current working directory.
647 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
648 Previously no event was raised.
650 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
651 parameter supplied by the sender MAIL FROM command. Previously it was
652 ignored, and only the check_spool_space option value for the required
655 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
656 the size of the signing public-key. Previously it was instead giving
657 the size of the signature hash.
659 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
660 the default. See the (new) dkim_verify_min_keysizes option.
662 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
663 and an ACL use a lookup for checking either the local_part or domain,
664 stale data could be accessed. Ensure that variable references are
665 dropped between messages.
667 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
668 by the client was not checked as pointing within response data before
669 being used. A malicious client could thus cause an out-of-bounds read and
670 possibly gain authentication. Fix by adding the check.
672 JH/42 Internationalisation: change the default for downconversion in the smtp
673 transport to be "if needed". Previously it was "as previously set" for
674 the message, which usually meant "if needed" for message-submission but
675 "no" for everything else. However, MTAs have been seen using SMTPUTF8
676 even when the envelope addresses did not need it, resulting in forwarding
677 failures to non-supporting MTAs. A downconvert in such cases will be
678 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
679 transport. The change does mean that addresses needing conversion will
680 be converted when previously a delivery failure would occur.
682 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
683 response was received it would be used unchecked in a fail-DSN, violating
684 standards on line-length limits. Truncate if needed.
686 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
687 doesn't work. (Additionally add a new main config option to configure the
694 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
695 side, to decode a TLS 1.3 packet capture.
697 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
698 Previously the default library behaviour applied, sending two, each in
701 JH/03 Debug output for ACL now gives the config file name and line number for
704 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
706 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
708 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
709 buffer overrun for (non-chunking) other transports.
711 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
712 TLS1.3, means that a server rejecting a client certificate is not visible
713 to the client until the first read of encrypted data (typically the
714 response to EHLO). Add detection for that case and treat it as a failed
715 TLS connection attempt, so that the normal retry-in-clear can work (if
716 suitably configured).
718 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
719 and/or domain. Found and fixed by Jason Betts.
721 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
722 configuration). If a CNAME target was not a wellformed name pattern, a
725 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
726 the OS reports them interleaved with other addresses.
728 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
729 used both for input and for a verify callout, both encrypted, SMTP
730 responses being sent by the server could be lost. This resulted in
731 dropped connections and sometimes bounces generated by a peer sending
734 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
735 string. Previously it was possible to cause undefined behaviour in a
736 library routine (usually a crash). Found by "zerons".
738 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
741 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
742 API was removed, so update to use the newer ones.
744 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
745 any timeout set, is taking a long time. Previously we would hang on to a
746 rotated logfile "forever" if the input was arriving with long gaps
747 (a previous attempt to fix addressed lack, for a long time, of initial
750 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
751 shared (NFS) environment. The length of the tempfile name is now
752 4 + 16 ("hdr.$message_exim_id") which might break on file
753 systems which restrict the file name length to lower values.
756 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
757 shared (NFS) environment.
759 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
760 did for all versions <4.90). Notably -M, -m, --invert, -I may be
763 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
764 on some platforms for bit 31.
766 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
767 to changes apparently associated with TLS1.3 handling some of the APIs
768 previously used were either nonfunctional or inappropriate. Strings
769 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
770 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
771 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
772 This affects log line X= elements, the $tls_{in,out}_cipher variables,
773 and the use of specific cipher names in the encrypted= ACL condition.
775 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
777 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
778 verification result was not updated unless hosts_require_ocsp applied.
780 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
781 queue_list_requires_admin set to false, non-admin users were denied the
784 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
785 directory-of-certs mode. Previously they were advertised despite the
788 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
789 A single TCP connection by a client will now hold a TLS connection open
790 for multiple message deliveries, by default. Previously the default was to
793 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
794 default. If built with the facility, DANE will be used. The facility
795 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
797 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
798 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
799 must be defined and you must still, unless you define DISABLE_TLS, manage
800 the the include-dir and library-file requirements that go with that
801 choice. Non-TLS builds are still supported.
803 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
806 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
807 default. If the platform supports and has the facility enabled, it will
808 be requested on all coneections.
810 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
811 controlled by the build-time option SUPPORT_PIPE_CONNECT.
813 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
815 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
816 success-DSN messages. Previously the From: header was always the default
817 one for these; the option was ignored.
819 JH/28 Fix the timeout on smtp response to apply to the whole response.
820 Previously it was reset for every read, so a teergrubing peer sending
821 single bytes within the time limit could extend the connection for a
822 long time. Credit to Qualsys Security Advisory Team for the discovery.
824 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
825 delivery address, which leaked information of the results of local
826 forwarding. Change to the original envelope recipient address, per
829 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
830 requested. Previously not bounce was generated and a log entry of
831 error ignored was made.
833 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
835 JH/32 Introduce a general tainting mechanism for values read from the input
836 channel, and values derived from them. Refuse to expand any tainted
837 values, to catch one form of exploit.
839 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
840 was unused and the unexpanded text used for the test. Found and
841 fixed by Ruben Jenster.
843 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
844 an attempt to use a TLS library read routine dereffed a nul pointer,
847 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
848 them in our smtp response.
850 JH/36 Have the general router option retry_use_local_part default to true when
851 any of the restrictive preconditions are set (to anything). Previously it
852 was only for check_local user. The change removes one item of manual
853 configuration which is required for proper retries when a remote router
854 handles a subset of addresses for a domain.
856 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
857 link count into consideration.
859 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
860 caused the extension of big_buffer, the following lines were ignored.
862 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
863 accordance with RFC 2308. Previously there was no expiry, so a longlived
864 receive process (eg. due to ACL delays) versus a short SOA value could
867 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
869 JH/39 Promote DMARC support to mainline.
871 JH/40 Bug 2452: Add a References: header to DSNs.
873 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
874 parameters. The relevant library call is documented as "Deprecated: This
875 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
876 3.6.0, DH parameters are negotiated following RFC7919."
878 HS/06 Change the default of dnssec_request_domains to "*"
880 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
881 carried on and emitted a BDAT command, even when PIPELINING was not
884 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
885 buffer was used for the filename, resulting in a trap when tainted
886 arguments (eg. $domain) were used.
888 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
889 recommended to avoid a possible server-load attack. The feature can be
890 re-enabled via the openssl_options main cofiguration option.
892 JH/45 local_scan API: documented the current smtp_printf() call. This changed
893 for version 4.90 - adding a "more data" boolean to the arguments.
894 Bumped the ABI version number also, this having been missed previously;
895 release versions 4.90 to 4.92.3 inclusive were effectively broken in
896 respect of usage of smtp_printf() by either local_scan code or libraries
897 accessed via the ${dlfunc } expansion item. Both will need coding
898 adjustment for any calls to smtp_printf() to match the new function
899 signature; a FALSE value for the new argument is always safe.
901 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
902 the file-offset (which the Linux syscall does, and exim expects); this
903 resulted in an indefinite loop.
905 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
906 to do ARC verification. The Authentication-Results: header line added
907 by the configuration then had no ARC item.
909 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
910 an incorrect mode was used for reading the result, resulting in it being
917 JH/01 Remove code calling the customisable local_scan function, unless a new
918 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
920 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
921 non-signal-safe functions being used.
923 JH/03 Bug 2269: When presented with a received message having a stupidly large
924 number of DKIM-Signature headers, disable DKIM verification to avoid
925 a resource-consumption attack. The limit is set at twenty.
927 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
928 report of oldest_pass in ${authres } in consequence, and separate out
929 some descriptions of reasons for verification fail.
931 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
932 files in the spool were present and unlocked. A queue-runner could spot
933 them, resulting in a duplicate delivery. Fix that by doing the unlock
934 after the unlink. Investigation by Tim Stewart. Take the opportunity to
935 add more error-checking on spoolfile handling while that code is being
938 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
939 No known attacks, no CVE, this is defensive hardening.
941 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
942 a queue-runner could start a delivery while other operations were ongoing.
943 Cutthrough delivery was a common victim, resulting in duplicate delivery.
944 Found and investigated by Tim Stewart. Fix by using the open message data
945 file handle rather than opening another, and not locally closing it (which
946 releases a lock) for that case, while creating the temporary .eml format
947 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
949 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
950 $sender_verify_failure/$recipient_verify_failure to "random".
952 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
955 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
956 Previously this would segfault.
958 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
961 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
962 like zero, since the resolver should be doing this for us, But we need one
963 as a CNAME but no MX presence gets the CNAME returned; we need to check
964 that doesn't point to an MX to declare it "no MX returned" rather than
965 "error, loop". A new main option is added so the older capability of
966 following some limited number of chain links is maintained.
968 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
970 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
971 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
972 now-deprecated earlier definitions used only specified the range up to TLS
973 1.2 (in the older-version library docs).
975 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
977 JH/15 Rework TLS client-side context management. Stop using a global, and
978 explicitly pass a context around. This enables future use of TLS for
979 connections to service-daemons (eg. malware scanning) while a client smtp
980 connection is using TLS; with cutthrough connections this is quite likely.
982 JH/16 Fix ARC verification to do AS checks in reverse order.
984 JH/17 Support a "tls" option on the ${readsocket } expansion item.
986 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
987 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
988 Previously the "utf8" would be re-prepended for every additional message.
990 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
991 Previously thery were accepted, resulting in issues when attempting to
992 forward messages to a non-supporting MTA.
994 PP/02 Let -n work with printing macros too, not just options.
996 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
997 one parent address was copied, and bogus data was used at delivery-logging
998 time. Either a crash (after delivery) or bogus log data could result.
999 Discovery and analysis by Tim Stewart.
1001 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1002 Previously if the string ended mid-character, we did not insert the
1003 promised '?' replacement.
1005 PP/04 Documentation: current string operators work on bytes, not codepoints.
1007 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1008 should pack well giving a smaller memory footprint so better caching and
1009 therefore performance. Group the declarations where this can't be done so
1010 that the byte-sized flag variables are not interspersed among pointer
1011 variables, giving a better chance of good packing by the compiler.
1013 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1014 non-null, to avoid issues with sites running BATV. Previously reports were
1015 sent with an empty envelope sender so looked like bounces.
1017 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1018 The ignore_error flag wasn't being returned from the filter subprocess so
1019 was not set for later routers. Investigation and fix by Matthias Kurz.
1021 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1022 and a msg:complete for the whole, when a message is manually removed using
1023 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1025 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1026 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1027 hope that is portable enough.
1029 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1030 requiring a known-CA anchor certificate; make it now rely entirely on the
1031 TLSA as an anchor. Checking the name on the leaf cert against the name
1032 on the A-record for the host is still done for TA (but not for EE mode).
1034 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1035 included in delivery lines for non-proxied connections, when compiled with
1036 SUPPORT_SOCKS and running with proxy logging enabled.
1038 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1039 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1040 move the existing event to fire before the normal logging of message
1041 failure so that custom logging is bracketed by normal logging.
1043 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1044 msg:fail:internal event. Developement by Matthias Kurz.
1046 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1047 far too small for todays use of crypto signatures stored there. Go all
1048 the way to the max DNS message size of 64kB, even though this might be
1049 overmuch for IOT constrained device use.
1051 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1052 copy a string over itself. The library routine is documented as not
1053 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1055 JH/32 For main options check_spool_space and check_inode_space, where the
1056 platform supports 64b integers, support more than the previous 2^31 kB
1057 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1058 the previous G, M, k.
1060 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1061 $authenticated_fail_id variable on authentication failure. Previously
1064 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1065 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1066 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1067 GNUTLS_SEC_PARAM_MEDIUM.
1069 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1070 side. Previously we would continue as if no SNI had been received.
1072 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1073 "<" character, which should be a list-separator specification, we walked
1074 off past the nul-terimation.
1076 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1077 causes) even when the retry time is not yet met. Previously they were
1078 not, meaning that when (say) an account was over-quota and temp-rejecting,
1079 and multiple senders' messages were queued, only one sender would get
1080 notified on each configured delay_warning cycle.
1082 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1084 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1085 cipher-suites, an error can be left on the stack even for a succeeding
1086 accept; this results in impossible error messages when a later operation
1089 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1090 return error codes indicating retry. Under TLS1.3 this becomes required.
1092 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1093 it only wrote the new authenticators, resulting in a lack of tracking of
1094 peer changes of ESMTP extensions until the next cache flush.
1096 JH/41 Fix the loop reading a message header line to check for integer overflow,
1097 and more-often against header_maxsize. Previously a crafted message could
1098 induce a crash of the recive process; now the message is cleanly rejected.
1100 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1101 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1107 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1108 When redis_servers is set to a list of > 1 element, and the Redis servers
1109 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1110 case of MOVED into a DEFER case instead, thus moving the query onto the
1111 next server in the list. For a cluster of N elements, all N servers must
1112 be defined in redis_servers.
1114 GF/02 Catch and remove uninitialized value warning in exiqsumm
1115 Check for existence of @ARGV before looking at $ARGV[0]
1117 JH/01 Replace the store_release() internal interface with store_newblock(),
1118 which internalises the check required to safely use the old one, plus
1119 the allocate and data copy operations duplicated in both (!) of the
1120 extant use locations.
1122 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1123 modifier. This matches the restriction on the commandline.
1125 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1126 Previously only the last row was returned.
1128 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1129 we assumed that tags in the header were well-formed, and parsed the
1130 element content after inspecting only the first char of the tag.
1131 Assumptions at that stage could crash the receive process on malformed
1134 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1135 While running the DKIM ACL we operate on the Permanent memory pool so that
1136 variables created with "set" persist to the DATA ACL. Also (at any time)
1137 DNS lookups that fail create cache records using the Permanent pool. But
1138 expansions release any allocations made on the current pool - so a dnsdb
1139 lookup expansion done in the DKIM ACL releases the memory used for the
1140 DNS negative-cache, and bad things result. Solution is to switch to the
1141 Main pool for expansions.
1142 While we're in that code, add checks on the DNS cache during store_reset,
1143 active in the testsuite.
1144 Problem spotted, and debugging aided, by Wolfgang Breyha.
1146 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1147 When none of the hosts presented to a transport match an already-open
1148 connection, close it and proceed with the list. Previously we would
1149 queue the message. Spotted by Lena with Yahoo, probably involving
1152 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1153 Previously a spurious "250 OK id=" response was appended to the proper
1156 JH/08 The "support for" informational output now, which built with Content
1157 Scanning support, has a line for the malware scanner interfaces compiled
1158 in. Interface can be individually included or not at build time.
1160 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1161 by the template makefile "src/EDITME". The "STREAM" support for an older
1162 ClamAV interface method is removed.
1164 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1165 rows affected is given instead).
1167 JH/11 The runtime Berkeley DB library version is now additionally output by
1168 "exim -d -bV". Previously only the compile-time version was shown.
1170 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1171 SMTP connection. Previously, when one had more recipients than the
1172 first, an abortive onward connection was made. Move to full support for
1173 multiple onward connections in sequence, handling cutthrough connection
1174 for all multi-message initiating connections.
1176 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1177 routers. Previously, a multi-recipient message would fail to match the
1178 onward-connection opened for the first recipient, and cause its closure.
1180 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1181 a timeout on read on a GnuTLS initiating connection, resulting in the
1182 initiating connection being dropped. This mattered most when the callout
1183 was marked defer_ok. Fix to keep the two timeout-detection methods
1186 JH/15 Relax results from ACL control request to enable cutthrough, in
1187 unsupported situations, from error to silently (except under debug)
1188 ignoring. This covers use with PRDR, frozen messages, queue-only and
1191 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1193 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1194 metadata, resulting in a crash in free().
1196 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1197 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1198 Broken also in d185889f4, with init system revamp.
1200 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1201 Previously we abruptly closed the connection after reading a malware-
1202 found indication; now we go on to read the "scan ok" response line,
1205 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1206 ACL. Previously, a crash would result.
1208 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1209 macro text after a replacement (previously it was only once per line) and
1210 by skipping builtin macros when searching for an uppercase lead character.
1212 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1213 for the build is renamed.
1215 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1216 was allocated for every new TLS startup, meaning one per message. Fix
1217 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1219 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1220 reported the original. Fix to report (as far as possible) the ACL
1221 result replacing the original.
1223 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1224 OpenSSL. Certificate information is loaded for every new TLS startup,
1225 and the resources needed to be freed.
1227 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1229 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1232 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1233 DATA response info to the (existing) per-recipient response info for
1234 the "C=" log element. It can have useful tracking info from the
1235 destination system. Patch from Simon Arlott.
1237 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1238 length value. Previously this would segfault.
1240 HS/02 Support Avast multiline protoocol, this allows passing flags to
1241 newer versions of the scanner.
1243 JH/28 Ensure that variables possibly set during message acceptance are marked
1244 dead before release of memory in the daemon loop. This stops complaints
1245 about them when the debug_store option is enabled. Discovered specifically
1246 for sender_rate_period, but applies to a whole set of variables.
1247 Do the same for the queue-runner and queue-list loops, for variables set
1248 from spool message files. Do the same for the SMTP per-message loop, for
1249 certain variables indirectly set in ACL operations.
1251 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1252 as a multi-recipient message from a mailinglist manager). The coding had
1253 an arbitrary cutoff number of characters while checking for more input;
1254 enforced by writing a NUL into the buffer. This corrupted long / fast
1255 input. The problem was exposed more widely when more pipelineing of SMTP
1256 responses was introduced, and one Exim system was feeding another.
1257 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1258 receiving system, and refused recipients seen by the sending system
1259 (propating to people being dropped from mailing lists).
1260 Discovered and pinpointed by David Carter.
1262 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1263 replaced by the ${authresults } expansion.
1265 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1267 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1268 allows proper process termination in container environments.
1270 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1271 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1273 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1274 and "err_perm", deprecated since 4.83 when the RFC-defined words
1275 "temperror" and "permerror" were introduced.
1277 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1278 transport-filters or DKIM-signing. The restriction was lost in the
1279 consolidation of verify-callout and delivery SMTP handling.
1280 Extend the restriction to also cover ARC-signing.
1282 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1283 in defer=pass mode supply a 450 to the initiator. Previously the message
1286 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1287 tls_require_ciphers is used as before.
1289 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1290 "pass_unscanned". Only tmpfails from the scanner are written to
1291 the paniclog, as they may require admin intervention (permission
1292 denied, license issues). Other scanner errors (like decompression
1293 bombs) do not cause a paniclog entry.
1295 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1296 Previously it was possible to log spurious information in receive log
1299 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1300 triggered odd behaviour from Outlook Express clients.
1302 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1305 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1306 since the IETF WG has not yet settled on that versus the original
1307 "bare" representation.
1309 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1310 Previously the millisecond value corrupted the output.
1311 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1312 corrupted the output.
1318 JH/01 Rework error string handling in TLS interface so that the caller in
1319 more cases is responsible for logging. This permits library-sourced
1320 string to be attached to addresses during delivery, and collapses
1321 pairs of long lines into single ones.
1323 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1324 during configuration. Wildcards are allowed and expanded.
1326 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1327 This permits better logging.
1329 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1330 do not close it down and have the child transport start it up again on
1331 the passed-on TCP connection. Instead, proxy the child (and any
1332 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1333 affected: the continued delivery log lines do not have any DNSSEC, TLS
1334 Certificate or OCSP information. TLS cipher information is still logged.
1336 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1337 identical IP addresses on different listening ports. Will also affect
1340 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1341 add noisy ifdef guards to special-case this sillyness.
1342 Patch from Bernd Kuhls.
1344 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1345 than 255 are no longer allowed.
1347 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1348 Disable the session-cache too, which might reduce our load. Since we
1349 currrectly use a new context for every connection, both as server and
1350 client, there is no benefit for these.
1351 GnuTLS appears to not support tickets server-side by default (we don't
1352 call gnutls_session_ticket_enable_server()) but client side is enabled
1353 by default on recent versions (3.1.3 +) unless the PFS priority string
1356 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1357 <https://reproducible-builds.org/specs/source-date-epoch/>.
1359 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1360 the check for any unsuccessful recipients did not notice the limit, and
1361 erroneously found still-pending ones.
1363 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1364 MSG_MORE. Only in-clear (not on TLS connections).
1366 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1367 transport-filter is involved we can buffer the headers in memory for
1368 creating the signature, and read the spool data file once for the
1369 signature and again for transmission.
1371 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1372 4.77 as the kernel support then wasn't solid, having issues in 64bit
1373 mode. Now, it's been long enough. Add support for FreeBSD also.
1375 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1376 case where the routing stage had gathered several addresses to send to
1377 a host before calling the transport for the first, we previously failed
1378 to close down TLS in the old transport process before passing the TCP
1379 connection to the new process. The new one sent a STARTTLS command
1380 which naturally failed, giving a failed delivery and bloating the retry
1381 database. Investigation and fix prototype from Wolfgang Breyha.
1383 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1384 false-negatives in the check that the sender had not preempted a response
1385 or prompt from Exim (running as a server), due to that code's lack of
1386 awareness of the SMTP input buffering.
1388 PP/04 Add commandline_checks_require_admin option.
1389 Exim drops privileges sanely, various checks such as -be aren't a
1390 security problem, as long as you trust local users with access to their
1391 own account. When invoked by services which pass untrusted data to
1392 Exim, this might be an issue. Set this option in main configuration
1393 AND make fixes to the calling application, such as using `--` to stop
1396 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1397 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1398 responses to those, into a single TLS record each way (this usually means
1399 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1402 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1403 affects you only if you're dancing at the edge of the param size limits.
1404 If you are, and this message makes sense to you, then: raise the
1405 configured limit or use OpenSSL 1.1. Nothing we can do for older
1408 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1409 cmdline element to get the documented default one. Previously it was
1412 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1413 are unsuable for non-message-specific reasons - eg. connection timeout,
1414 banner-time rejection.
1416 JH/16 Fix logging of delivery remote port, when specified by router, under
1419 PP/06 Repair manualroute's ability to take options in any order, even if one
1420 is the name of a transport.
1423 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1425 JH/17 Change the list-building routines interface to use the expanding-string
1426 triplet model, for better allocation and copying behaviour.
1428 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1429 Previously it was constructed the first time a possibly-matching string
1430 was met in the configuration file input during startup; now it is done
1433 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1434 compatible one, to avoid the (poorly documented) possibility of a config
1435 file in the working directory redirecting the DB files, possibly correpting
1436 some existing file. CVE-2017-10140 assigned for BDB.
1438 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1439 cache-hot. Previously, although the result was properly cached, the
1440 initial verify call returned a defer.
1442 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1443 the main verify for receipient in uncached-mode.
1445 JH/22 Retire historical build files to an "unsupported" subdir. These are
1446 defined as "ones for which we have no current evidence of testing".
1448 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1449 if present. Previously it was ignored.
1451 JH/24 Start using specified-initialisers in C structure init coding. This is
1452 a C99 feature (it's 2017, so now considered safe).
1454 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1455 if was a fixed-sized field and bitmask ops via macros; it is now more
1458 PP/07 GitHub PR 56: Apply MariaDB build fix.
1459 Patch provided by Jaroslav Škarvada.
1461 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1462 during Coverity cleanups [4.87 JH/47]
1463 Diagnosis and fix provided by Michael Fischer v. Mollard.
1465 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1466 the right size to place the terminating semicolon on its own folded
1467 line, the header hash was calculated to an incorrect value thanks to
1468 the (relaxed) space the fold became.
1470 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1471 and confused the parent.
1473 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1474 which could crash as a result. This could lead to undeliverable messages.
1476 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1479 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1480 matters on fast-turnover and PID-randomising systems, which were getting
1481 out-of-order delivery.
1483 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1484 a possibly-overlapping copy. The symptom was that "Remote host closed
1485 connection in response to HELO" was logged instead of the actual 4xx
1488 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1489 Previously only that bufferd was discarded, resulting in SYMTP command
1492 JH/32 DKIM: when a message has multiple signatures matching an identity given
1493 in dkim_verify_signers, run the dkim acl once for each. Previously only
1494 one run was done. Bug 2189.
1496 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1497 "panic the current process" to "deliberately defer". The panic log is
1498 still written with the problem list name; the mail and reject logs now
1499 get a temp-reject line for the message that was being handled, saying
1500 something like "domains check lookup or other defer". The SMTP 451
1501 message is still "Temporary local problem".
1503 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1504 A crafted sequence of BDAT commands could result in in-use memory beeing
1505 freed. CVE-2017-16943.
1507 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1508 from SMTP input. Previously it was always done; now only done for DATA
1509 and not BDAT commands. CVE-2017-16944.
1511 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1512 to the message (such as an overlong header line). Previously this was
1513 not done and we did not exit BDAT mode. Followon from the previous item
1514 though a different problem.
1520 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1521 than -2003 did; needs libidn2 in addition to libidn.
1523 JH/02 The path option on a pipe transport is now expanded before use.
1525 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1526 Patch provided by "Björn", documentation fix added too.
1528 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1529 missing a wire-to-host endian conversion.
1531 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1532 close after a BDAT command line could be taken as a following command,
1533 giving a synch failure. Fix by only checking for synch immediately
1534 before acknowledging the chunk.
1536 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1537 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1538 macro. Patches provided by Josh Soref.
1540 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1541 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1542 by IANA as a service extension. However, John Klensin suggests that we
1545 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1546 the dkim code may be unix-mode line endings rather than smtp wire-format
1547 CRLF, so prepend a CR to any bare LF.
1549 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1550 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1552 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1553 the first were themselves being wrongly included in the feed into dkim
1554 processing; with most chunk sizes in use this resulted in an incorrect
1555 body hash calculated value.
1557 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1558 DKIM signature block, for verification. Although advised against by
1559 standards it is specifically not ruled illegal.
1561 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1563 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1564 missing a body hash (the bh= tag).
1566 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1567 It seems that HAProxy sends the Proxy Protocol information in clear and
1568 only then does a TLS startup, so do the same.
1570 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1571 TCP connections (such as for Spamd) unless the daemon successfully set
1572 Fast Open mode on its listening sockets. This fixes breakage seen on
1573 too-old kernels or those not configured for Fast Open, at the cost of
1574 requiring both directions being enabled for TFO, and TFO never being used
1575 by non-daemon-related Exim processes.
1577 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1578 endings, at least on the first header line. Try to canonify any that get
1579 past that check, despite the cost.
1581 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1582 now limited to an arbitrary five deep, while parsing addresses with the
1583 strip_excess_angle_brackets option enabled.
1585 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1586 instead leave the unprompted TLS handshake in socket buffer for the
1587 TLS library to consume.
1589 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1591 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1593 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1594 region in one operation, for speed, and this leaves assigned pointers
1595 dangling. Add checks run only under the testsuite which checks all
1596 variables at a store-reset and panics on a dangling pointer; add code
1597 explicitly nulling out all the variables discovered. Fixes one known
1598 bug: a transport crash, where a dangling pointer for $sending_ip_address
1599 originally assigned in a verify callout, is re-used.
1601 PP/06 Drop '.' from @INC in various Perl scripts.
1603 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1605 PP/08 Reduce a number of compilation warnings under clang; building with
1606 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1607 should be warning-free.
1609 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1611 HS/01 Fix portability problems introduced by PP/08 for platforms where
1612 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1614 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1615 chunk. This allows us to accept broken chunked messages. We need a more
1616 general solution here.
1618 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1619 already-broken messages in the queue.
1621 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1623 JH/19 Fix reference counting bug in routing-generated-address tracking.
1629 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1630 supports it and a size is available (ie. the sending peer gave us one).
1632 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1633 years of being deprecated). The replacements are the ACLs
1634 acl_smtp_mime and acl_not_smtp_mime.
1636 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1637 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1638 or even an in-clear connection were permitted. Now, if the host lookup
1639 was dnssec and dane was requested then the host is only used if the
1640 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1641 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1642 if one fails this test.
1643 This means that a poorly-configured remote DNS will make it incommunicado;
1644 but it protects against a DNS-interception attack on it.
1646 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1647 non-noisy when a race steals the message being considered.
1649 JH/05 If main configuration option tls_certificate is unset, generate a
1650 self-signed certificate for inbound TLS connections.
1652 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1653 in rewrites and routers.
1655 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1656 and logged a warning sing 4.83; now they are a configuration file error.
1658 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1659 (lacking @domain). Apply the same qualification processing as RCPT.
1661 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1663 JH/10 Support ${sha256:} applied to a string (as well as the previous
1666 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1667 a cutthrough deliver is pending, as we always want to make a connection.
1668 This also avoids re-routing the message when later placing the cutthrough
1669 connection after a verify cache hit.
1670 Do not update it with the verify result either.
1672 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1673 when routing results in more than one destination address.
1675 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1676 signing (which inhibits the cutthrough capability). Previously only
1677 the presence of an option was tested; now an expansion evaluating as
1678 empty is permissible (obviously it should depend only on data available
1679 when the cutthrough connection is made).
1681 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1682 the relevant preceding SMTP command did not note the pipelining mode.
1684 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1685 Previously they were not counted.
1687 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1688 as one having no matching records. Previously we deferred the message
1689 that needed the lookup.
1691 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1692 distinguished as "(=".
1694 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1695 for missing MX records. Previously it only worked for missing A records.
1697 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1699 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1700 after the data-go-ahead and data-ack. Patch from Jason Betts.
1702 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1703 even for a "none" policy. Patch from Tony Meyer.
1705 JH/22 Fix continued use of a connection for further deliveries. If a port was
1706 specified by a router, it must also match for the delivery to be
1709 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1710 When one of the recipients of a message was unsuitable for the connection
1711 (has no matching addresses), we lost track of needing to mark it
1712 deferred. As a result mail would be lost.
1714 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1716 JH/25 Decoding ACL controls is now done using a binary search; the source code
1717 takes up less space and should be simpler to maintain. Merge the ACL
1718 condition decode tables also, with similar effect.
1720 JH/26 Fix problem with one_time used on a redirect router which returned the
1721 parent address unchanged. A retry would see the parent address marked as
1722 delivered, so not attempt the (identical) child. As a result mail would
1725 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1726 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1727 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1730 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1731 Main config options check_{log,spool}_{inodes,space} are now
1732 100 inodes, 10MB unless set otherwise in the configuration.
1734 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1735 Previously it only applied to the main-section connection policy
1738 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1740 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1741 by me. Added RFC7919 DH primes as an alternative.
1743 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1744 are not in the system include path.
1746 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1747 GnuTLS, when a session startup failed (eg because the client disconnected)
1748 Exim did stdio operations after fclose. This was exposed by a recent
1749 change which nulled out the file handle after the fclose.
1751 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1752 signed directly by the cert-signing cert, rather than an intermediate
1753 OCSP-signing cert. This is the model used by LetsEncrypt.
1755 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1757 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1758 an incoming connection.
1760 HS/02 Bug 1802: Do not half-close the connection after sending a request
1763 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1764 fallback to "prime256v1".
1766 JH/34 SECURITY: Use proper copy of DATA command in error message.
1767 Could leak key material. Remotely exploitable. CVE-2016-9963.
1773 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1774 and 3.4.4 - once the server is enabled to respond to an OCSP request
1775 it does even when not requested, resulting in a stapling non-aware
1776 client dropping the TLS connection.
1778 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1779 support variable-length bit vectors. No functional change.
1781 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1782 The I= interface field on outgoing lines is now after the H= remote
1783 host field, same as incoming lines. There is a separate
1784 outgoing_interface log selector which allows you to disable the
1787 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1788 If not running log_selector +smtp_connection the mainlog would be held
1789 open indefinitely after a "too many connections" event, including to a
1790 deleted file after a log rotate. Leave the per net connection logging
1791 leaving it open for efficiency as that will be quickly detected by the
1792 check on the next write.
1794 HS/01 Bug 1671: Fix post transport crash.
1795 Processing the wait-<transport> messages could crash the delivery
1796 process if the message IDs didn't exist for some reason. When
1797 using 'split_spool_directory=yes' the construction of the spool
1798 file name failed already, exposing the same netto behaviour.
1800 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1801 mime_regex ACL conditions.
1803 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1804 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1805 to HELO, local diagnostic string.
1807 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1808 log line to debug. Even when configured with a tls authenticator many
1809 client connections are expected to not authenticate in this way, so
1810 an authenticate fail is not an error.
1812 HS/02 Add the Exim version string to the process info. This way exiwhat
1813 gives some more detail about the running daemon.
1815 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1816 matter for fast-change records such as DNSBLs.
1818 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1819 retry record, even if constant. There may be multiple transports with
1820 different interface settings and the retry behaviour needs to be kept
1823 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1825 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1827 JH/10 Bug 840: fix log_defer_output option of pipe transport
1829 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1830 hosts_randomize. This matters a lot when combined with mua_wrapper.
1832 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1833 ${quote_pgsql:<string>} operator.
1835 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1837 JH/14 Change header file name for freeradius-client. Relevant if compiling
1838 with Radius support; from the Gentoo tree and checked under Fedora.
1840 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1842 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1843 option transport_filter, meaning no filtering.
1845 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1847 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1848 defaults to "*" (all hosts). The variable is now available when not built
1849 with TLS, default unset, mainly to enable keeping the testsuite sane.
1850 If a server certificate is not supplied (via tls_certificate) an error is
1851 logged, and clients will find TLS connections fail on startup. Presumably
1852 they will retry in-clear.
1853 Packagers of Exim are strongly encouraged to create a server certificate
1854 at installation time.
1856 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1857 with the $config_file variable.
1859 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1860 in transport context, after the attempt, and per-recipient. The latter type
1861 is per host attempted. The event data is the error message, and the errno
1862 information encodes the lookup type (A vs. MX) used for the (first) host,
1863 and the trailing two digits of the smtp 4xx response.
1865 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1866 to write to mainlog (or rejectlog, paniclog) in the window between file
1867 creation and permissions/ownership being changed. Particularly affects
1868 installations where exicyclog is run as root, rather than exim user;
1869 result is that the running daemon panics and dies.
1871 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1873 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1874 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1875 "pri" and "weight". Note that the previous implicit priority given by the
1876 list order is no longer honoured.
1878 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1879 for DKIM processing.
1881 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1882 by defining SUPPORT_SOCKS.
1884 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1885 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1886 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1887 variables are renamed to proxy_{local,external}_{address,port}.
1889 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1890 for a build by defining SUPPORT_I18N
1892 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1893 of the query string, and make ${quote_redis:} do that quoting.
1895 JH/29 Move Events support from Experimental to mainline, enabled by default
1896 and removable for a build by defining DISABLE_EVENT.
1898 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1900 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1901 cached by the daemon.
1903 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1904 by defining LOOKUP_REDIS. The libhiredis library is required.
1906 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1907 keys are given for lookup.
1909 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1910 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1911 only supported when built with TLS support. The PolarSSL SHA routines
1912 are still used when the TLS library is too old for convenient support.
1914 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1915 openssl_options), for security. OpenSSL forces this from version 1.1.0
1916 server-side so match that on older versions.
1918 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1919 allocation for $value could be released as the expansion processing
1920 concluded, but leaving the global pointer active for it.
1922 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1923 and to use the domains and local_parts ACL conditions.
1925 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1926 incorrectly not doubled on cutthrough transmission, hence seen as a
1927 body-termination at the receiving system - resulting in truncated mails.
1928 Commonly the sender saw a TCP-level error, and retransmitted the message
1929 via the normal store-and-forward channel. This could result in duplicates
1930 received - but deduplicating mailstores were liable to retain only the
1931 initial truncated version.
1933 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1935 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1937 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1938 we're in there, support oversigning also; bug 1309.
1940 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1942 HS/04 Add support for keep_environment and add_environment options.
1944 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1945 either intentional arithmetic overflow during PRNG, or testing config-
1948 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1949 delivery resulted in actual delivery. Cancel cutthrough before DATA
1952 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1953 rejects a recipient: pass the reject to the originator.
1955 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1956 Many were false-positives and ignorable, but it's worth fixing the
1959 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1960 for the new environment-manipulation done at startup. Move the routines
1961 from being local to tls.c to being global via the os.c file.
1963 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1964 an extract embedded as result-arg for a map, the first arg for extract
1965 is unavailable so we cannot tell if this is a numbered or keyed
1966 extraction. Accept either.
1972 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1975 JH/02 The smtp transport option "multi_domain" is now expanded.
1977 JH/03 The smtp transport now requests PRDR by default, if the server offers
1980 JH/04 Certificate name checking on server certificates, when exim is a client,
1981 is now done by default. The transport option tls_verify_cert_hostnames
1982 can be used to disable this per-host. The build option
1983 EXPERIMENTAL_CERTNAMES is withdrawn.
1985 JH/05 The value of the tls_verify_certificates smtp transport and main options
1986 default to the word "system" to access the system default CA bundle.
1987 For GnuTLS, only version 3.0.20 or later.
1989 JH/06 Verification of the server certificate for a TLS connection is now tried
1990 (but not required) by default. The verification status is now logged by
1991 default, for both outbound TLS and client-certificate supplying inbound
1994 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1997 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1998 Status Notification (bounce) messages are now MIME format per RFC 3464.
1999 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2000 under the control of the dsn_advertise_hosts option, and routers may
2001 have a dsn_lasthop option.
2003 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2004 default, modifiable by a malware= option. The list separator for
2005 the options can now be changed in the usual way. Bug 68.
2007 JH/10 The smtp_receive_timeout main option is now expanded before use.
2009 JH/11 The incoming_interface log option now also enables logging of the
2010 local interface on delivery outgoing connections.
2012 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2013 if the interface and destination host and port all match.
2015 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2018 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2019 Patch from Andrew Lewis.
2021 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2022 now supports optional time-restrictions, weighting, and priority
2023 modifiers per server. Patch originally by <rommer@active.by>.
2025 JH/16 The spamd_address main option now supports a mixed list of local
2026 and remote servers. Remote servers can be IPv6 addresses, and
2027 specify a port-range.
2029 JH/17 Bug 68: The spamd_address main option now supports an optional
2030 timeout value per server.
2032 JH/18 Bug 1581: Router and transport options headers_add/remove can
2033 now have the list separator specified.
2035 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2038 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2041 JH/21 Support for the A6 type of dns record is withdrawn.
2043 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2044 rather than the verbs used.
2046 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2047 from 255 to 1024 chars.
2049 JH/24 Verification callouts now attempt to use TLS by default.
2051 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2052 are generic router options now. The defaults didn't change.
2054 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2055 Original patch from Alexander Shikoff, worked over by JH.
2057 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2058 files if ZCAT_COMMAND is not executable.
2060 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2062 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2064 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2065 Normally benign, it bites when the pair was led to by a CNAME;
2066 modern usage is to not canonicalize the domain to a CNAME target
2067 (and we were inconsistent anyway for A-only vs AAAA+A).
2069 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2071 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2072 when evaluating $sender_host_dnssec.
2074 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2075 $sender_helo_dnssec variable.
2077 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2079 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2081 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2083 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2084 documented as working, but never had. Support all but $spam_report.
2086 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2087 added for tls authenticator.
2089 HS/03 Add perl_taintmode main config option
2095 TL/01 When running the test suite, the README says that variables such as
2096 no_msglog_check are global and can be placed anywhere in a specific
2097 test's script, however it was observed that placement needed to be near
2098 the beginning for it to behave that way. Changed the runtest perl
2099 script to read through the entire script once to detect and set these
2100 variables, reset to the beginning of the script, and then run through
2101 the script parsing/test process like normal.
2103 TL/02 The BSD's have an arc4random API. One of the functions to induce
2104 adding randomness was arc4random_stir(), but it has been removed in
2105 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2106 function when detected.
2108 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2109 cause callback expansion.
2111 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2112 syntax errors in an expansion can be treated as a string instead of
2113 logging or causing an error, due to the internal use of bool_lax
2114 instead of bool when processing it.
2116 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2117 server certificates when making smtp deliveries.
2119 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2121 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2123 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2125 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2126 Merged patch from Sebastian Wiedenroth.
2128 JH/05 Fix results-pipe from transport process. Several recipients, combined
2129 with certificate use, exposed issues where response data items split
2130 over buffer boundaries were not parsed properly. This eventually
2131 resulted in duplicates being sent. This issue only became common enough
2132 to notice due to the introduction of connection certificate information,
2133 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2135 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2136 size buffer was used, resulting in syntax errors when an expansion
2139 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2140 version 3.3.6 or later.
2142 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2143 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2144 both become "event_action", the variables become $event_name, $event_data
2145 and $event_defer_errno. There is a new variable $verify_mode, usable in
2146 routers, transports and related events. The tls:cert event is now also
2147 raised for inbound connections, if the main configuration event_action
2150 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2151 early OCSP support, but no stapling (appears to be less than 1.0.0).
2153 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2154 server certificate names available under the smtp transport option
2155 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2158 JH/10 Time-related extraction expansions from certificates now use the main
2159 option "timezone" setting for output formatting, and are consistent
2160 between OpenSSL and GnuTLS compilations. Bug 1541.
2162 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2163 encoded parameter in the incoming message. Bug 1558.
2165 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2166 include certificate info, eximon was claiming there were spoolfile
2169 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2171 JH/14 Log delivery-related information more consistently, using the sequence
2172 "H=<name> [<ip>]" wherever possible.
2174 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2175 are problematic for Debian distribution, omit them from the release
2178 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2180 JH/16 Fix string representation of time values on 64bit time_t architectures.
2183 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2184 output list separator was used.
2189 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2190 checkers that were complaining about end of non-void function with no
2193 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2194 This was a regression introduced in 4.83 by another bugfix.
2196 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2198 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2199 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2205 TF/01 Correctly close the server side of TLS when forking for delivery.
2207 When a message was received over SMTP with TLS, Exim failed to clear up
2208 the incoming connection properly after forking off the child process to
2209 deliver the message. In some situations the subsequent outgoing
2210 delivery connection happened to have the same fd number as the incoming
2211 connection previously had. Exim would try to use TLS and fail, logging
2212 a "Bad file descriptor" error.
2214 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2215 utilities have not been installed.
2217 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2218 temporary space as the ACL may create new global variables.
2220 TL/01 LDAP support uses per connection or global context settings, depending
2221 upon the detected version of the libraries at build time.
2223 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2224 to extract and use the src ip:port in logging and expansions as if it
2225 were a direct connection from the outside internet. PPv2 support was
2226 updated based on HAProxy spec change in May 2014.
2228 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2230 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2231 Properly escape header and check for NULL return.
2233 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2236 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2238 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2239 characters in header names, implemented as a verify condition.
2240 Contributed by Michael Fischer v. Mollard.
2242 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2243 results permerror and temperror. Previous values are deprecated but
2244 still accepted. In a future release, err_perm and err_temp will be
2245 completely removed, which will be a backward incompatibility if the
2246 ACL tests for either of these two old results. Patch contributed by
2247 user bes-internal on the mailing list.
2249 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2251 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2252 selectors, in both main and reject logs.
2254 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2257 JH/07 Add malware type "sock" for talking to simple daemon.
2259 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2261 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2262 routers/transports under cutthrough routing.
2264 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2265 numbers. Touch up "bool" conditional to keep the same definition.
2267 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2269 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2271 JH/12 Expand items in router/transport headers_add or headers_remove lists
2272 individually rather than the list as a whole. Bug 1452.
2274 Required for reasonable handling of multiple headers_ options when
2275 they may be empty; requires that headers_remove items with embedded
2276 colons must have them doubled (or the list-separator changed).
2278 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2279 view the policy declared in the DMARC record. Currently, $dmarc_status
2280 is a combined value of both the record presence and the result of the
2283 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2285 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2286 dnslookup router and the smtp transport (applying to the forward
2289 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2290 of ldap servers used for a specific lookup. Patch provided by Heiko
2293 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2294 New variable $lookup_dnssec_authenticated for observability.
2296 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2297 Patch submitted by Lars Timman.
2299 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2301 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2302 Requires trusted mode and valid format message id, aborts otherwise.
2303 Patch contributed by Heiko Schlichting.
2305 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2306 certextract with support for various fields. Bug 1358.
2308 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2309 is requested by default, modifiable by smtp transport option
2312 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2313 operate on certificate variables to give certificate fingerprints
2314 Also new ${sha256:cert_variable}.
2316 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2318 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2319 Christian Aistleitner.
2321 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2323 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2324 file. Patch from Wolfgang Breyha.
2326 JH/25 Expand the coverage of the delivery $host and $host_address to
2327 client authenticators run in verify callout. Bug 1476.
2329 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2330 align with daemon_smtp_ports. Bug 72.
2332 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2333 support and error reporting did not work properly.
2335 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2336 and is readable. Patch from Andrew Colin Kissa.
2338 TL/14 Enhance documentation of ${run expansion and how it parses the
2339 commandline after expansion, particularly in the case when an
2340 unquoted variable expansion results in an empty value.
2342 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2344 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2345 by renaming away from it. Addresses GH issue 15, reported by
2348 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2350 TL/15 SECURITY: prevent double expansion in math comparison functions
2351 (can expand unsanitized data). Not remotely exploitable.
2358 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2360 PP/02 Make -n do something, by making it not do something.
2361 When combined with -bP, the name of an option is not output.
2363 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2366 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2367 $sender_host_name and config options to manage this, and basic check
2370 PP/05 DSCP support for outbound connections and control modifier for inbound.
2372 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2373 (Only plugin which currently uses this is kerberos4, which nobody should
2374 be using, but we should make it available and other future plugins might
2375 conceivably use it, even though it would break NAT; stuff *should* be
2376 using channel bindings instead).
2378 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2379 name; added for Sendmail compatibility; requires admin caller.
2380 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2381 just ignore it); requires trusted caller.
2382 Also parse but ignore: -Ac -Am -X<logfile>
2385 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2387 TL/02 Add +smtp_confirmation as a default logging option.
2389 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2390 Patch by Magnus Holmgren from 2007-02-20.
2392 TL/04 Bugzilla 1281 - Spec typo.
2393 Bugzilla 1283 - Spec typo.
2394 Bugzilla 1290 - Spec grammar fixes.
2396 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2398 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2400 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2401 mailing list by by Dmitry Isaikin.
2403 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2405 JH/02 Support "G" suffix to numbers in ${if comparisons.
2407 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2409 NM/01 Bugzilla 1197 - Spec typo
2410 Bugzilla 1196 - Spec examples corrections
2412 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2414 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2415 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2418 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2419 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2421 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2422 "acl {{name}{arg}...}", and optional args on acl condition
2425 JH/05 Permit multiple router/transport headers_add/remove lines.
2427 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2429 JH/07 Avoid using a waiting database for a single-message-only transport.
2430 Performance patch from Paul Fisher. Bugzilla 1262.
2432 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2435 JH/09 Add $headers_added variable, with content from use of ACL modifier
2436 add_header (but not yet added to the message). Bugzilla 199.
2438 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2439 Pulled from Bugzilla 817 by Wolfgang Breyha.
2441 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2443 (nb: this is the same fix as in Exim 4.80.1)
2445 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2448 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2451 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2453 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2454 advertise SMTP AUTH mechanism to us, instead of a generic
2455 protocol violation error. Also, make Exim more robust to bad
2456 data from the Dovecot auth socket.
2458 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2460 When a queue runner is handling a message, Exim first routes the
2461 recipient addresses, during which it prunes them based on the retry
2462 hints database. After that it attempts to deliver the message to
2463 any remaining recipients. It then updates the hints database using
2466 So if a recipient address works intermittently, it can get repeatedly
2467 deferred at routing time. The retry hints record remains fresh so the
2468 address never reaches the final cutoff time.
2470 This is a fairly common occurrence when a user is bumping up against
2471 their storage quota. Exim had some logic in its local delivery code
2472 to deal with this. However it did not apply to per-recipient defers
2473 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2475 This change adds a proper retry rule check during routing so that the
2476 final cutoff time is checked against the message's age. We only do
2477 this check if there is an address retry record and there is not a
2478 domain retry record; this implies that previous attempts to handle
2479 the address had the retry_use_local_parts option turned on. We use
2480 this as an approximation for the destination being like a local
2481 delivery, as in LMTP.
2483 I suspect this new check makes the old local delivery cutoff check
2484 redundant, but I have not verified this so I left the code in place.
2486 TF/02 Correct gecos expansion when From: is a prefix of the username.
2488 Test 0254 submits a message to Exim with the header
2492 When I ran the test suite under the user fanf2, Exim expanded
2493 the header to contain my full name, whereas it should have added
2494 a Resent-Sender: header. It erroneously treats any prefix of the
2495 username as equal to the username.
2497 This change corrects that bug.
2499 GF/01 DCC debug and logging tidyup
2500 Error conditions log to paniclog rather than rejectlog.
2501 Debug lines prefixed by "DCC: " to remove any ambiguity.
2503 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2505 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2506 Bug spotted by Jeremy Harris; was flawed since initial commit.
2507 Would have resulted in OCSP responses post-SNI triggering an Exim
2508 NULL dereference and crash.
2510 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2512 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2513 Bug detection, analysis and fix by Samuel Thibault.
2514 Bugzilla 1331, Debian bug #698092.
2516 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2518 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2519 Server implementation by Todd Lyons, client by JH.
2520 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2521 config variable "prdr_enable" controls whether the server
2522 advertises the facility. If the client requests PRDR a new
2523 acl_data_smtp_prdr ACL is called once for each recipient, after
2524 the body content is received and before the acl_smtp_data ACL.
2525 The client is controlled by both of: a hosts_try_prdr option
2526 on the smtp transport, and the server advertisement.
2527 Default client logging of deliveries and rejections involving
2528 PRDR are flagged with the string "PRDR".
2530 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2531 fclose(). Diagnosis by Todd Lyons.
2533 PP/17 Update configure.default to handle IPv6 localhost better.
2534 Patch by Alain Williams (plus minor tweaks).
2537 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2538 This is now consistent with GnuTLS, and is now documented: the
2539 previous undocumented portable approach to treating the option as
2540 unset was to force an expansion failure. That still works, and
2541 an empty string is now equivalent.
2543 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2544 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2545 not performing validation itself.
2547 PP/20 Added force_command boolean option to pipe transport.
2548 Patch from Nick Koston, of cPanel Inc.
2550 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2553 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2555 PP/21 Fix eximon continuous updating with timestamped log-files.
2556 Broken in a format-string cleanup in 4.80, missed when I repaired the
2557 other false fix of the same issue.
2558 Report and fix from Heiko Schlichting.
2561 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2562 Report from Prashanth Katuri.
2564 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2565 It's SecureTransport, so affects any MacOS clients which use the
2566 system-integrated TLS libraries, including email clients.
2568 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2569 using a MIME ACL for non-SMTP local injection.
2570 Report and assistance in diagnosis by Warren Baker.
2572 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2574 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2576 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2577 last id that failed so it may be referenced in subsequent ACL's.
2579 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2582 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2583 ldap library initialization, allowing self-signed CA's to be
2584 used. Also properly sets require_cert option later in code by
2585 using NULL (global ldap config) instead of ldap handle (per
2586 session). Bug diagnosis and testing by alxgomz.
2588 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2589 the src/util/ subdirectory.
2591 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2592 renamed to Transport Post Delivery Action by Jeremy Harris, as
2595 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2596 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2597 redis_servers = needs to be configured which will be used by the redis
2598 lookup. Patch from Warren Baker, of The Packet Hub.
2600 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2602 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2603 hostname or reverse DNS when processing a host list. Used suggestions
2604 from multiple comments on this bug.
2606 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2608 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2609 Missed a few lines, added it to make the runtest require no keyboard
2612 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2613 contains upper case chars. Make router use caseful_local_part.
2615 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2616 support when GnuTLS has been built with p11-kit.
2622 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2624 This, or similar/improved, will also be change PP/11 of 4.82.
2630 PP/01 Handle short writes when writing local log-files.
2631 In practice, only affects FreeBSD (8 onwards).
2632 Bugzilla 1053, with thanks to Dmitry Isaikin.
2634 NM/01 Bugzilla 949 - Documentation tweak
2636 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2639 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2641 PP/02 Implemented gsasl authenticator.
2643 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2645 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2646 `pkg-config foo` for cflags/libs.
2648 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2649 with rest of GSASL and with heimdal_gssapi.
2651 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2652 `pkg-config foo` for cflags/libs for the TLS implementation.
2654 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2655 properties get this fed in as external SSF. A number of robustness
2656 and debugging improvements to the cyrus_sasl authenticator.
2658 PP/08 cyrus_sasl server now expands the server_realm option.
2660 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2661 Patch by Jeremy Harris.
2663 PP/10 Added dbmjz lookup type.
2665 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2667 PP/12 MAIL args handles TAB as well as SP, for better interop with
2668 non-compliant senders.
2669 Analysis and variant patch by Todd Lyons.
2671 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2672 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2673 Patch from Dirk Mueller <dmueller@suse.com>
2675 PP/13 tls_peerdn now print-escaped for spool files.
2676 Observed some $tls_peerdn in wild which contained \n, which resulted
2677 in spool file corruption.
2679 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2680 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2681 or write after TLS renegotiation, which otherwise led to messages
2684 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2685 as a tracking header (ie: a signed header comes before the signature).
2686 Patch from Wolfgang Breyha.
2688 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2689 comma-sep list; embedded commas doubled.
2691 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2693 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2695 Report and patch from Dmitry Banschikov.
2697 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2698 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2699 failures. We appear to now support TLS1.1+ with Exim.
2701 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2702 lets Exim select keys and certificates based upon TLS SNI from client.
2703 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2704 before an outbound SMTP session. New log_selector, +tls_sni.
2706 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2707 NULL dereference. Report and patch from Alun Jones.
2709 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2710 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2711 Not seeing resolver debug output on NetBSD, but suspect this is a
2712 resolver implementation change.
2714 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2715 Left warnings. Added "eximon gdb" invocation mode.
2717 PP/21 Defaulting "accept_8bitmime" to true, not false.
2719 PP/22 Added -bw for inetd wait mode support.
2721 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2722 locate the relevant includes and libraries. Made this the default.
2724 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2725 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2727 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2728 This may cause build issues on older platforms.
2730 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2731 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2732 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2733 Added SNI support via GnuTLS too.
2734 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2736 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2738 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2739 Applied second patch from Janne, implementing suggestion to default
2740 multiple-strings-in-record handling to match SPF spec.
2742 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2744 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2745 read-only, out of scope).
2746 Patch from Wolfgang Breyha, report from Stuart Northfield.
2748 PP/29 Fix three issues highlighted by clang analyser static analysis.
2749 Only crash-plausible issue would require the Cambridge-specific
2750 iplookup router and a misconfiguration.
2751 Report from Marcin Mirosław.
2753 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2755 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2756 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2757 As part of this, removing so much warning spew let me fix some minor
2758 real issues in debug logging.
2760 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2761 assignment on my part. Fixed.
2763 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2764 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2765 Janne Snabb (who went above and beyond: thank you).
2767 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2768 string otherwise requires a connection and a bunch more work and it's
2769 relatively easy to get wrong. Should also expose TLS library linkage
2772 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2773 64-bit ${eval} (JH/03).
2775 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2776 GNU libc to support some of the 64-bit stuff, should not lead to
2777 conflicts. Defined before os.h is pulled in, so if a given platform
2778 needs to override this, it can.
2780 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2781 protection layer was required, which is not implemented.
2782 Bugzilla 1254, patch from Wolfgang Breyha.
2784 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2785 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2786 tls_dhparam take prime identifiers. Also unbreak combination of
2787 OpenSSL+DH_params+TLSSNI.
2789 PP/39 Disable SSLv2 by default in OpenSSL support.
2795 PP/01 Solaris build fix for Oracle's LDAP libraries.
2796 Bugzilla 1109, patch from Stephen Usher.
2798 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2800 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2803 TF/02 Fix a couple more cases where we did not log the error message
2804 when unlink() failed. See also change 4.74-TF/03.
2806 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2807 lock up or crash if it happened to be inside a call to libc when it
2808 got a SIGUSR1 from exiwhat.
2810 The SIGUSR1 handler appends the current process status to the process
2811 log which is later printed by exiwhat. It used to use the general
2812 purpose logging code to do this, but several functions it calls are
2813 not safe for signals.
2815 The new output code in the SIGUSR1 handler is specific to the process
2816 log, and simple enough that it's easy to inspect for signal safety.
2817 Removing some special cases also simplifies the general logging code.
2818 Removing the spurious timestamps from the process log simplifies
2821 TF/04 Improved ratelimit ACL condition.
2823 The /noupdate option has been deprecated in favour of /readonly which
2824 has clearer semantics. The /leaky, /strict, and /readonly update modes
2825 are mutually exclusive. The update mode is no longer included in the
2826 database key; it just determines when the database is updated. (This
2827 means that when you upgrade Exim will forget old rate measurements.)
2829 Exim now checks that the per_* options are used with an update mode that
2830 makes sense for the current ACL. For example, when Exim is processing a
2831 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2832 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2833 must specify per_mail/readonly. If you omit the update mode it defaults to
2834 /leaky where that makes sense (as before) or /readonly where required.
2836 The /noupdate option is now undocumented but still supported for
2837 backwards compatibility. It is equivalent to /readonly except that in
2838 ACLs where /readonly is required you may specify /leaky/noupdate or
2839 /strict/noupdate which are treated the same as /readonly.
2841 A useful new feature is the /count= option. This is a generalization
2842 of the per_byte option, so that you can measure the throughput of other
2843 aggregate values. For example, the per_byte option is now equivalent
2844 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2846 The per_rcpt option has been generalized using the /count= mechanism
2847 (though it's more complicated than the per_byte equivalence). When it is
2848 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2849 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2850 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2851 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2852 Note that using per_rcpt with a non-readonly update mode in more than
2853 one ACL will cause the recipients to be double-counted. (The per_mail
2854 and per_byte options don't have this problem.)
2856 The handling of very low rates has changed slightly. If the computed rate
2857 is less than the event's count (usually one) then this event is the first
2858 after a long gap. In this case the rate is set to the same as this event's
2859 count, so that the first message of a spam run is counted properly.
2861 The major new feature is a mechanism for counting the rate of unique
2862 events. The new per_addr option counts the number of different
2863 recipients that someone has sent messages to in the last time period. It
2864 behaves like per_rcpt if all the recipient addresses are different, but
2865 duplicate recipient addresses do not increase the measured rate. Like
2866 the /count= option this is a general mechanism, so the per_addr option
2867 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2868 example, measure the rate that a client uses different sender addresses
2869 with the options per_mail/unique=$sender_address. There are further
2870 details in the main documentation.
2872 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2874 TF/06 Removed a few PCRE remnants.
2876 TF/07 Automatically extract Exim's version number from tags in the git
2877 repository when doing development or release builds.
2879 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2880 Bugzilla 879. Patch from Paul Fisher.
2882 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2883 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2886 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2888 PP/05 Variable $av_failed, true if the AV scanner deferred.
2889 Bugzilla 1078. Patch from John Horne.
2891 PP/06 Stop make process more reliably on build failure.
2892 Bugzilla 1087. Patch from Heiko Schlittermann.
2894 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2895 Bugzilla 1089. Patch from Heiko Schlittermann.
2897 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2898 Bugzilla 1131. Patch from Holger Weiß.
2900 PP/09 Handle IPv6 addresses with SPF.
2901 Bugzilla 860. Patch from Wolfgang Breyha.
2903 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2905 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2908 PP/11 match_* no longer expand right-hand-side by default.
2909 New compile-time build option, EXPAND_LISTMATCH_RHS.
2910 New expansion conditions, "inlist", "inlisti".
2912 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2914 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2916 PP/14 fix log_write() format string regression from TF/03.
2917 Bugzilla 1152. Patch from Dmitry Isaikin.
2923 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2925 PP/02 Harmonised TLS library version reporting; only show if debugging.
2926 Layout now matches that introduced for other libraries in 4.74 PP/03.
2928 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2930 PP/04 New "dns_use_edns0" global option.
2932 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2935 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2936 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2938 TK/01 Updated PolarSSL code to 0.14.2.
2939 Bugzilla 1097. Patch from Andreas Metzler.
2941 PP/07 Catch divide-by-zero in ${eval:...}.
2942 Fixes bugzilla 1102.
2944 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2947 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2948 format-string attack -- SECURITY: remote arbitrary code execution.
2950 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2951 time unintentionally subject to list matching rules, letting the header
2952 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2953 arbitrary string expansion). This allowed for information disclosure.
2955 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2956 INT_MIN/-1 -- value coerced to INT_MAX.
2962 NM/01 Workaround for PCRE version dependency in version reporting
2965 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2966 This fixes portability to compilers other than gcc, notably
2967 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2969 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2970 makefiles for portability to HP-UX and POSIX correctness.
2972 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2973 Also via indented variable definition in the Makefile.
2974 (Debugging by Oliver Heesakkers).
2976 PP/02 Restore caching of spamd results with expanded spamd_address.
2977 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2979 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2980 Improves build reliability. Fix from: Frank Elsner
2982 NM/02 Fix wide character breakage in the rfc2047 coding
2983 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2985 NM/03 Allow underscore in dnslist lookups
2986 Fixes bug 1026. Patch from Graeme Fowler
2988 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2989 Code patches from Adam Ciarcinski of NetBSD.
2991 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2994 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2995 is logged, to avoid truncation. Patch from John Horne.
2997 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2998 Patch from Jakob Hirsch.
3000 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3001 SQL string expansion failure details.
3002 Patch from Andrey Oktyabrski.
3004 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3005 Patch from Simon Arlott.
3007 PP/09 New lookups functionality failed to compile on old gcc which rejects
3008 extern declarations in function scope.
3009 Patch from Oliver Fleischmann
3011 PP/10 Use sig_atomic_t for flags set from signal handlers.
3012 Check getgroups() return and improve debugging.
3013 Fixed developed for diagnosis in bug 927 (which turned out to be
3016 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3017 Patch from Mark Zealey.
3019 PP/12 Bugzilla 1056: Improved spamd server selection.
3020 Patch from Mark Zealey.
3022 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3023 Based on patch from Heiko Schlittermann.
3025 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3026 Patch from Uwe Doering, sign-off by Michael Haardt.
3028 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3029 variable declaration deep within a block. Bug and patch from
3032 PP/15 lookups-Makefile IRIX compatibility coercion.
3034 PP/16 Make DISABLE_DKIM build knob functional.
3036 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3037 Patch by Simon Arlott
3039 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3040 variable argument macros. Our copy now differs from upstream.
3046 TF/01 Failure to get a lock on a hints database can have serious
3047 consequences so log it to the panic log.
3049 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3050 controlled using the smtp_confirmation log selector.
3052 TF/03 Include the error message when we fail to unlink a spool file.
3054 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3055 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3056 for maintaining out-of-tree patches for some time.
3058 PP/01 Bugzilla 139: Documentation and portability issues.
3059 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3060 Handle per-OS dynamic-module compilation flags.
3062 PP/02 Let /dev/null have normal permissions.
3063 The 4.73 fixes were a little too stringent and complained about the
3064 permissions on /dev/null. Exempt it from some checks.
3065 Reported by Andreas M. Kirchwitz.
3067 PP/03 Report version information for many libraries, including
3068 Exim version information for dynamically loaded libraries. Created
3069 version.h, now support a version extension string for distributors
3070 who patch heavily. Dynamic module ABI change.
3072 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3073 privilege escalation vulnerability whereby the Exim run-time user
3074 can cause root to append content of the attacker's choosing to
3077 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3080 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3081 If dropping privileges for untrusted macros, we disabled normal logging
3082 on the basis that it would fail; for the Exim run-time user, this is not
3083 the case, and it resulted in successful deliveries going unlogged.
3084 Fixed. Reported by Andreas Metzler.
3090 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3091 only prepend for the Resent-* case. Fixes regression introduced in
3092 Exim 4.70 by NM/22 for Bugzilla 607.
3094 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3095 increasing numbers of administrators be bitten by this.
3097 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3099 PP/03 Bugzilla 994: added openssl_options main configuration option.
3101 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3103 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3105 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3107 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3108 without a peer certificate, leading to a segfault because of an
3109 assumption that peers always have certificates. Be a little more
3110 paranoid. Problem reported by Martin Tscholak.
3112 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3113 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3114 NB: ClamAV planning to remove STREAM in "middle of 2010".
3115 CL also introduces -bmalware, various -d+acl logging additions and
3116 more caution in buffer sizes.
3118 PP/09 Implemented reverse_ip expansion operator.
3120 PP/10 Bugzilla 937: provide a "debug" ACL control.
3122 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3124 PP/12 Bugzilla 973: Implement --version.
3126 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3128 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3130 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3132 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3133 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3134 ignore trailing whitespace.
3136 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3138 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3141 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3142 Notification from Dr Andrew Aitchison.
3144 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3145 ExtendedDetectionInfo response format.
3146 Notification from John Horne.
3148 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3151 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3152 XSL and documented dependency on system catalogs, with examples of how
3155 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3158 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3159 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3160 configuration file which is writeable by the Exim user or group.
3162 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3163 of configuration files to cover files specified with the -C option if
3164 they are going to be used with root privileges, not just the default
3167 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3168 option (effectively making it always true).
3170 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3171 files to be used while preserving root privileges.
3173 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3174 that rogue child processes cannot use them.
3176 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3177 run-time user, instead of root.
3179 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3180 Exim run-time user without dropping privileges.
3182 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3183 result string, instead of calling string_vformat() twice with the same
3186 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3187 for other users. Others should always drop root privileges if they use
3188 -C on the command line, even for a whitelisted configure file.
3190 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3192 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3198 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3199 $data_path, and $header_path variables; fixed documentation bugs and
3202 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3203 exipick to access non-standard spools, including the "frozen" queue
3206 NM/01 Bugzilla 965: Support mysql stored procedures.
3207 Patch from Alain Williams
3209 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3211 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3212 Patch from Andreas Metzler
3214 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3215 Patch from Kirill Miazine
3217 NM/05 Bugzilla 671: Added umask to procmail example.
3219 JJ/03 installed exipick 20100323.0, fixing doc bug
3221 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3222 directory. Notification and patch from Dan Rosenberg.
3224 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3226 TK/02 Improve log output when DKIM signing operation fails.
3228 MH/01 Treat the transport option dkim_domain as a colon separated
3229 list, not as a single string, and sign the message with each element,
3230 omitting multiple occurences of the same signer.
3232 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3233 Bugzilla 985, 986. Patch by Simon Arlott
3235 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3236 Patch by Simon Arlott
3238 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3239 MBX locking. Notification from Dan Rosenberg.
3245 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3247 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3249 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3251 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3253 NM/04 Bugzilla 915: Flag broken perl installation during build.
3259 TK/01 Added patch by Johannes Berg that expands the main option
3260 "spamd_address" if it starts with a dollar sign.
3262 TK/02 Write list of recipients to X-Envelope-Sender header when building
3263 the mbox-format spool file for content scanning (suggested by Jakob
3266 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3267 (http://www.dcc-servers.net/) support via dccifd. Activated by
3268 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3270 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3271 by Mark Daniel Reidel <mr@df.eu>.
3273 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3274 When building exim an external PCRE library is now needed -
3275 PCRE is a system library on the majority of modern systems.
3276 See entry on PCRE_LIBS in EDITME file.
3278 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3279 conversation. Added nologin parameter to request.
3280 Patch contributed by Kirill Miazine.
3282 TF/01 Do not log submission mode rewrites if they do not change the address.
3284 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3286 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3287 log files in place. Contributed by Roberto Lima.
3289 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3291 TF/03 Bugzilla 615: When checking the local_parts router precondition
3292 after a local_part_suffix or local_part_prefix option, Exim now
3293 does not use the address's named list lookup cache, since this
3294 contains cached lookups for the whole local part.
3296 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3297 Robert Millan. Documentation is in experimental-spec.txt.
3299 TF/04 Bugzilla 668: Fix parallel build (make -j).
3301 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3303 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3304 Patch provided by Jan Srzednicki.
3306 TF/05 Leading white space used to be stripped from $spam_report which
3307 wrecked the formatting. Now it is preserved.
3309 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3310 that they are available at delivery time.
3312 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3314 TF/08 TLS error reporting now respects the incoming_interface and
3315 incoming_port log selectors.
3317 TF/09 Produce a more useful error message if an SMTP transport's hosts
3318 setting expands to an empty string.
3320 NM/06 Bugzilla 744: EXPN did not work under TLS.
3321 Patch provided by Phil Pennock.
3323 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3324 Patch provided by Richard Godbee.
3326 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3327 acl_smtp_notquit, added index entry.
3329 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3330 Patch provided by Eugene Bujak.
3332 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3333 accept(). Patch provided by Maxim Dounin.
3335 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3336 Patch provided by Phil Pennock.
3338 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3340 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3341 Patch provided by Brad "anomie" Jorsch.
3343 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3344 Patch provided by Dean Brooks.
3346 TK/05 Add native DKIM support (does not depend on external libraries).
3348 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3349 Patch provided by Graeme Fowler.
3351 NM/16 Bugzilla 851: Documentation example syntax fix.
3353 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3355 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3358 NM/19 Bugzilla 745: TLS version reporting.
3359 Patch provided by Phil Pennock.
3361 NM/20 Bugzilla 167: bool: condition support.
3362 Patch provided by Phil Pennock.
3364 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3365 clients. Patch provided by Phil Pennock.
3367 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3368 Patch provided by Brad "anomie" Jorsch.
3370 NM/23 Bugzilla 687: Fix misparses in eximstats.
3371 Patch provided by Heiko Schlittermann.
3373 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3374 Patch provided by Heiko Schlittermann.
3376 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3377 plus update to original patch.
3379 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3381 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3382 Patch provided by David Brownlee.
3384 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3386 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3388 NM/30 Bugzilla 888: TLS documentation bugfixes.
3390 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3392 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3393 Unlike the original bugzilla I have changed all shell scripts in src tree.
3395 NM/33 Bugzilla 898: Transport filter timeout fix.
3396 Patch by Todd Rinaldo.
3398 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3399 Patch by Serge Demonchaux.
3401 NM/35 Bugzilla 39: Base64 decode bug fixes.
3402 Patch by Jakob Hirsch.
3404 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3406 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3408 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3410 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3416 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3417 ALT-N's libdkim that I have put here:
3418 http://duncanthrax.net/exim-experimental/
3420 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3421 were called 'true' and it seems that C99 defines that as a reserved
3422 keyword to be used with 'bool' variable types. That means you could
3423 not include C99-style headers which use bools without triggering
3424 build errors in sieve.c.
3426 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3427 as mailq or other aliases. Changed the --help handling significantly
3428 to do whats expected. exim_usage() emits usage/help information.
3430 SC/01 Added the -bylocaldomain option to eximstats.
3432 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3434 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3436 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3442 PH/01 Another patch from the Sieve maintainer.
3444 PH/02 When an IPv6 address is converted to a string for single-key lookup
3445 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3446 dots are used instead of colons so that keys in lsearch files need not
3447 contain colons. This was done some time before quoting was made available
3448 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3449 (notated using the quote facility) so as to distinguish them from IPv4
3450 keys. This meant that lookups for IP addresses in host lists did not work
3451 for iplsearch lookups.
3453 This has been fixed by arranging for IPv6 addresses to be expressed with
3454 colons if the lookup type is iplsearch. This is not incompatible, because
3455 previously such lookups could never work.
3457 The situation is now rather anomalous, since one *can* have colons in
3458 ordinary lsearch keys. However, making the change in all cases is
3459 incompatible and would probably break a number of configurations.
3461 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3464 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3465 conversion specification without a maximum field width, thereby enabling
3466 a rogue spamd server to cause a buffer overflow. While nobody in their
3467 right mind would setup Exim to query an untrusted spamd server, an
3468 attacker that gains access to a server running spamd could potentially
3469 exploit this vulnerability to run arbitrary code as the Exim user.
3471 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3472 $primary_hostname instead of what libspf2 thinks the hosts name is.
3474 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3475 a directory entry by the name of the lookup key. Previously, if a
3476 symlink pointed to a non-existing file or a file in a directory that
3477 Exim lacked permissions to read, a lookup for a key matching that
3478 symlink would fail. Now it is enough that a matching directory entry
3479 exists, symlink or not. (Bugzilla 503.)
3481 PH/03 The body_linecount and body_zerocount variables are now exported in the
3484 PH/04 Added the $dnslist_matched variable.
3486 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3487 This means they are set thereafter only if the connection becomes
3490 PH/06 Added the client_condition to authenticators so that some can be skipped
3491 by clients under certain conditions.
3493 PH/07 The error message for a badly-placed control=no_multiline_responses left
3494 "_responses" off the end of the name.
3496 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3498 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3499 (without spaces) instead of just copying the configuration text.
3501 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3503 PH/11 Added $max_received_linelength.
3505 PH/12 Added +ignore_defer and +include_defer to host lists.
3507 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3508 way in which PCRE > 7.0 is built.
3510 PH/14 Implemented queue_only_load_latch.
3512 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3513 MAIL command. The effect was to mangle the value on 64-bit systems.
3515 PH/16 Another patch from the Sieve maintainer.
3517 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3519 PH/18 If a system quota error occurred while trying to create the file for
3520 a maildir delivery, the message "Mailbox is full" was not appended to the
3521 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3522 only to a quota excession during the actual writing of the file.
3524 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3525 characters?) which causes problems in log lines. The DN values are now
3526 passed through string_printing() before being added to log lines.
3528 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3529 and InterBase are left for another time.)
3531 PH/21 Added message_body_newlines option.
3533 PH/22 Guard against possible overflow in moan_check_errorcopy().
3535 PH/23 POSIX allows open() to be a macro; guard against that.
3537 PH/24 If the recipient of an error message contained an @ in the local part
3538 (suitably quoted, of course), incorrect values were put in $domain and
3539 $local_part during the evaluation of errors_copy.
3545 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3546 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3549 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3550 issue a MAIL command.
3552 PH/02 In an ACL statement such as
3554 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3556 if a client was not listed at all, or was listed with a value other than
3557 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3558 the condition was not true (as it should be), so access was not denied.
3559 The bug was that the ! inversion was incorrectly passed on to the second
3560 item. This has been fixed.
3562 PH/03 Added additional dnslists conditions == and =& which are different from
3563 = and & when the dns lookup returns more than one IP address.
3565 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3566 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3568 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3569 FSYNC, which compiles an option called disable_fsync that allows for
3570 bypassing fsync(). The documentation is heavily laced with warnings.
3572 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3574 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3575 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3576 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3577 including adding "make clean"; (3) Added -fPIC when compiling the test
3578 dynamically loaded module, to get rid of a warning.
3580 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3581 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3582 = 0s. The bug is otherwise harmless.
3584 PH/07 There was a bug in the dovecot authenticator such that the value of
3585 $auth1 could be overwritten, and so not correctly preserved, after a
3586 successful authentication. This usually meant that the value preserved by
3587 the server_setid option was incorrect.
3589 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3591 PH/09 Installed PCRE release 7.0.
3593 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3594 run for batched SMTP input. It is now run at the start of every message
3595 in the batch. While fixing this I discovered that the process information
3596 (output by running exiwhat) was not always getting set for -bs and -bS
3597 input. This is fixed, and it now also says "batched" for BSMTP.
3599 PH/11 Added control=no_pipelining.
3601 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3602 patch, slightly modified), and move the expansion of helo_data till after
3603 the connection is made in the smtp transport (so it can use these
3606 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3608 PH/14 Added log_selector = +pid.
3610 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3612 PH/16 Add ${if forany and ${if forall.
3614 PH/17 Added dsn_from option to vary the From: line in DSNs.
3616 PH/18 Flush SMTP output before performing a callout, unless control =
3617 no_callout_flush is set.
3619 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3620 was true (the default) a successful delivery failed to delete the retry
3621 item, thus causing premature timeout of the address. The bug is now
3624 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3626 PH/21 Long custom messages for fakedefer and fakereject are now split up
3627 into multiline responses in the same way that messages for "deny" and
3628 other ACL rejections are.
3630 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3631 with slight modification.
3633 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3634 draft, changing the syntax and factoring some duplicate code".
3636 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3637 for deliveries of the second and subsequent messages over the same SMTP
3640 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3641 ${reduce, with only minor "tidies".
3643 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3645 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3646 expansion side effects.
3648 PH/27 When a message times out after an over-quota error from an Exim-imposed
3649 quota, the bounce message says "mailbox is full". This message was not
3650 being given when it was a system quota that was exceeded. It now should
3653 MH/03 Made $recipients available in local_scan(). local_scan() already has
3654 better access to the recipient list through recipients_list[], but
3655 $recipients can be useful in postmaster-provided expansion strings.
3657 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3658 in the case of a MAIL command with additional options following the
3659 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3660 were accidentally chopped off.
3662 PH/29 SMTP synchronization checks are implemented when a command is read -
3663 there is a check that no more input is waiting when there shouldn't be
3664 any. However, for some commands, a delay in an ACL can mean that it is
3665 some time before the response is written. In this time, more input might
3666 arrive, invalidly. So now there are extra checks after an ACL has run for
3667 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3668 pipelining has not been advertised.
3670 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3672 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3673 returned in the SMTP response when smtp_return_error_details was false.
3674 This has been fixed.
3676 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3677 instead of the C I/O that was originally supplied, because problems were
3678 reported on Solaris.
3680 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3681 Exim which did not show up earlier: it was assuming that a call to
3682 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3683 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3684 was a macro that became an assignment, so it seemed to work. This has
3685 changed to a proper function call with a void return, hence the compile
3686 error. Exim's code has been fixed.
3688 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3691 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3693 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3695 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3696 because it turns out to be too big or there is a timeout) there is no
3697 "Completed" line in the log. When some messages of this type were
3698 selected by exigrep, they were listed as "not completed". Others were
3699 picked up by some special patterns. I have improved the selection
3700 criteria to be more general.
3702 PH/38 The host_find_failed option in the manualroute router can now be set
3703 to "ignore", to completely ignore a host whose IP address cannot be
3704 found. If all hosts are ignored, the behaviour is controlled by the new
3705 host_all_ignored option.
3707 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3708 homing or because of multiple MX records with /mx) generated more than
3709 one IP address, and the following item turned out to be the local host,
3710 all the secondary addresses of the first item were incorrectly removed
3711 from the list, along with the local host and any following hosts (which
3712 is what is supposed to happen).
3714 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3715 whoever called Exim into the -H file. In the case of the daemon it was
3716 behaving confusingly. When first started, it used values for whoever
3717 started the daemon, but after a SIGHUP it used the Exim user (because it
3718 calls itself on a restart). I have changed the code so that it now always
3721 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3722 message are rejected with the same error (e.g. no authentication or bad
3723 sender address), and a DATA command is nevertheless sent (as can happen
3724 with PIPELINING or a stupid MUA), the error message that was given to the
3725 RCPT commands is included in the rejection of the DATA command. This is
3726 intended to be helpful for MUAs that show only the final error to their
3729 PH/42 Another patch from the Sieve maintainer.
3731 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3732 Eximstats - Fixed some broken HTML links and added missing column headers
3734 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3735 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3737 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3739 PH/43 Yet another patch from the Sieve maintainer.
3741 PH/44 I found a way to check for a TCP/IP connection going away before sending
3742 the response to the final '.' that terminates a message, but only in the
3743 case where the client has not sent further data following the '.'
3744 (unfortunately, this is allowed). However, in many cases there won't be
3745 any further data because there won't be any more messages to send. A call
3746 to select() can be used: if it shows that the input is "ready", there is
3747 either input waiting, or the socket has been closed. An attempt to read
3748 the next input character can distinguish the two cases. Previously, Exim
3749 would have sent an OK response which the client would never have see.
3750 This could lead to message repetition. This fix should cure that, at
3751 least in a lot of common cases.
3753 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3754 advertised in response to EHLO.
3760 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3761 fixed by 4.65/MH/01 (is this a record?) are fixed:
3763 (i) An empty string was always treated as zero by the numeric comparison
3764 operators. This behaviour has been restored.
3766 (ii) It is documented that the numeric comparison operators always treat
3767 their arguments as decimal numbers. This was broken in that numbers
3768 starting with 0 were being interpreted as octal.
3770 While fixing these problems I realized that there was another issue that
3771 hadn't been noticed. Values of message_size_limit (both the global option
3772 and the transport option) were treated as octal if they started with 0.
3773 The documentation was vague. These values are now always treated as
3774 decimal, and I will make that clear in the documentation.
3780 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3781 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3784 MH/01 Don't check that the operands of numeric comparison operators are
3785 integers when their expansion is in "skipping" mode (fixes bug
3786 introduced by 4.64-PH/07).
3788 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3789 child addresses, Exim now panics and dies. Previously, because the count
3790 is held in a short int, deliveries were likely to be lost. As such a
3791 large number of recipients for a single message is ridiculous
3792 (performance will be very, very poor), I have chosen to impose a limit
3793 rather than extend the field.
3799 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3800 leftover -K file (the existence of which was triggered by #402).
3801 While we were at it, introduced process PID as part of the -K
3802 filename. This should rule out race conditions when creating
3805 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3806 processing considerably. Previous code took too long for large mails,
3807 triggering a timeout which in turn triggers #401.
3809 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3810 in the DK code in transports.c. sendfile() is not really portable,
3811 hence the _LINUX specificness.
3813 TF/01 In the add_headers option to the mail command in an Exim filter,
3814 there was a bug that Exim would claim a syntax error in any
3815 header after the first one which had an odd number of characters
3818 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3819 callout verification, Exim cached a "reject" for the entire domain. This
3820 is correct for most verifications, but it is not correct for a recipient
3821 verification with use_sender or use_postmaster set, because in that case
3822 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3823 case of MAIL FROM:<> rejection from other early rejections (e.g.
3824 rejection of HELO). When verifying a recipient using a non-null MAIL
3825 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3826 Whatever the result of the callout, the value of the domain cache is
3827 left unchanged (for any other kind of callout, getting as far as trying
3828 RCPT means that the domain itself is ok).
3830 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3833 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3834 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3835 session, the daemon ignores ECONNECT errors and logs others; it now
3836 ignores EPIPE as well.
3838 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3839 (quoted-printable decoding).
3841 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3842 later the small subsequent patch to fix an introduced bug.
3844 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3846 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3848 PH/08 An error is now given if message_size_limit is specified negative.
3850 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3851 to be given (somewhat) arbitrary names.
3853 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3856 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3857 miscellaneous code fixes
3859 PH/10 Added the log_reject_target ACL modifier to specify where to log
3862 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3863 hostname. This is wrong, because it relates to the incoming message (and
3864 probably the interface on which it is arriving) and not to the outgoing
3865 callout (which could be using a different interface). This has been
3866 changed to use the value of the helo_data option from the smtp transport
3867 instead - this is what is used when a message is actually being sent. If
3868 there is no remote transport (possible with a router that sets up host
3869 addresses), $smtp_active_hostname is used.
3871 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3872 tweaks were necessary in order to get it to work (see also 21 below):
3873 (a) The code assumed that strncpy() returns a negative number on buffer
3874 overflow, which isn't the case. Replaced with Exim's string_format()
3876 (b) There were several signed/unsigned issues. I just did the minimum
3877 hacking in of casts. There is scope for a larger refactoring.
3878 (c) The code used strcasecmp() which is not a standard C function.
3879 Replaced with Exim's strcmpic() function.
3880 (d) The code set only $1; it now sets $auth1 as well.
3881 (e) A simple test gave the error "authentication client didn't specify
3882 service in request". It would seem that Dovecot has changed its
3883 interface. Fortunately there's a specification; I followed it and
3884 changed what the client sends and it appears to be working now.
3886 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3889 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3890 address A is aliased to B and C, where B exists and C does not. Without
3891 -v the output is "A verified" because verification stops after a
3892 successful redirection if more than one address is generated. However,
3893 with -v the child addresses are also verified. Exim was outputting "A
3894 failed to verify" and then showing the successful verification for C,
3895 with its parentage. It now outputs "B failed to verify", showing B's
3896 parentage before showing the successful verification of C.
3898 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3899 look up a TXT record in a specific list after matching in a combined
3902 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3903 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3904 they consult the DNS. I had assumed they would set it the way they
3905 wanted; and indeed my experiments on Linux seem to show that in some
3906 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3907 To be on the safe side, however, I have now made the interface to
3908 host_find_byname() similar to host_find_bydns(), with an argument
3909 containing the DNS resolver options. The host_find_byname() function now
3910 sets these options at its start, just as host_find_bydns() does. The smtp
3911 transport options dns_qualify_single and dns_search_parents are passed to
3912 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3913 of host_find_byname() use the default settings of RES_DEFNAMES
3914 (qualify_single) but not RES_DNSRCH (search_parents).
3916 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3917 spool_read_header() do less string testing, by means of a preliminary
3918 switch on the second character of optional "-foo" lines. (This is
3919 overdue, caused by the large number of possibilities that now exist.
3920 Originally there were few.) While I was there, I also converted the
3921 str(n)cmp tests so they don't re-test the leading "-" and the first
3922 character, in the hope this might squeeze out yet more improvement.
3924 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3925 flag allowing group syntax was set by the header_syntax check but not
3926 turned off, possible causing trouble later; (2) The flag was not being
3927 set at all for the header_verify test, causing "group"-style headers to
3928 be rejected. I have now set it in this case, and also caused header_
3929 verify to ignore an empty address taken from a group. While doing this, I
3930 came across some other cases where the code for allowing group syntax
3931 while scanning a header line wasn't quite right (mostly, not resetting
3932 the flag correctly in the right place). These bugs could have caused
3933 trouble for malformed header lines. I hope it is now all correct.
3935 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3936 with the "reply" argument non-NULL. The code, however (which originally
3937 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3938 but it didn't always do it. This confused somebody who was copying the
3939 code for some other use. I have removed all the tests.
3941 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3942 feature that was used to support insecure browsers during the U.S. crypto
3943 embargo. It requires special client support, and Exim is probably the
3944 only MTA that supported it -- and would never use it because real RSA is
3945 always available. This code has been removed, because it had the bad
3946 effect of slowing Exim down by computing (never used) parameters for the
3947 RSA_EXPORT functionality.
3949 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3950 authenticator to fail if there's a tab character in the incoming data
3951 (there should never be unless someone is messing about, as it's supposed
3952 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3953 if the connection is using TLS or if the remote IP is the same as the
3954 local IP, and the "valid-client-cert option" if a client certificate has
3957 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3958 authenticators. This can be used for authorization after authentication
3959 succeeds. (In the case of plaintext, it servers for both authentication
3962 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3963 if any retry times were supplied.
3965 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3966 connection, where there is no client IP address to check. In this
3967 situation, the verify now always succeeds.
3969 PH/25 Applied John Jetmore's -Mset patch.
3971 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3973 PH/27 In a string expansion for a processed (not raw) header when multiple
3974 headers of the same name were present, leading whitespace was being
3975 removed from all of them, but trailing whitespace was being removed only
3976 from the last one. Now trailing whitespace is removed from each header
3977 before concatenation. Completely empty headers in a concatenation (as
3978 before) are ignored.
3980 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3981 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3983 PH/29 [Removed. This was a change that I later backed out, and forgot to
3984 correct the ChangeLog entry (that I had efficiently created) before
3985 committing the later change.]
3987 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3988 address errors (4xx response to RCPT) over the same connection as other
3989 messages routed to the same hosts. Such deliveries are always "forced",
3990 so retry times are not inspected. This resulted in far too many retries
3991 for the affected addresses. The effect occurred only when there were more
3992 hosts than the hosts_max_try setting in the smtp transport when it had
3993 the 4xx errors. Those hosts that it had tried were not added to the list
3994 of hosts for which the message was waiting, so if all were tried, there
3995 was no problem. Two fixes have been applied:
3997 (i) If there are any address or message errors in an SMTP delivery, none
3998 of the hosts (tried or untried) are now added to the list of hosts
3999 for which the message is waiting, so the message should not be a
4000 candidate for sending over the same connection that was used for a
4001 successful delivery of some other message. This seems entirely
4002 reasonable: after all the message is NOT "waiting for some host".
4003 This is so "obvious" that I'm not sure why it wasn't done
4004 previously. Hope I haven't missed anything, but it can't do any
4005 harm, as the worst effect is to miss an optimization.
4007 (ii) If, despite (i), such a delivery is accidentally attempted, the
4008 routing retry time is respected, so at least it doesn't keep
4009 hammering the server.
4011 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4012 in ${readsocket because some servers need this prod.
4014 PH/32 Added some extra debug output when updating a wait-xxx database.
4016 PH/33 The hint "could be header name not terminated by colon", which has been
4017 given for certain expansion errors for a long time, was not being given
4018 for the ${if def:h_colon_omitted{... case.
4020 PH/34 The spec says: "With one important exception, whenever a domain list is
4021 being scanned, $domain contains the subject domain." There was at least
4022 one case where this was not true.
4024 PH/35 The error "getsockname() failed: connection reset by peer" was being
4025 written to the panic log as well as the main log, but it isn't really
4026 panic-worthy as it just means the connection died rather early on. I have
4027 removed the panic log writing for the ECONNRESET error when getsockname()
4030 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4031 runs only) independently of the message's sender address. This meant
4032 that, if the 4xx error was in fact related to the sender, a different
4033 message to the same recipient with a different sender could confuse
4034 things. In particular, this can happen when sending to a greylisting
4035 server, but other circumstances could also provoke similar problems.
4036 I have changed the default so that the retry time for these errors is now
4037 based a combination of the sender and recipient addresses. This change
4038 can be overridden by setting address_retry_include_sender=false in the
4041 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4042 remote server are returned as part of bounce messages. This was not
4043 happening for LMTP over a pipe (the lmtp transport), but now it is the
4044 same for both kinds of LMTP.
4046 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4047 in header lines that were in fact CNAMEs. This is no longer the case.
4049 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4050 and queue runs started by the daemon processed all messages. This has
4051 been fixed so that -R and -S can now usefully be given with -q<time>.
4053 PH/40 Import PCRE release 6.7 (fixes some bugs).
4055 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4057 PH/42 Give an error if -q is specified more than once.
4059 PH/43 Renamed the variables $interface_address and $interface_port as
4060 $received_ip_address and $received_port, to make it clear that these
4061 values apply to message reception, and not to the outgoing interface when
4062 a message is delivered. (The old names remain recognized, of course.)
4064 PH/44 There was no timeout on the connect() call when using a Unix domain
4065 socket in the ${readsocket expansion. There now is.
4067 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4068 be meaningful with "accept".
4070 SC/01 Eximstats V1.43
4071 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4073 SC/02 Eximstats V1.44
4074 Use a glob alias rather than an array ref in the generated
4075 parser. This improves both readability and performance.
4077 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4078 Collect SpamAssassin and rejection statistics.
4079 Don't display local sender or destination tables unless
4080 there is data to show.
4081 Added average volumes into the top table text output.
4083 SC/04 Eximstats V1.46
4084 Collect data on the number of addresses (recipients)
4085 as well as the number of messages.
4087 SC/05 Eximstats V1.47
4088 Added 'Message too big' to the list of mail rejection
4089 reasons (thanks to Marco Gaiarin).
4091 SC/06 Eximstats V1.48
4092 Mainlog lines which have GMT offsets and are too short to
4093 have a flag are now skipped.
4095 SC/07 Eximstats V1.49 (Alain Williams)
4096 Added the -emptyok flag.
4098 SC/08 Eximstats V1.50
4099 Fixes for obtaining the IP address from reject messages.
4101 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4102 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4103 whitespace changes from 4.64-PH/27
4105 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4108 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4109 are found, allow negative numbers in numeric criteria)
4111 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4113 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4114 to match changes made in 4.64-PH/43
4116 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4118 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4119 "domain" is actually an IP address, and force a failure. This locks out
4120 those revolvers/nameservers that support "A-for-A" lookups, in
4121 contravention of the specifications.
4123 PH/48 When a host name was looked up from an IP address, and the subsequent
4124 forward lookup of the name timed out, the host name was left in
4125 $sender_host_name, contrary to the specification.
4127 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4128 restricted to single-key lookups, Exim was not diagnosing an error if
4129 * or *@ was used with a query-style lookup.
4131 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4133 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4134 long ago, but noone interested enough thought of it. Let's just say that
4135 the "1.1" means that there are some new functions that weren't there at
4136 some point in the past.
4138 PH/51 Error processing for expansion failure of helo_data from an smtp
4139 transport during callout processing was broken.
4141 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4142 tested/used via the -bh/-bhc/-bs options.
4144 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4145 bug, fixed in subsequent PCRE releases).
4147 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4148 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4150 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4156 SC/01 Use a glob alias rather than an array ref in eximstats generated
4157 parser. This improves both readability and performance.
4159 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4160 Don't display local sender or destination tables in eximstats unless
4161 there is data to show.
4162 Added average volumes into the eximstats top table text output.
4164 SC/03 Collect data on the number of addresses (recipients) as well
4165 as the number of messages in eximstats.
4167 TF/01 Correct an error in the documentation for the redirect router. Exim
4168 does (usually) call initgroups() when daemonizing.
4170 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4171 with consistent privilege compared to when running as a daemon.
4173 TF/03 Note in the spec that $authenticated_id is not set for local
4174 submissions from trusted users.
4176 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4177 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4179 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4180 by adding some example configuration directives to the default
4181 configuration file. A little bit of work is required to uncomment the
4182 directives and define how usernames and passwords are checked, but
4183 there is now a framework to start from.
4185 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4186 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4187 without this. I don't know how relevant this is to other LDAP libraries.
4189 PH/02 Add the verb name to the "unknown ACL verb" error.
4191 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4193 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4195 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4196 directory not expanded when it should be if an expanded home directory
4197 was set for the address (which is overridden by the transport).
4199 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4202 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4203 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4204 because it is too late at that time, and has no effect.
4206 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4207 security issue with \' (bugzilla #107). I could not use the
4208 PQescapeStringConn() function, because it needs a PGconn value as one of
4211 PH/08 When testing addresses using -bt, indicate those final addresses that
4212 are duplicates that would not cause an additional delivery. At least one
4213 person was confused, thinking that -bt output corresponded to deliveries.
4214 (Suppressing duplicates isn't a good idea as you lose the information
4215 about possibly different redirections that led to the duplicates.)
4217 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4218 systems where poll() doesn't work, in particular OS X.
4220 PH/10 Added more information to debugging output for retry time not reached.
4222 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4223 operations in malware.c.
4225 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4228 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4229 syslog_duplication set false, log lines that would normally be written
4230 both the the main log and to the reject log were not written to syslog at
4233 PH/14 In the default configuration, change the use of "message" in ACL warn
4234 statements to "add_header".
4236 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4237 not followed by a command (e.g. "seen endif").
4239 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4240 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4243 PH/17 Added extra conditions to the default value of delay_warning_condition
4247 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4248 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4249 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4252 The Auto-Submitted: and various List- headers are standardised, whereas I
4253 don't think Precedence: ever was.
4255 PH/18 Refactored debugging code in route_finduser() to show more information,
4256 in particular, the error code if getpwnam() issues one.
4258 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4259 This is apparently needed in addition to the PH/07 change above to avoid
4260 any possible encoding problems.
4262 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4263 but not after initializing Perl.
4265 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4266 output them only if debugging. By default they are written stderr,
4267 apparently, which is not desirable.
4269 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4272 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4275 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4277 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4278 authenticated or an ident call has been made. Suppress the default
4279 values for $authenticated_id and $authenticated_sender (but permit -oMai
4280 and -oMas) when testing with -bh.
4282 PH/24 Re-jigged the order of the tests in the default configuration so that the
4283 tests for valid domains and recipients precede the DNS black list and CSA
4284 tests, on the grounds that those ones are more expensive.
4286 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4287 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4288 This bug exists in every version of Exim that I still have, right back to
4291 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4292 However, an attempt to turn on case-sensitivity in a regex key by
4293 including (?-i) didn't work because the subject string was already
4294 lowercased, and the effects were non-intuitive. It turns out that a
4295 one-line patch can be used to allow (?-i) to work as expected.
4301 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4302 other effects) broke the use of negated acl sub-conditions.
4304 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4307 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4308 "Deny" causes Exim to reject the incoming connection with a 554 error.
4309 Unfortunately, if there is a major crisis, such as a disk failure,
4310 tcp-wrappers gives "deny", whereas what one would like would be some
4311 kind of temporary error. A kludge has been added to help with this.
4312 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4313 554 error is used if errno is still zero or contains ENOENT (which occurs
4314 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4317 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4319 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4320 errors. Otherwise a message that provokes a temporary error (when other
4321 messages do not) can cause a whole host to time out.
4323 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4324 addresses were routed directly to files or pipes from a redirect router.
4325 File deliveries just didn't batch; pipe deliveries might have suffered
4328 PH/06 A failure to get a lock for a hints database would erroneously always say
4329 "Failed to get write lock", even when it was really a read lock.
4331 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4332 of 0600. This has been changed to use the value of the lockfile_mode
4333 option (which defaults to 0600).
4335 PH/08 Applied small patch from the Sieve maintainer.
4337 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4338 folder from quota calculations, a direct delivery into this folder messed
4339 up the contents of the maildirsize file. This was because the regex was
4340 used only to exclude .Trash (or whatever) when the size of the mailbox
4341 was calculated. There was no check that a delivery was happening into an
4342 excluded directory. This bug has been fixed by ignoring all quota
4343 processing for deliveries into excluded directories.
4345 PH/10 Added the maildirfolder_create_regex option to appendfile.
4351 PH/01 The code for finding all the local interface addresses on a FreeBSD
4352 system running IPv6 was broken. This may well have applied to all BSD
4353 systems, as well as to others that have similar system calls. The broken
4354 code found IPv4 interfaces correctly, but gave incorrect values for the
4355 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4356 that it would not match correctly against @[] and not recognize the IPv6
4359 PH/02 The ipliteral router was not recognizing addresses of the form user@
4360 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4362 PH/03 Added disable_ipv6.
4364 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4365 decoded form, because it is most often used to construct To: headers
4366 lines in autoreplies, and the decoded form may well be syntactically
4367 invalid. However, $reply_address has leading white space removed, and all
4368 newlines turned into spaces so that the autoreply transport does not
4371 PH/05 If group was specified without a user on a router, and no group or user
4372 was specified on a transport, the group from the router was ignored.
4374 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4375 for visible compile-time settings that can be used to change these
4376 numbers, for those that want even more. Backwards compatibility with old
4377 spool files has been maintained. However, going back to a previous Exim
4378 release will lost any variables that are in spool files.
4380 PH/07 Two small changes when running in the test harness: increase delay when
4381 passing a TCP/IP connection to a new process, in case the original
4382 process has to generate a bounce, and remove special handling of
4383 127.0.0.2 (sic), which is no longer necessary.
4385 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4386 be the same on different OS.
4388 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4391 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4392 whether --show-vars was specified or not
4394 JJ/02 exipick: Added support for new ACL variable spool format introduced
4397 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4398 syntactically invalid From: or Reply-to: line, and a filter used this to
4399 generate an autoreply, and therefore failed to obtain an address for the
4400 autoreply, Exim could try to deliver to a non-existent relative file
4401 name, causing unrelated and misleading errors. What now happens is that
4402 it logs this as a hard delivery error, but does not attempt to create a
4405 PH/11 The exinext utility has a -C option for testing purposes, but although
4406 the given file was scanned by exinext itself; it wasn't being passed on
4407 when Exim was called.
4409 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4410 an end-of-file indication when reading a command response.
4412 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4413 compiled. In many other places in Exim, IPv6 addresses are always
4414 recognized, so I have changed this. It also means that IPv4 domain
4415 literals of the form [IPV4:n.n.n.n] are now always recognized.
4417 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4418 used if the router is not running as root, for example, when verifying at
4419 ACL time, or when using -bh. The debugging output from this situation was
4420 non-existent - all you got was a failure to exec. I have made two
4423 (a) Failures to set uid/gid, the current directory, or a process leader
4424 in a subprocess such as that created by queryprogram now generate
4425 suitable debugging output when -d is set.
4427 (b) The queryprogram router detects when it is not running as root,
4428 outputs suitable debugging information if -d is set, and then runs
4429 the subprocess without attempting to change uid/gid.
4431 PH/15 Minor change to Makefile for building test_host (undocumented testing
4434 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4435 additional section of a DNS packet that returns MX or SRV records.
4436 Instead, it always explicitly searches for A/AAAA records. This avoids
4437 major problems that occur when a DNS server includes only records of one
4438 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4439 fixed another bug: if SRV records were looked up and the corresponding
4440 address records were *not* found in the additional section, the port
4441 values from the SRV records were lost.
4443 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4444 using the correct key (the original address) when searching the retry
4445 rules in order to find which one to use for generating the retry hint.
4447 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4448 adding the default one. Similarly, if it contains a Reply-To: header, the
4449 errors_reply_to option, if set, is not used.
4451 PH/19 When calculating a retry time, Exim used to measure the "time since
4452 failure" by looking at the "first failed" field in the retry record. Now
4453 it does not use this if it is later than than the arrival time of the
4454 message. Instead it uses the arrival time. This makes for better
4455 behaviour in cases where some deliveries succeed, thus re-setting the
4456 "first failed" field. An example is a quota failure for a huge message
4457 when small messages continue to be delivered. Without this change, the
4458 "time since failure" will always be short, possible causing more frequent
4459 delivery attempts for the huge message than are intended.
4460 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4462 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4463 $1, $2, $3) because the numerical variables can be reset during some
4464 expansion items (e.g. "match"), thereby losing the authentication data.
4466 PH/21 Make -bV show the size of off_t variables so that the test suite can
4467 decide whether to run tests for quotas > 2G.
4469 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4470 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4471 filecount value is greater than 2G or if a quota value is greater than 2G
4472 on a system where the size of off_t is not greater than 4, a panic error
4475 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4476 never match. The debug and -bh output now contains an explicit error
4477 message indicating a malformed IPv4 address or mask.
4479 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4480 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4481 PH/23 above applies.
4483 PH/25 Do not write to syslog when running in the test harness. The only
4484 occasion when this arises is a failure to open the main or panic logs
4485 (for which there is an explicit test).
4487 PH/26 Added the /no_tell option to "control=freeze".
4489 PH/27 If a host name lookup failed very early in a connection, for example, if
4490 the IP address matched host_lookup and the reverse lookup yielded a name
4491 that did not have a forward lookup, an error message of the form "no IP
4492 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4493 could be logged. Now it outputs the IP address instead of "NULL".
4495 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4496 allows the sender and the authenticated sender to be set when
4497 submitting a message from within Exim. Since child_open_exim() is
4498 documented for local_scan(), the new function should be too.
4500 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4501 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4502 results in an empty string is now treated as unset.
4504 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4506 PH/31 Added sender_verify_fail logging option.
4508 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4509 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4510 tidied the source and removed it altogether.
4512 PH/33 When a queue run was abandoned because the load average was too high, a
4513 log line was always written; now it is written only if the queue_run log
4514 selector is set. In addition, the log line for abandonment now contains
4515 information about the queue run such as the pid. This is always present
4516 in "start" and "stop" lines but was omitted from the "abandon" line.
4518 PH/34 Omit spaces between a header name and the colon in the error message that
4519 is given when verify = headers_syntax fails (if there are lots of them,
4520 the message gets confusing).
4522 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4523 names, as there are now some PTR records that contain slashes. This check
4524 is only to protect against broken name servers that fall over on strange
4525 characters, so the fact that it applies to all lookups doesn't matter.
4527 PH/36 Now that the new test suite is complete, we can remove some of the
4528 special code in Exim that was needed for the old test suite. For example,
4529 sorting DNS records because real resolvers return them in an arbitrary
4530 order. The new test suite's fake resolver always returns records in the
4533 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4534 bounces) except when queue_only is set, to avoid logging races between
4535 the different processes.
4537 PH/38 Panic-die if .include specifies a non-absolute path.
4539 PH/39 A tweak to the "H" retry rule from its user.
4541 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4542 a label. They prevented compilation on older perls.
4544 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4545 a warning to be raised on newish perls.
4547 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4548 on queue. Changes to match documented behaviour of showing count of
4549 messages matching specified criteria.
4551 PH/40 Changed the default ident timeout from 30s to 5s.
4553 PH/41 Added support for the use of login_cap features, on those BSD systems
4554 that have them, for controlling the resources used by pipe deliveries.
4556 PH/42 The content-scanning code uses fopen() to create files in which to put
4557 message data. Previously it was not paying any attention to the mode of
4558 the files. Exim runs with umask(0) because the rest of the code creates
4559 files with open(), and sets the required mode explicitly. Thus, these
4560 files were ending up world-writeable. This was not a big issue, because,
4561 being within the spool directory, they were not world-accessible. I have
4562 created a function called modefopen, which takes an additional mode
4563 argument. It sets umask(777), creates the file, chmods it to the required
4564 mode, then resets the umask. All the relevant calls to fopen() in the
4565 content scanning code have been changed to use this function.
4567 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4568 to 24 hours. This avoids potential overflow problems when processing G
4569 and H retry rules. I suspect nobody ever tinkers with this value.
4571 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4573 PH/45 When the plaintext authenticator is running as a client, the server's
4574 challenges are checked to ensure they are valid base64 strings. By
4575 default, the authentication attempt is cancelled if an invalid string is
4576 received. Setting client_ignore_invalid_base64 true ignores these errors.
4577 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4578 they are received. Thus, the responses can be made to depend on the
4579 challenges. If an invalid string is ignored, an empty string is placed in
4582 PH/46 Messages that are created by the autoreply transport now contains a
4583 References: header, in accordance with RFCs 2822 and 3834.
4585 PH/47 Added authenticated_sender_force to the smtp transport.
4587 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4589 PH/49 Installed latest patch from the Sieve maintainer.
4591 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4592 was also set, the appendfile transport was unnecessarily scanning a
4593 directory of message files (e.g. for maildir delivery) to find the count
4594 of files (along with the size), even though it did not need this
4595 information. It now does the scan only if it needs to find either the
4596 size of the count of files.
4598 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4600 PH/52 Two bugs concerned with error handling when the smtp transport is
4603 (i) Exim was not creating retry information for temporary errors given
4604 for individual recipients after the DATA command when the smtp transport
4605 was used in LMTP mode. This meant that they could be retried too
4606 frequently, and not timed out correctly.
4608 (ii) Exim was setting the flag that allows error details to be returned
4609 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4610 recipients that were returned after the DATA command.
4612 PH/53 This is related to PH/52, but is more general: for any failing address,
4613 when detailed error information was permitted to be returned to the
4614 sender, but the error was temporary, then after the final timeout, only
4615 "retry timeout exceeded" was returned. Now it returns the full error as
4616 well as "retry timeout exceeded".
4618 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4619 do this, and (what is worse) MTAs that accept it.
4621 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4622 will now be deprecated.
4624 PH/56 New os.c-cygwin from the Cygwin maintainer.
4626 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4627 formats (previously only available in exim formats via -bpr, -bpru,
4628 and -bpra. Now also available in native and exiqgrep formats)
4630 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4631 with very large, slow to parse queues
4633 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4635 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4637 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4638 responses to authentication challenges, though it was showing the
4639 challenges; (ii) I've removed the CR characters from the debug output for
4642 PH/58 Allow for the insertion of a newline as well as a space when a string
4643 is turned into more than one encoded-word during RFC 2047 encoding. The
4644 Sieve code now uses this.
4646 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4647 data_4xx, lost_connection, tls_required.
4649 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4650 message was being sent as an SMTP response.
4652 PH/61 Add -l and -k options to exicyclog.
4654 PH/62 When verifying, if an address was redirected to one new address, so that
4655 verification continued, and the new address failed or deferred after
4656 having set something in $address_data, the value of $address_data was not
4657 passed back to the ACL. This was different to the case when no
4658 redirection occurred. The value is now passed back in both cases.
4660 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4661 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4662 use login_cap.h, so on its own it isn't the distinguishing feature. The
4663 new name refers directly to the setclassresources() function.
4665 PH/65 Added configuration files for NetBSD3.
4667 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4669 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4670 is preferred over IPv4.
4672 PH/68 The bounce_return_message and bounce_return_body options were not being
4673 honoured for bounces generated during the reception of non-SMTP messages.
4674 In particular, this applied to messages rejected by the ACL. This bug has
4675 been fixed. However, if bounce_return_message is true and bounce_return_
4676 body is false, the headers that are returned for a non-SMTP message
4677 include only those that have been read before the error was detected.
4678 (In the case of an ACL rejection, they have all been read.)
4680 PH/69 The HTML version of the specification is now built in a directory called
4681 spec_html instead of spec.html, because the latter looks like a path with
4682 a MIME-type, and this confuses some software.
4684 PH/70 Catch two compiler warnings in sieve.c.
4686 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4687 function verify_get_ident() calls ip_connect() to connect a socket, but
4688 if the "connect()" function timed out, ip_connect() used to close the
4689 socket. However, verify_get_ident() also closes the socket later, and in
4690 between Exim writes to the log, which may get opened at this point. When
4691 the socket was closed in ip_connect(), the log could get the same file
4692 descriptor number as the socket. This naturally causes chaos. The fix is
4693 not to close the socket in ip_connect(); the socket should be closed by
4694 the function that creates it. There was only one place in the code where
4695 this was missing, in the iplookup router, which I don't think anybody now
4696 uses, but I've fixed it anyway.
4698 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4699 well as to direct DNS lookups. Otherwise the handling of names in host
4700 lists is inconsistent and therefore confusing.
4706 PH/01 Two changes to the default runtime configuration:
4708 (1) Move the checks for relay_from_hosts and authenticated clients from
4709 after to before the (commented out) DNS black list checks.
4711 (2) Add control=submission to the relay_from_hosts and authenticated
4712 clients checks, on the grounds that messages accepted by these
4713 statements are most likely to be submissions.
4715 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4717 (1) Generate an error if the third argument for the ${prvs expansion is
4720 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4723 (3) Reset the variables that are obtained from the first argument of
4724 ${prvscheck and used in the second argument before leaving the code,
4725 because their memory is reclaimed, so using them afterwards may do
4728 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4729 one (it's much easier than Tom thought :-).
4731 (5) Because of (4), we can now allow for the use of $prvscheck_result
4732 inside the third argument.
4734 PH/03 For some reason, the default setting of PATH when running a command from
4735 a pipe transport was just "/usr/bin". I have changed it to
4738 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4739 anything to be listed in the output from -bV.
4741 PH/05 When a filter generated an autoreply, the entire To: header line was
4742 quoted in the delivery log line, like this:
4744 => >A.N.Other <ano@some.domain> <original@ddress> ...
4746 This has been changed so that it extracts the operative address. There
4747 may be more than one such address. If so, they are comma-separated, like
4750 => >ano@some.domain,ona@other.domain <original@ddress> ...
4752 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4753 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4754 not being looked up in the rDNS to get a host name, Exim was showing the
4755 IP address twice in Received: lines, even though the IP addresses were
4756 identical. For example:
4758 Received: from [1.2.3.4] (helo=[1.2.3.4])
4760 However, if the real host name was known, it was omitting the HELO data
4761 if it matched the actual IP address. This has been tidied up so that it
4762 doesn't show the same IP address twice.
4764 PH/07 When both +timestamp and +memory debugging was on, the value given by
4765 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4766 called by the debug printing, thereby overwriting the timestamp buffer.
4767 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4769 PH/08 When the original message was included in an autoreply transport, it
4770 always said "this is a copy of the message, including all the headers",
4771 even if body_only or headers_only was set. It now gives an appropriate
4774 PH/09 Applied a patch from the Sieve maintainer which:
4776 o fixes some comments
4777 o adds the (disabled) notify extension core
4778 o adds some debug output for the result of if/elsif tests
4779 o points to the current vacation draft in the documentation
4780 and documents the missing references header update
4784 o fixes a bug in processing the envelope test (when testing
4785 multiple envelope elements, the last element determined the
4788 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4789 Electronic Mail") by including:
4791 Auto-submitted: auto-generated
4793 in the messages that it generates (bounce messages and others, such as
4794 warnings). In the case of bounce messages for non-SMTP messages, there was
4795 also a typo: it was using "Auto_submitted" (underscore instead of
4796 hyphen). Since every message generated by Exim is necessarily in response
4797 to another message, thes have all been changed to:
4799 Auto-Submitted: auto-replied
4801 in accordance with these statements in the RFC:
4803 The auto-replied keyword:
4805 - SHOULD be used on messages sent in direct response to another
4806 message by an automatic process,
4808 - MUST NOT be used on manually-generated messages,
4810 - MAY be used on Delivery Status Notifications (DSNs) and Message
4811 Disposition Notifications (MDNs),
4813 - MUST NOT be used on messages generated by automatic or periodic
4814 processes, except for messages which are automatic responses to
4817 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4818 to the default Received: header definition.
4820 PH/12 Added log selector acl_warn_skipped (default on).
4822 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4823 variables because (a) they are in the wrong storage pool and (b) even if
4824 they were copied, it wouldn't work properly because of the caching.
4826 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4827 checking when decoding. Apparently there are clients that generate
4828 overlong encoded strings. Why am I not surprised?
4830 PH/15 If the first argument of "${if match_address" was not empty, but did not
4831 contain an "@" character, Exim crashed. Now it writes a panic log message
4832 and treats the condition as false.
4834 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4836 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4837 extension "envelope-auth". The code is finished and in agreement with
4838 other implementations, but there is no documentation so far and in fact,
4839 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4840 not changing the active code.
4842 Print executed "if" and "elsif" statements when debugging is used. This
4843 helps a great deal to understand what a filter does.
4845 Document more things not specified clearly in RFC3028. I had all this
4846 sorted out, when out of a sudden new issues came to my mind. Oops."
4848 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4851 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4852 canonical form (as documented). However, after a host name lookup from
4853 the IP address, check_host() was doing a simple string comparison with
4854 addresses acquired from the DNS when checking that the found name did
4855 have the original IP as one of its addresses. Since any found IPv6
4856 addresses are likely to be in abbreviated form, the comparison could
4857 fail. Luckily, there already exists a function for doing the comparison
4858 by converting both addresses to binary, so now that is used instead of
4859 the text comparison.
4861 PH/20 There was another similar case to PH/19, when a complete host name was
4862 given in a host list; looking up its IP address could give an abbreviated
4863 form, whereas the current host's name might or might not be abbreviated.
4864 The same fix has been applied.
4870 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4871 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4874 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4875 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4877 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4879 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4880 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4881 Also, for "Auto-Submitted", treat anything other than "no" as
4882 non-personal, in accordance with RFC 3834. (Previously it treated
4883 anything starting "auto-" as non-personal.)
4885 TF/01 The control=submission/name=... option had a problem with syntax
4886 errors if the name included a slash character. The /name= option
4887 now slurps the rest of the string, so it can include any characters
4888 but it must come last in the list of options (after /sender_retain
4891 PH/05 Some modifications to the interface to the fake nameserver for the new
4899 TK/01 Added the "success_on_redirect" address verification option. See
4900 NewStuff for rationale and an example.
4902 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4904 PH/02 Patch to exigrep to allow it to work on syslog lines.
4906 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4907 fread() to skip over the body file's header line, because in Cygwin the
4908 header line is locked and is inaccessible.
4910 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4911 co-exist for some time) to make it clear that it is the Exim ID that is
4912 referenced, not the Message-ID: header line.
4914 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4915 string_format() function, because snprintf() does not exist on all
4918 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4919 ${stat: expansion item.
4921 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4922 protocol synchronization error", to keep the pedants happy.
4924 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4925 well as for IRIX systems, when gcc is being used. See the host.c source
4928 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4930 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4933 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4934 transport and to the smtp transport in LMTP mode.
4936 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4938 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4939 run a filter in a subprocess. This could lead to confusion in subsequent
4940 lookups in the parent process. There should also be a search_tidyup() at
4941 the end of the subprocess.
4943 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4944 only if the host matched helo_try_verify_hosts, which caused the
4945 verification to occur when the EHLO/HELO command was issued. The ACL just
4946 tested the remembered result. Now, if a previous verification attempt has
4947 not happened, "verify = helo" does it there and then.
4949 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4951 TK/03 Fix log output including CR from clamd.
4953 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4954 exist provoked a memory error which could cause a segfault.
4956 PH/15 Installed PCRE 6.2
4958 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4960 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4961 of the problem. Specifically, suggested +O2 rather than +O1 for the
4964 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4966 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4967 record, the port was not correctly logged when the outgoing_port log
4968 selector was set (it logged the transort's default port).
4970 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4971 fallback_hosts, and "hosts" in the smtp transport.
4973 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4974 host errors such as "Connection refused".
4976 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4977 authentication with radiusclient 0.4.9:
4979 - Error returned from rc_read_config was caught wrongly
4980 - Username/password not passed on to radius server due to wrong length.
4982 The presumption is that some radiusclient API changes for 4.51/PH/17
4983 were not taken care of correctly. The code is still untested by me (my
4984 Linux distribution still has 0.3.2 of radiusclient), but it was
4985 contributed by a Radius user.
4987 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4988 expanding the "port" option of the smtp transport.
4990 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4991 while reading a MIME header. Thanks to Tom Hughes for a patch.
4993 PH/24 Include config.h inside local_scan.h so that configuration settings are
4996 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4997 that in an ACL for RCPT (for example), you can examine exactly what was
5000 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5001 commands, but it was not correctly comparing the address with the actual
5002 client host address. Thus, it would show the EHLO address in Received:
5003 header lines when this was not necessary.
5005 PH/27 Added the % operator to ${eval:}.
5007 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5008 it should be ignoring failures (because with -C, for example, it has lost
5009 privilege). It wasn't ignoring creation failures other than "already
5012 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5015 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5016 because an input line was too long, either on its own, or by virtue of
5017 too many continuations, the temporary file was not being removed, and the
5018 return code was incorrect.
5020 PH/31 Missing "BOOL" in function definition in filtertest.c.
5022 PH/32 Applied Sieve patches from the maintainer.
5024 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5026 PH/33 Added "verify = not_blind".
5028 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5029 Local/Makefile (with some defaults set). These are used in built scripts
5030 such as exicyclog, but they have never been used in the exim_install
5031 script (though there are many overriding facilities there). I have
5032 arranged that the exim_install script now takes note of these two
5035 PH/35 Installed configuration files for Dragonfly.
5037 PH/36 When a locally submitted message by a trusted user did not contain a
5038 From: header, and the sender address was obtained from -f or from an SMTP
5039 MAIL command, and the trusted user did not use -F to supply a sender
5040 name, $originator_name was incorrectly used when constructing a From:
5041 header. Furthermore, $originator_name was used for submission mode
5042 messages from external hosts without From: headers in a similar way,
5043 which is clearly wrong.
5045 PH/37 Added control=suppress_local_fixups.
5047 PH/38 When log_selector = +received_sender was set, and the addition of the
5048 sender made the log line's construction buffer exactly full, or one byte
5049 less than full, an overflow happened when the terminating "\n" was
5052 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5053 when the result of a list match is failure because a DNS lookup failed.
5055 PH/40 RM_COMMAND is now used in the building process.
5057 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5058 the "build-* directories that it finds.
5060 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5061 address in a domain literal was a prefix of an interface address.
5063 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5064 when verifying a sender address, unless rewrite_headers is false.
5066 PH/44 Wrote a long comment about why errors_to addresses are verified as
5067 recipients, not senders.
5069 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5070 the ratelimit ACL was added.
5072 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5074 PH/46 Added extra information about PostgreSQL errors to the error string.
5076 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5077 avoiding the need to install special zones in a real server. This is
5078 backwards compatible; if it can't find the fake resolver, it drops back.
5079 Thus, both old and new test suites can be run.
5081 TF/02 Added util/ratelimit.pl
5083 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5084 clock is set back in time.
5086 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5087 Candler <B.Candler@pobox.com>.
5089 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5090 OK for addresses that are the result of redirections.
5092 PH/48 A number of further additions for the benefit of the new test suite,
5093 including a fake gethostbyname() that interfaces to the fake DNS resolver
5096 TF/06 The fix for widen_domains has also been applied to qualify_single and
5097 search_parents which are the other dnslookup options that can cause
5100 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5103 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5105 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5106 service" if it encountered an MX record with an empty target hostname.
5107 The message is now "an MX or SRV record indicated no SMTP service".
5109 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5110 if the DNS of the sending site is misconfigured. This is quite a
5111 common situation. This change restores the behaviour of treating a
5112 helo verification defer as a failure.
5114 PH/51 If self=fail was set on a router, the bounce message did not include the
5115 actual error message.
5121 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5123 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5124 command itself ended in error, the underlying message about the transport
5125 filter timeout was being overwritten with the pipe command error. Now the
5126 underlying error message should be appended to the second error message.
5128 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5130 PH/02 Reduce the amount of output that "make" produces by default. Full output
5131 can still be requested.
5133 PH/03 The warning log line about a condition test deferring for a "warn" verb
5134 was being output only once per connection, rather than after each
5135 occurrence (because it was using the same function as for successful
5136 "warn" verbs). This seems wrong, so I have changed it.
5138 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5139 it should not have, which might have caused a crash in the right
5140 circumstances, but probably never did.
5142 PH/04 Installed a modified version of Tony Finch's patch to make submission
5143 mode fix the return path as well as the Sender: header line, and to
5144 add a /name= option so that you can make the user's friendly name appear
5147 TF/03 Added the control = fakedefer ACL modifier.
5149 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5150 Mark Lowes for thorough testing.
5152 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5154 TK/03 Merged latest SRS patch from Miles Wilton.
5156 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5157 with the definition in sysexits.h (which is #included earlier).
5158 Fortunately, Exim does not actually use EX_OK. The code used to try to
5159 preserve the sysexits.h value, by assuming that macro definitions were
5160 scanned for macro replacements. I have been disabused of this notion,
5161 so now the code just undefines EX_OK before #including unistd.h.
5163 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5164 in the smtp transport. When a block could not be written in a single
5165 write() function, the timeout was being re-applied to each part-write.
5166 This seems wrong - if the receiver was accepting one byte at a time it
5167 would take for ever. The timeout is now adjusted when this happens. It
5168 doesn't have to be particularly precise.
5170 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5171 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5173 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5174 without a domain if the check to <postmaster@domain> fails.
5176 SC/01 Eximstats: added -xls and the ability to specify output files
5177 (patch written by Frank Heydlauf).
5179 SC/02 Eximstats: use FileHandles for outputting results.
5181 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5183 SC/04 Eximstats: fixed display of large numbers with -nvr option
5185 SC/05 Eximstats: fixed merging of reports with empty tables.
5187 SC/06 Eximstats: added the -include_original_destination flag
5189 SC/07 Eximstats: removed tabs and trailing whitespace.
5191 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5193 TK/06 MBOX spool code: Add real "From " MBOX separator line
5194 so the .eml file is really in mbox format (even though
5195 most programs do not really care). Patch from Alex Miller.
5197 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5198 The latter is generated from $received_to and is only set if the
5199 message has one envelope recipient. SA can use these headers,
5200 obviously out-of-the-box. Patch from Alex Miller.
5202 PH/08 The ${def test on a variable was returning false if the variable's
5203 value was "0", contrary to what the specification has always said!
5204 The result should be true unless the variable is empty.
5206 PH/09 The syntax error of a character other than { following "${if
5207 def:variable_name" (after optional whitespace) was not being diagnosed.
5208 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5209 accidental colon was present, for example, could give incorrect results.
5211 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5212 result is used (not including appendfile, where other changes are about
5215 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5216 This involved changing a lot of size variables from int to off_t. It
5217 should work with maildirs and everything.
5219 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5220 spamd dying while we are connected to it.
5222 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5225 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5226 function for BDB 4.3.
5228 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5230 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5231 the sources for PCRE have been re-arranged and more files are now
5234 PH/15 The code I had for printing potentially long long variables in PH/11
5235 above was not the best (it lost precision). The length of off_t variables
5236 is now inspected at build time, and an appropriate printing format (%ld
5237 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5238 to be "long long int" or "long int". This is needed for the internal
5239 formatting function string_vformat().
5241 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5242 the configuration file to be ":syslog", then the script "guesses" where
5243 the logs files are, rather than using the compiled in default. In our
5244 case the guess is not the same as the compiled default, so the script
5245 suddenly stopped working when I started to use syslog. The patch checks
5246 to see if log_file_path is "". If so, it attempts to read it from exim
5247 with no configuration file to get the compiled in version, before it
5248 falls back to the previous guessing code."
5250 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5251 implementing BATV in an Exim configuration. See NewStuff for the gory
5254 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5255 Makefile that are specific to HP-UX.
5257 PH/18 If the "use_postmaster" option was set for a recipient callout together
5258 with the "random" option, the postmaster address was used as the MAIL
5259 FROM address for the random test, but not for the subsequent recipient
5260 test. It is now used for both.
5262 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5263 patch removes a few documentation additions to RFC 3028, because the
5264 latest draft now contains them. It adds the new en;ascii-case comparator
5265 and a new error check for 8bit text in MIME parts. Comparator and
5266 require names are now matched exactly. I enabled the subaddress
5267 extension, but it is not well tested yet (read: it works for me)."
5269 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5270 rework some of the code of TK/09 above to avoid the hardwired use of
5271 "%lld" and "long long". Replaced the call to snprintf() with a call to
5274 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5275 records point to non-existent hosts", "retry timeout exceeded", and
5276 "retry time not reached for any host after a long failure period".
5278 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5279 experimental DomainKeys support:
5281 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5282 (2) On an error such as an illegally used "control", the wrong name for
5283 the control was given.
5285 These problems did NOT occur unless DomainKeys support was compiled.
5287 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5289 PH/24 Added ${if match_ip condition.
5291 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5292 databases so that it will be absolutely obvious if a crash occurs in the
5293 DB library. This is a regular occurrence (often caused by mis-matched
5296 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5297 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5298 picked up on a user's system that detects such things. There doesn't seem
5299 to be a gcc warning option for this - only an attribute that has to be
5300 put on the function's prototype. It seems that in Fedora Core 4 they have
5301 set this on a number of new functions. No doubt there will be more in due
5304 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5305 specify a transport. However, if an address that was verified by such a
5306 router was the subject of a callout, Exim crashed because it tried to
5307 read the rcpt_include_affixes from the non-existent transport. Now it
5308 just assumes that the setting of that option is false. This bug was
5309 introduced by 4.51/PH/31.
5311 PH/28 Changed -d+all to exclude +memory, because that information is very
5312 rarely of interest, but it makes the output a lot bigger. People tend to
5313 do -d+all out of habit.
5315 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5316 code in os-type was giving problems when libc.so lives in lib64, like on
5319 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5320 aren't the modern standard, and it seems that some systems' include files
5321 don't always have them. Exim was already checking for some of the newer
5322 ones like T_AAAA, and defining it itself. I've added checks for all the
5323 record types that Exim uses.
5325 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5326 not automatically generating a new one, as it is supposed to. This
5327 prevented TLS from working. If the file did exist, but contained invalid
5328 data, a new version was generated, as expected. It was only the case of a
5329 non-existent file that was broken.
5331 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5332 with a change in libdomainkeys > 0.64.
5334 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5335 from DNS. If the selector record carries the flag, it now has
5336 precedence over the domain-wide flag.
5338 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5340 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5341 the use of a port name that isn't defined in /etc/services) occurred, the
5342 message was deferred as in a normal delivery, and thus remained on the
5343 spool, instead of being failed because of the mua_wrapper setting. This
5344 is now fixed, and I tidied up some of the mua_wrapper messages at the
5347 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5348 the messages in a hash of arrays rather than using individual hashes.
5349 This is a bit cleaner and results in dramatic memory savings, albeit
5350 at a slight CPU cost.
5352 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5353 as requested by Marc Sherman.
5355 SC/10 Eximstats: added histograms for user specified patterns as requested
5358 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5360 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5361 fopen() in the content-scanning modules that did not already have it.
5367 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5368 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5370 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5372 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5374 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5375 to test Sieve filters that use "vacation".
5377 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5378 that changes the way the GnuTLS parameters are stored in the cache file.
5379 The new format can be generated externally. For backward compatibility,
5380 if the data in the cache doesn't make sense, Exim assumes it has read an
5381 old-format file, and it generates new data and writes a new file. This
5382 means that you can't go back to an older release without removing the
5385 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5386 work if there are any delivery delays because "one_time" forces the
5387 parent to be marked "delivered", so its unseen clone is never tried
5388 again. For this reason, Exim now forbids the simultaneous setting of
5391 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5392 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5393 entry if you want to know the details. The fix, however, neglected to
5394 consider the case when local delivery batching is involved. The test for
5395 "previously delivered" was not happening when checking to see if an
5396 address could be batched with a previous (undelivered) one; under
5397 certain circumstances this could lead to multiple deliveries to the same
5400 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5401 in its include files, and this causes problems building Exim.
5403 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5404 header_syntax) but Exim was just ignoring anything given after a slash.
5405 In particular, this caused confusion with an attempt to use "verify =
5406 reverse_host_lookup/defer_ok". An error is now given when options are
5407 supplied for verify items that do not have them. (Maybe reverse_host_
5408 lookup should have a defer_ok option, but that's a different point.)
5410 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5411 defined by RFC 821) to 2048, because there were problems with some AUTH
5412 commands, and RFC 1869 says the size should be increased for extended
5413 SMTP commands that take arguments.
5415 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5418 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5419 "unknown" error; now it says that the functionality isn't in the binary.
5421 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5422 an address' error message when a string expansion fails (syntax or
5423 whatever). Otherwise the password may appear in the log. Following change
5424 PH/42 below, there is no longer a chance of it appearing in a bounce
5427 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5429 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5430 of its addresses was ever tried. (Bugzilla bug #2.)
5432 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5433 the result incorrectly in the debug output. (It correctly added a newline
5434 to what was transported.)
5436 TF/01 Added $received_time.
5438 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5439 commented out examples of how to interface to a virus scanner and to
5440 SpamAssassin. Also added commented examples of av_scanner and
5441 spamd_address settings.
5443 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5444 and controls are allowed in which ACLs. There were a couple of minor
5445 errors. Some of the entries in the conditions table (which is a table of
5446 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5447 negation of where the condition IS allowed.
5449 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5451 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5452 header file does not have a version number, so I've had to invent a new
5453 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5454 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5455 radiusclient), but it was contributed by a Radius user.
5457 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5458 files or directories, for OpenSSL.
5460 PH/19 When an Exim process that is running as root has to create an Exim log
5461 file, it does so in a subprocess that runs as exim:exim so as to get the
5462 ownership right at creation (otherwise, other Exim processes might see
5463 the file with the wrong ownership). There was no test for failure of this
5464 fork() call, which would lead to the process getting stuck as it waited
5465 for a non-existent subprocess. Forks do occasionally fail when resources
5466 run out. I reviewed all the other calls to fork(); they all seem to check
5469 PH/20 When checking for unexpected SMTP input at connect time (before writing
5470 the banner), Exim was not dealing correctly with a non-positive return
5471 from the read() function. If the client had disconnected by this time,
5472 the result was a log entry for a synchronization error with an empty
5473 string after "input=" when read() returned zero. If read() returned -1
5474 (an event I could not check), uninitialized data bytes were printed.
5475 There were reports of junk text (parts of files, etc) appearing after
5478 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5480 PH/22 Added support for macro redefinition, and (re)definition in between
5481 driver and ACL definitions.
5483 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5484 forgetting to use the resulting value; it was using the unexpanded value.
5486 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5487 hadn't been configured. The fix is from Juergen Kreileder, who
5488 understands it better than I do:
5490 "Here's what I see happening with three configured cyrus_sasl
5491 authenticators configured (plain, login, cram-md5):
5493 On startup auth_cyrus_sasl_init() gets called for each of these.
5494 This means three calls to sasl_listmech() without a specified mech_list.
5495 => SASL tests which mechs of all available mechs actually work
5496 => three warnings about OTP not working
5497 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5499 With the patch, sasl_listmech() also gets called three times. But now
5500 SASL's mech_list option is set to the server_mech specified in the the
5501 authenticator. Or in other words, the answer from sasl_listmech()
5502 gets limited to just the mech you're testing for (which is different
5504 => the return list contains just 'plain' or 'login', 'cram-md5' or
5505 nothing depending on the value of ob->server_mech.
5507 I've just tested the patch: Authentication still works fine,
5508 unavailable mechs specified in the exim configuration are still
5509 caught, and the auth.log warnings about OTP are gone."
5511 PH/25 When debugging is enabled, the contents of the command line are added
5512 to the debugging output, even when log_selector=+arguments is not
5515 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5516 answer is "GNU", and only if the return is "GNU/something" is the answer
5519 PH/27 $acl_verify_message is now set immediately after the failure of a
5520 verification in an ACL, and so is available in subsequent modifiers. In
5521 particular, the message can be preserved by coding like this:
5523 warn !verify = sender
5524 set acl_m0 = $acl_verify_message
5526 Previously, $acl_verify_message was set only while expanding "message"
5527 and "log_message" when a very denied access.
5529 PH/28 Modified OS/os.c-Linux with
5531 -#ifndef OS_LOAD_AVERAGE
5532 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5534 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5535 nomenclature these days.)
5537 PH/29 Installed patch from the Sieve maintainer that adds the options
5538 sieve_useraddress and sieve_subaddress to the redirect router.
5540 PH/30 In these circumstances:
5541 . Two addresses routed to the same list of hosts;
5542 . First host does not offer TLS;
5543 . First host accepts first address;
5544 . First host gives temporary error to second address;
5545 . Second host offers TLS and a TLS session is established;
5546 . Second host accepts second address.
5547 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5548 and peerdn, if requested) that were in fact used only for the second
5551 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5552 attention to any local part prefix or suffix that was matched by the
5553 router that accepted the address. It now behaves in the same way as it
5554 does for delivery: the affixes are removed from the local part unless
5555 rcpt_include_affixes is set on the transport.
5557 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5558 timeout during the DATA phase of an incoming message.
5560 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5561 applied a patch sanctioned by the Sieve maintainer.
5563 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5564 the uid or gid is negative. A case of a negative gid caused this to be
5565 noticed. The fix allows for either to be negative.
5567 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5568 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5571 PH/36 The change PH/12 above was broken. Fixed it.
5573 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5574 the grounds that routing the same address twice would always produce the
5575 same answer. This might have been true once, but it is certainly no
5576 longer true now. Routing a child address may depend on the previous
5577 routing that produced that child. Some complicated redirection strategies
5578 went wrong when messages had multiple recipients, and made Exim's
5579 behaviour dependent on the order in which the addresses were given.
5581 I have moved the duplicate checking until after the routing is complete.
5582 Exim scans the addresses that are assigned to local and remote
5583 transports, and removes any duplicates. This means that more work will be
5584 done, as duplicates will always all be routed, but duplicates are
5585 presumably rare, so I don't expect this is of any significance.
5587 For deliveries to pipes, files, and autoreplies, the duplicate checking
5588 still happens during the routing process, since they are not going to be
5591 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5592 It corrects a timeout issue with spamd. This is Ian's comment: "The
5593 background is that sometimes spamd either never reads data from a
5594 connection it has accepted, or it never writes response data. The exiscan
5595 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5596 blindly assumes that writes won't block so it may never time out."
5598 PH/39 Allow G after quota size as well as K and M.
5600 PH/40 The value set for $authenticated_id in an authenticator may not contain
5601 binary zeroes or newlines because the value is written to log lines and
5602 to spool files. There was no check on this. Now the value is run through
5603 the string_printing() function so that such characters are converted to
5604 printable escape sequences.
5606 PH/41 $message_linecount is a new variable that contains the total number of
5607 lines in the message. Compare $body_linecount, which is the count for the
5610 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5611 bounce and delay warning messages, except in certain special cases, which
5614 (a) An SMTP error message from a remote host;
5615 (b) A message specified in a :fail: redirection;
5616 (c) A message specified in a "fail" command in a system filter;
5617 (d) A message specified in a FAIL return from the queryprogram router;
5618 (e) A message specified by the cannot_route_message router option.
5620 In these cases only, Exim does include the error details in bounce and
5621 warning messages. There are also a few cases where bland messages such
5622 as "unrouteable address" or "local delivery error" are given.
5624 PH/43 $value is now also set for the "else" part of a ${run expansion.
5626 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5627 being worked on, but at least Exim now implements the latest version to
5630 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5631 process to complete was treated as a delivery failure, a timeout while
5632 writing the message to the pipe was logged, but erroneously treated as a
5633 successful delivery. Such timeouts include transport filter timeouts. For
5634 consistency with the overall process timeout, these timeouts are now
5635 treated as errors, giving rise to delivery failures by default. However,
5636 there is now a new Boolean option for the pipe transport called
5637 timeout_defer, which, if set TRUE, converts the failures into defers for
5638 both kinds of timeout. A transport filter timeout is now identified in
5641 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5642 systems where "make" and "gmake" are different, calling "gmake" at top
5643 level broke things. I've arranged for the value of $(MAKE) to be passed
5644 from the Makefile to this script so that it can call the same version of
5648 A note about Exim versions 4.44 and 4.50
5649 ----------------------------------------
5651 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5652 changes of various kinds. As a consequence, a big documentation update was
5653 needed. This delayed the release for rather longer than seemed good, especially
5654 in the light of a couple of (minor) security issues. Therefore, the changes
5655 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5656 release. So 4.44 and 4.50 are in effect two different branches that both start
5659 I have left the 4.50 change log unchanged; it contains all the changes since
5660 4.43. The change log for 4.44 is below; many of its items are identical to
5661 those for 4.50. This seems to be the most sensible way to preserve the
5662 historical information.
5668 1. Minor wording change to the doc/README.SIEVE file.
5670 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5671 computation of the current number of files was incorrect.
5673 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5674 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5677 4. Give more explanation in the error message when the command for a transport
5678 filter fails to execute.
5680 5. There are several places where Exim runs a non-Exim command in a
5681 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5682 was being done only for the command run by the queryprogram router. It is
5683 now done for all such subprocesses. The other cases are: ${run, transport
5684 filters, and the commands run by the lmtp and pipe transports.
5686 6. Added CONFIGURE_GROUP build-time option.
5688 7. Some older OS have a limit of 256 on the maximum number of file
5689 descriptors. Exim was using setrlimit() to set 1000 as a large value
5690 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5691 systems. I've change it so that if it can't get 1000, it tries for 256.
5693 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5694 was an oversight, and furthermore, ever since the addition of extra
5695 controls (e.g. 4.43/32), the checks on when to allow different forms of
5696 "control" were broken. There should now be diagnostics for all cases when a
5697 control that does not make sense is encountered.
5699 9. Added the /retain_sender option to "control=submission".
5701 10. $recipients is now available in the predata ACL (oversight).
5703 11. Tidy the search cache before the fork to do a delivery from a message
5704 received from the command line. Otherwise the child will trigger a lookup
5705 failure and thereby defer the delivery if it tries to use (for example) a
5706 cached ldap connection that the parent has called unbind on.
5708 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5709 of $address_data from the recipient verification was clobbered by the
5710 sender verification.
5712 13. The value of address_data from a sender verification is now available in
5713 $sender_address_data in subsequent conditions in the ACL statement.
5715 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5717 15. Added a new option "connect=<time>" to callout options, to set a different
5720 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5721 was its contents. (It was OK if the option was not defined at all.)
5723 17. A "Completed" log line is now written for messages that are removed from
5724 the spool by the -Mrm option.
5726 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5727 information about exactly what failed.
5729 19. Added -dd to debug only the daemon process.
5731 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5732 handles timeouts, both on the server side and network timeouts. Renamed the
5733 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5735 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5736 It is now set to "smtps".
5738 22. $host_address is now set to the target address during the checking of
5739 ignore_target_hosts.
5741 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5742 being passed; this would have caused $sender_host_name to have been used if
5743 matching the list had actually called for a host name (not very likely,
5744 since this list is usually IP addresses). A host name is now passed as
5747 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5748 code that specifies a non-restarting handler (typically sigaction() in
5749 modern systems) in an attempt to fix a rare and obscure crash bug.
5751 25. Narrowed the window for a race in the daemon that could cause it to ignore
5752 SIGCHLD signals. This is not a major problem, because they are used only to
5753 wake it up if nothing else does.
5755 26. A malformed maildirsize file could cause Exim to calculate negative values
5756 for the mailbox size or file count. Odd effects could occur as a result.
5757 The maildirsize information is now recalculated if the size or filecount
5760 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5761 support for a long time. Removed HAVE_SYS_VFS_H.
5763 28. Installed the latest version of exipick from John Jetmore.
5765 29. In an address list, if the pattern was not a regular expression, an empty
5766 subject address (from a bounce message) matched only if the pattern was an
5767 empty string. Non-empty patterns were not even tested. This was the wrong
5768 because it is perfectly reasonable to use an empty address as part of a
5769 database query. An empty address is now tested by patterns that are
5770 lookups. However, all the other forms of pattern expect the subject to
5771 contain a local part and a domain, and therefore, for them, an empty
5772 address still always fails if the pattern is not itself empty.
5774 30. Exim went into a mad DNS loop when attempting to do a callout where the
5775 host was specified on an smtp transport, and looking it up yielded more
5776 than one IP address.
5778 31. Re-factored the code for checking spool and log partition space into a
5779 function that finds that data and another that does the check. The former
5780 is then used to implement four new variables: $spool_space, $log_space,
5781 $spool_inodes, and $log_inodes.
5783 32. The RFC2047 encoding function was originally intended for short strings
5784 such as real names; it was not keeping to the 75-character limit for
5785 encoded words that the RFC imposes. It now respects the limit, and
5786 generates multiple encoded words if necessary. To be on the safe side, I
5787 have increased the buffer size for the ${rfc2047: expansion operator from
5790 33. It is now permitted to omit both strings after an "if" condition; if the
5791 condition is true, the result is "true". As before, when the second string
5792 is omitted, a false condition yields an empty string. This makes it less
5793 cumbersome to write custom ACL and router conditions.
5795 34. Failure to deliver a bounce message always caused it to be frozen, even if
5796 there was an errors_to setting on the router. The errors_to setting is now
5799 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5800 canonical form (fully expanded) before being placed in
5801 $sender_host_address.
5803 36. The table in the code that translates DNS record types into text (T_A to
5804 "A" for instance) was missing entries for NS and CNAME. It is just possible
5805 that this could have caused confusion if both these types were looked up
5806 for the same domain, because the text type is used as part of Exim's
5807 per-process caching. But the chance of anyone hitting this buglet seems
5810 37. The dnsdb lookup has been extended in a number of ways.
5812 (1) There is a new type, "zns", which walks up the domain tree until it
5813 finds some nameserver records. It should be used with care.
5815 (2) There is a new type, "mxh", which is like "mx" except that it returns
5816 just the host names, not the priorities.
5818 (3) It is now possible to give a list of domains (or IP addresses) to be
5819 looked up. The behaviour when one of the lookups defers can be
5820 controlled by a keyword.
5822 (4) It is now possible to specify the separator character for use when
5823 multiple records are returned.
5825 38. The dnslists ACL condition has been extended: it is now possible to supply
5826 a list of IP addresses and/or domains to be looked up in a particular DNS
5829 39. Added log_selector=+queue_time_overall.
5831 40. When running the queue in the test harness, wait just a tad after forking a
5832 delivery process, to get repeatability of debugging output.
5834 41. Include certificate and key file names in error message when GnuTLS fails
5835 to set them up, because the GnuTLS error message doesn't include the name
5836 of the failing file when there is a problem reading it.
5838 42. Allow both -bf and -bF in the same test run.
5840 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5842 44. The "Exiscan patch" is now merged into the mainline Exim source.
5844 45. Sometimes the final signoff response after QUIT could fail to get
5845 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5846 < 0 before doing a fflush(). This bug looks as though it goes back to the
5847 introduction of TLS in release 3.20, but "sometimes" must have been rare
5848 because the tests only now provoked it.
5850 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5851 (this can affect the format of dates).
5853 47. exim_tidydb, when checking for the continued existence of a message for
5854 which it has found a message-specific retry record, was not finding
5855 messages that were in split spool directories. Consequently, it was
5856 deleting retry records that should have stayed in existence.
5858 48. Steve fixed some bugs in eximstats.
5860 49. The SPA authentication driver was not abandoning authentication and moving
5861 on to the next authenticator when an expansion was forced to fail,
5862 contradicting the general specification for all authenticators. Instead it
5863 was generating a temporary error. It now behaves as specified.
5865 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5866 (the order specifies the preference for clients). The order is now AES256,
5867 AES128, 3DES, ARCFOUR128.
5869 51. Small patch to Sieve code - explicitly set From: when generating an
5872 52. Exim crashed if a remote delivery caused a very long error message to be
5873 recorded - for instance if somebody sent an entire SpamAssassin report back
5874 as a large number of 550 error lines. This bug was coincidentally fixed by
5875 increasing the size of one of Exim's internal buffers (big_buffer) that
5876 happened as part of the Exiscan merge. However, to be on the safe side, I
5877 have made the code more robust (and fixed the comments that describe what
5880 53. Now that there can be additional text after "Completed" in log lines (if
5881 the queue_time_overall log selector is set), a one-byte patch to exigrep
5882 was needed to allow it to recognize "Completed" as not the last thing in
5885 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5886 patch that reportedly fixes this has been added. I am not expert enough to
5887 create a test for it. This is what the patch creator wrote:
5889 "I found a little strange behaviour of ldap code when working with
5890 Windows 2003 AD Domain, where users was placed in more than one
5891 Organization Units. When I tried to give exim partial DN, the exit code
5892 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5893 But simultaneously result of request was absolutely normal ldap result,
5894 so I produce this patch..."
5896 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5897 REFERENCE, so I have modified the code to exclude the patch when that macro
5900 55. Some experimental protocols are using DNS PTR records for new purposes. The
5901 keys for these records are domain names, not reversed IP addresses. The
5902 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5903 leaves it alone. Component reversal etc. now happens only for IP addresses.
5906 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5908 57. Double the size of the debug message buffer (to 2048) so that more of very
5909 long debug lines gets shown.
5911 58. The exicyclog utility now does better if the number of log files to keep
5912 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5914 59. Two changes related to the smtp_active_hostname option:
5916 (1) $smtp_active_hostname is now available as a variable.
5917 (2) The default for smtp_banner uses $smtp_active_hostname instead
5918 of $primary_hostname.
5920 60. The host_aton() function is supposed to be passed a string that is known
5921 to be a valid IP address. However, in the case of IPv6 addresses, it was
5922 not checking this. This is a hostage to fortune. Exim now panics and dies
5923 if the condition is not met. A case was found where this could be provoked
5924 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5925 components; fortuitously, this particular loophole had already been fixed
5926 by change 4.50/55 above.
5928 If there are any other similar loopholes, the new check in host_aton()
5929 itself should stop them being exploited. The report I received stated that
5930 data on the command line could provoke the exploit when Exim was running as
5931 exim, but did not say which command line option was involved. All I could
5932 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5933 running as the user.
5936 61. There was a buffer overflow vulnerability in the SPA authentication code
5937 (which came originally from the Samba project). I have added a test to the
5938 spa_base64_to_bits() function which I hope fixes it.
5941 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5942 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5944 63. The daemon start-up calls getloadavg() while still root for those OS that
5945 need the first call to be done as root, but it missed one case: when
5946 deliver_queue_load_max is set with deliver_drop_privilege. This is
5947 necessary for the benefit of the queue runner, because there is no re-exec
5948 when deliver_drop_privilege is set.
5950 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5951 This has been fixed.
5953 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5954 was in use, was not putting the data itself into the right store pool;
5955 consequently, it could be overwritten for a subsequent message in the same
5956 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5959 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5961 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5962 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5963 the function were treating the return as a boolean value, which happened to
5964 work because 0=false and not-0=true, but is not correct code.
5966 68. The host_aton() function was not handling scoped IPv6 addresses (those
5967 with, for example, "%eth0" on the end) correctly.
5969 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5970 negated items (that is, ~something) in unsigned ints. Some compilers
5971 apparently mutter when there is no cast.
5973 70. If an address verification called from an ACL failed, and did not produce a
5974 user-specific message (i.e. there was only a "system" message), nothing was
5975 put in $acl_verify_message. In this situation, it now puts the system
5978 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5979 session; change 4.31/43 added the unwanted input to the log line - except
5980 that it did not do this in the start of session case. It now does.
5982 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5983 This is wrong and can cause the other end to generate a synchronization
5984 error if it is another Exim or anything else that does the synchronization
5985 check. A QUIT command is no longer sent after a timeout.
5987 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5988 during host lookups.
5990 74. The defer_ok option of callout verification was not working if it was used
5991 when verifying addresses in header lines, that is, for this case:
5993 verify = header_sender/callout=defer_ok
5995 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5996 those file descriptors could be used for SMTP connections. If anything
5997 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5998 could be sent to the SMTP client, causing chaos. The daemon now opens
5999 stdin, stdout, and stderr to /dev/null when it puts itself into the
6002 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6003 log by default. The user can override this with suitable Perl magic.
6005 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6006 the log message when discard triggers, was not working for the DATA ACL or
6007 for the non-SMTP ACL.
6009 78. Error message wording change in sieve.c.
6011 79. If smtp_accept_max_per_host was set, the number of connections could be
6012 restricted to fewer than expected, because the daemon was trying to set up
6013 a new connection before checking whether the processes handling previous
6014 connections had finished. The check for completed processes is now done
6015 earlier. On busy systems, this bug wouldn't be noticed because something
6016 else would have woken the daemon, and it would have reaped the completed
6019 80. If a message was submitted locally by a user whose login name contained one
6020 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6021 It caused a spool format error. I have fixed the spool reading code. A
6022 related problem was that the "from" clause in the Received: line became
6023 illegal because of the space(s). It is now covered by ${quote_local_part.
6025 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6028 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6030 83. Patch from Sieve maintainer to fix unterminated string problem in
6031 "vacation" handling.
6033 84. Some minor changes to the Linux configuration files to help with other
6034 OS variants using glibc.
6036 85. One more patch for Sieve to update vacation handling to latest spec.
6039 ----------------------------------------------------
6040 See the note above about the 4.44 and 4.50 releases.
6041 ----------------------------------------------------
6047 1. Change 4.43/35 introduced a bug that caused file counts to be
6048 incorrectly computed when quota_filecount was set in an appendfile
6051 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6052 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6055 3. Give more explanation in the error message when the command for a transport
6056 filter fails to execute.
6058 4. There are several places where Exim runs a non-Exim command in a
6059 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6060 was being done only for the command run by the queryprogram router. It is
6061 now done for all such subprocesses. The other cases are: ${run, transport
6062 filters, and the commands run by the lmtp and pipe transports.
6064 5. Some older OS have a limit of 256 on the maximum number of file
6065 descriptors. Exim was using setrlimit() to set 1000 as a large value
6066 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6067 systems. I've change it so that if it can't get 1000, it tries for 256.
6069 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6070 was an oversight, and furthermore, ever since the addition of extra
6071 controls (e.g. 4.43/32), the checks on when to allow different forms of
6072 "control" were broken. There should now be diagnostics for all cases when a
6073 control that does not make sense is encountered.
6075 7. $recipients is now available in the predata ACL (oversight).
6077 8. Tidy the search cache before the fork to do a delivery from a message
6078 received from the command line. Otherwise the child will trigger a lookup
6079 failure and thereby defer the delivery if it tries to use (for example) a
6080 cached ldap connection that the parent has called unbind on.
6082 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6083 of $address_data from the recipient verification was clobbered by the
6084 sender verification.
6086 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6087 was its contents. (It was OK if the option was not defined at all.)
6089 11. A "Completed" log line is now written for messages that are removed from
6090 the spool by the -Mrm option.
6092 12. $host_address is now set to the target address during the checking of
6093 ignore_target_hosts.
6095 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6096 being passed; this would have caused $sender_host_name to have been used if
6097 matching the list had actually called for a host name (not very likely,
6098 since this list is usually IP addresses). A host name is now passed as
6101 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6102 code that specifies a non-restarting handler (typically sigaction() in
6103 modern systems) in an attempt to fix a rare and obscure crash bug.
6105 15. Narrowed the window for a race in the daemon that could cause it to ignore
6106 SIGCHLD signals. This is not a major problem, because they are used only to
6107 wake it up if nothing else does.
6109 16. A malformed maildirsize file could cause Exim to calculate negative values
6110 for the mailbox size or file count. Odd effects could occur as a result.
6111 The maildirsize information is now recalculated if the size or filecount
6114 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6115 support for a long time. Removed HAVE_SYS_VFS_H.
6117 18. Updated exipick to current release from John Jetmore.
6119 19. Allow an empty sender to be matched against a lookup in an address list.
6120 Previously the only cases considered were a regular expression, or an
6123 20. Exim went into a mad DNS lookup loop when doing a callout where the
6124 host was specified on the transport, if the DNS lookup yielded more than
6127 21. The RFC2047 encoding function was originally intended for short strings
6128 such as real names; it was not keeping to the 75-character limit for
6129 encoded words that the RFC imposes. It now respects the limit, and
6130 generates multiple encoded words if necessary. To be on the safe side, I
6131 have increased the buffer size for the ${rfc2047: expansion operator from
6134 22. Failure to deliver a bounce message always caused it to be frozen, even if
6135 there was an errors_to setting on the router. The errors_to setting is now
6138 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6139 canonical form (fully expanded) before being placed in
6140 $sender_host_address.
6142 24. Updated eximstats to version 1.33
6144 25. Include certificate and key file names in error message when GnuTLS fails
6145 to set them up, because the GnuTLS error message doesn't include the name
6146 of the failing file when there is a problem reading it.
6148 26. Expand error message when OpenSSL has problems setting up cert/key files.
6151 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6152 (this can affect the format of dates).
6154 28. exim_tidydb, when checking for the continued existence of a message for
6155 which it has found a message-specific retry record, was not finding
6156 messages that were in split spool directories. Consequently, it was
6157 deleting retry records that should have stayed in existence.
6159 29. eximstats updated to version 1.35
6160 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6161 1.35 - bugfix such that pie charts by volume are generated correctly
6163 30. The SPA authentication driver was not abandoning authentication and moving
6164 on to the next authenticator when an expansion was forced to fail,
6165 contradicting the general specification for all authenticators. Instead it
6166 was generating a temporary error. It now behaves as specified.
6168 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6169 (the order specifies the preference for clients). The order is now AES256,
6170 AES128, 3DES, ARCFOUR128.
6172 31. Small patch to Sieve code - explicitly set From: when generating an
6175 32. Exim crashed if a remote delivery caused a very long error message to be
6176 recorded - for instance if somebody sent an entire SpamAssassin report back
6177 as a large number of 550 error lines. This bug was coincidentally fixed by
6178 increasing the size of one of Exim's internal buffers (big_buffer) that
6179 happened as part of the Exiscan merge. However, to be on the safe side, I
6180 have made the code more robust (and fixed the comments that describe what
6183 33. Some experimental protocols are using DNS PTR records for new purposes. The
6184 keys for these records are domain names, not reversed IP addresses. The
6185 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6186 leaves it alone. Component reversal etc. now happens only for IP addresses.
6189 34. The host_aton() function is supposed to be passed a string that is known
6190 to be a valid IP address. However, in the case of IPv6 addresses, it was
6191 not checking this. This is a hostage to fortune. Exim now panics and dies
6192 if the condition is not met. A case was found where this could be provoked
6193 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6194 components; fortuitously, this particular loophole had already been fixed
6195 by change 4.50/55 or 4.44/33 above.
6197 If there are any other similar loopholes, the new check in host_aton()
6198 itself should stop them being exploited. The report I received stated that
6199 data on the command line could provoke the exploit when Exim was running as
6200 exim, but did not say which command line option was involved. All I could
6201 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6202 running as the user.
6205 35. There was a buffer overflow vulnerability in the SPA authentication code
6206 (which came originally from the Samba project). I have added a test to the
6207 spa_base64_to_bits() function which I hope fixes it.
6210 36. The daemon start-up calls getloadavg() while still root for those OS that
6211 need the first call to be done as root, but it missed one case: when
6212 deliver_queue_load_max is set with deliver_drop_privilege. This is
6213 necessary for the benefit of the queue runner, because there is no re-exec
6214 when deliver_drop_privilege is set.
6216 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6217 was in use, was not putting the data itself into the right store pool;
6218 consequently, it could be overwritten for a subsequent message in the same
6219 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6222 38. Sometimes the final signoff response after QUIT could fail to get
6223 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6224 < 0 before doing a fflush(). This bug looks as though it goes back to the
6225 introduction of TLS in release 3.20, but "sometimes" must have been rare
6226 because the tests only now provoked it.
6232 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6233 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6234 Now it returns an integer. A number of calls were still expecting a T/F
6235 return. Fortuitously, in all cases, the tests worked in OK situations,
6236 which is the norm. However, things would have gone wrong on any write
6237 failures on the smtp file descriptor. This function is used when sending
6238 messages over SMTP and also when doing verify callouts.
6240 2. When Exim is called to do synchronous delivery of a locally submitted
6241 message (the -odf or -odi options), it no longer closes stderr before doing
6244 3. Implemented the mua_wrapper option.
6246 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6248 5. Implemented the functions header_remove(), header_testname(),
6249 header_add_at_position(), and receive_remove_recipient(), and exported them
6252 6. If an ACL "warn" statement specified the addition of headers, Exim already
6253 inserted X-ACL-Warn: at the start if there was no header name. However, it
6254 was not making this test for the second and subsequent header lines if
6255 there were newlines in the string. This meant that an invalid header could
6256 be inserted if Exim was badly configured.
6258 7. Allow an ACL "warn" statement to add header lines at the start or after all
6259 the Received: headers, as well as at the end.
6261 8. Added the rcpt_4xx retry error code.
6263 9. Added postmaster_mailfrom=xxx to callout verification option.
6265 10. Added mailfrom=xxxx to the callout verification option, for verify=
6268 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6269 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6271 12. Inserted some casts to stop certain compilers warning when using pointer
6272 differences as field lengths or precisions in printf-type calls (mostly
6273 affecting debugging statements).
6275 13. Added optional readline() support for -be (dynamically loaded).
6277 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6278 same clock tick as a message's arrival, so that its received time was the
6279 same as the "first fail" time on the retry record, and that message
6280 remained on the queue past the ultimate address timeout, every queue runner
6281 would try a delivery (because it was past the ultimate address timeout) but
6282 after another failure, the ultimate address timeout, which should have then
6283 bounced the address, did not kick in. This was a "< instead of <=" error;
6284 in most cases the first failure would have been in the next clock tick
6285 after the received time, and all would be well.
6287 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6288 being recognized when the domain list was tested by the match_domain
6289 condition in an expansion string.
6291 16. Added the ${str2b64: operator.
6293 17. Exim was always calling setrlimit() to set a large limit for the number of
6294 processes, without checking whether the existing limit was already
6295 adequate. (It did check for the limit on file descriptors.) Furthermore,
6296 errors from getrlimit() and setrlimit() were being ignored. Now they are
6297 logged to the main and panic logs, but Exim does carry on, to try to do its
6298 job under whatever limits there are.
6300 18. Imported PCRE 5.0.
6302 19. Trivial typo in log message " temporarily refused connection" (the leading
6305 20. If the log selector return_path_on_delivery was set and an address was
6306 redirected to /dev/null, the delivery process crashed because it assumed
6307 that a return path would always be set for a "successful" delivery. In this
6308 case, the whole delivery is bypassed as an optimization, and therefore no
6311 21. Internal re-arrangement: the function for sending a challenge and reading
6312 a response while authentication was assuming a zero-terminated challenge
6313 string. It's now changed to take a pointer and a length, to allow for
6314 binary data in such strings.
6316 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6318 23. Exim was not respecting finduser_retries when seeking the login of the
6319 uid under which it was called; it was always trying 10 times. (The default
6320 setting of finduser_retries is zero.) Also, it was sleeping after the final
6321 failure, which is pointless.
6323 24. Implemented tls_on_connect_ports.
6325 25. Implemented acl_smtp_predata.
6327 26. If the domain in control=submission is set empty, Exim assumes that the
6328 authenticated id is a complete email address when it generates From: or
6329 Sender: header lines.
6331 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6332 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6333 chown and chgrp in /bin and hostname in /usr/bin.
6335 28. Exim was keeping the "process log" file open after each use, just as it
6336 does for the main log. This opens the possibility of it remaining open for
6337 long periods when the USR1 signal hits a daemon. Occasional processlog
6338 errors were reported, that could have been caused by this. Anyway, it seems
6339 much more sensible not to leave this file open at all, so that is what now
6342 29. The long-running daemon process does not normally write to the log once it
6343 has entered its main loop, and it closes the log before doing so. This is
6344 so that log files can straightforwardly be renamed and moved. However,
6345 there are a couple of unusual error situations where the daemon does write
6346 log entries, and I had neglected to close the log afterwards.
6348 30. The text of an SMTP error response that was received during a remote
6349 delivery was being truncated at 512 bytes. This is too short for some of
6350 the long messages that one sometimes sees. I've increased the limit to
6353 31. It is now possible to make retry rules that apply only when a message has a
6354 specific sender, in particular, an empty sender.
6356 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6357 it possible to be selective about when SMTP synchronization is enforced.
6359 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6361 32. Implemented hosts_connection_nolog.
6363 33. Added an ACL for QUIT.
6365 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6368 35. Added mailbox_size and mailbox_filecount to appendfile.
6370 36. Added control = no_multiline_responses to ACLs.
6372 37. There was a bug in the logic of the code that waits for the clock to tick
6373 in the case where the clock went backwards by a substantial amount such
6374 that the microsecond fraction of "now" was more than the microsecond
6375 fraction of "then" (but the whole seconds number was less).
6377 38. Added support for the libradius Radius client library this is found on
6378 FreeBSD (previously only the radiusclient library was supported).
6384 1. When certain lookups returned multiple values in the form name=value, the
6385 quoting of the values was not always being done properly. Specifically:
6386 (a) If the value started with a double quote, but contained no whitespace,
6388 (b) If the value contained whitespace other than a space character (i.e.
6389 tabs or newlines or carriage returns) it was not quoted.
6390 This fix has been applied to the mysql and pgsql lookups by writing a
6391 separate quoting function and calling it from the lookup code. The fix
6392 should probably also be applied to nisplus, ibase and oracle lookups, but
6393 since I cannot test any of those, I have not disturbed their existing code.
6395 2. A hit in the callout cache for a specific address caused a log line with no
6396 reason for rejecting RCPT. Now it says "Previous (cached) callout
6397 verification failure".
6399 3. There was an off-by-one bug in the queryprogram router. An over-long
6400 return line was truncated at 256 instead of 255 characters, thereby
6401 overflowing its buffer with the terminating zero. As well as fixing this, I
6402 have increased the buffer size to 1024 (and made a note to document this).
6404 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6405 when Exim is waiting for an SMTP response from a remote server, Exim
6406 restarts its select() call on the socket, thereby resetting its timeout.
6407 This is not a problem when such interrupts are rare. Somebody set up a cron
6408 job to run exiwhat every 2 minutes, which is less than the normal select()
6409 timeout (5 or 10 minutes). This meant that the select() timeout never
6410 kicked in because it was always reset. I have fixed this by comparing the
6411 time when an interrupt arrives with the time at the start of the first call
6412 to select(). If more time than the timeout has elapsed, the interrupt is
6413 treated as a timeout.
6415 5. Some internal re-factoring in preparation for the addition of Sieve
6416 extensions (by MH). In particular, the "personal" test is moved to a
6417 separate function, and given an option for scanning Cc: and Bcc: (which is
6418 not set for Exim filters).
6420 6. When Exim created an email address using the login of the caller as the
6421 local part (e.g. when creating a From: or Sender: header line), it was not
6422 quoting the local part when it contained special characters such as @.
6424 7. Installed new OpenBSD configuration files.
6426 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6427 try to make them clearer.
6429 9. Callout options, other than the timeout value, were being ignored when
6430 verifying sender addresses in header lines. For example, when using
6432 verify = header_sender/callout=no_cache
6434 the cache was (incorrectly) being used.
6436 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6437 only the Cygwin environment.
6439 11. When return_path_on_delivery was set as a log selector, if different remote
6440 addresses in the same message used different return paths and parallel
6441 remote delivery occurred, the wrong values would sometimes be logged.
6442 (Whenever a remote delivery process finished, the return path value from
6443 the most recently started remote delivery process was logged.)
6445 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6446 lines when AUTH and/or TLS are in use. This is the "received protocol"
6447 field. Exim used to use "asmtp" for authenticated SMTP, without any
6448 indication (in the protocol name) for TLS use. Now it follows the RFC and
6449 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6450 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6451 names appear in log lines as well as in Received: header lines.
6453 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6454 extensions, and comparison tests, and to fix some bugs.
6456 14. Changes to the "personal" filter test:
6458 (1) The test was buggy in that it was just doing the equivalent of
6459 "contains" tests on header lines. For example, if a user's address was
6460 anne@some.where, the "personal" test would incorrectly be true for
6462 To: susanne@some.where
6464 This test is now done by extracting each address from the header in turn,
6465 and checking the entire address. Other tests that are part of "personal"
6466 are now done using regular expressions (for example, to check local parts
6467 of addresses in From: header lines).
6469 (2) The list of non-personal local parts in From: addresses has been
6470 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6471 taken from the Sieve specification recommendations.
6473 (3) If the message contains any header line starting with "List-" it is
6474 treated as non-personal.
6476 (4) The test for "circular" in the Subject: header line has been removed
6477 because it now seems ill-conceived.
6479 15. Minor typos in src/EDITME comments corrected.
6481 16. Installed latest exipick from John Jetmore.
6483 17. If headers_add on a router specified a text string that was too long for
6484 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6485 of string_sprintf() is now avoided.
6487 18. $message_body_size was not set (it was always zero) when running the DATA
6488 ACL and the local_scan() function.
6490 19. For the "mail" command in an Exim filter, no default was being set for
6491 the once_repeat time, causing a random time value to be used if "once" was
6492 specified. (If the value happened to be <= 0, no repeat happened.) The
6493 default is now 0s, meaning "never repeat". The "vacation" command was OK
6494 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6495 (I found it when inspecting the code).
6497 20. There is now an overall timeout for performing a callout verification. It
6498 defaults to 4 times the callout timeout, which applies to individual SMTP
6499 commands during the callout. The overall timeout applies when there is more
6500 than one host that can be tried. The timeout is checked before trying the
6501 next host. This prevents very long delays if there are a large number of
6502 hosts and all are timing out (e.g. when the network connections are timing
6503 out). The value of the overall timeout can be changed by specifying an
6504 additional sub-option for "callout", called "maxwait". For example:
6506 verify = sender/callout=5s,maxwait=20s
6508 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6509 to the end before writing, but this should make it even safer).
6511 22. Exim was forgetting that it had advertised PIPELINING for the second and
6512 subsequent messages on an SMTP connection. It was also not resetting its
6513 memory on STARTTLS and an internal HELO.
6515 23. When Exim logs an SMTP synchronization error within a session, it now
6516 records whether PIPELINING has been advertised or not.
6518 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6519 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6520 rather than long int.
6522 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6524 26. Added the never_mail option to autoreply.
6530 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6531 crash if the getsockname() call failed; this can happen if a connection is
6532 closed very soon after it is established. The problem was simply in the
6533 order in which certain operations were done, causing Exim to try to write
6534 to the SMTP stream before it had set up the file descriptor. The bug has
6535 been fixed by making things happen in the correct order.
6541 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6542 before the connection was closed, thus losing the rejection response.
6544 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6545 some early Solaris releases, but causes trouble in current releases where
6546 socklen_t is defined.
6548 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6551 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6554 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6555 that releasing the top of it at the end releases what was used for sub-
6556 expansions (unless the block got too big). However, discard this block if
6557 the first thing is a variable or header, so that we can use its block when
6558 it is dynamic (useful for very large $message_headers, for example).
6560 6. Lookups now cache *every* query, not just the most recent. A new, separate
6561 store pool is used for this. It can be recovered when all lookup caches are
6562 flushed. Lookups now release memory at the end of their result strings.
6563 This has involved some general refactoring of the lookup sources.
6565 7. Some code has been added to the store_xxx() functions to reduce the amount
6566 of flapping under certain conditions.
6568 8. log_incoming_interface used to affect only the <= reception log lines. Now
6569 it causes the local interface and port to be added to several more SMTP log
6570 lines, for example "SMTP connection from", and rejection lines.
6572 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6574 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6576 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6577 could be overwritten at the end of the current message (or the start of a
6578 new message if it was set in a HELO ACL). The value is now preserved for
6579 the duration of the SMTP connection.
6581 12. If a transport had a headers_rewrite setting, and a matching header line
6582 contained an unqualified address, that address was qualified, even if it
6583 did not match any rewriting rules. The underlying bug was that the values
6584 of the flags that permit the existence of unqualified sender and recipient
6585 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6586 non-local messages, and by -bnq for local messages) were not being
6587 preserved with the message after it was received.
6589 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6590 "next input=" as part of the text comprising the host identity instead of
6591 the correct text. The code was using the same buffer for two different
6592 strings. However, depending on which order the printing function evaluated
6593 its arguments, the bug did not always show up. Under Linux, for example, my
6594 test suite worked just fine.
6596 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6597 doesn't work with some older versions of Perl. It has been changed to "my",
6598 which in any case is probably the better facility to use.
6600 15. A really picky compiler found some instances of statements for creating
6601 error messages that either had too many or two few arguments for the format
6604 16. The size of the buffer for calls to the DNS resolver has been increased
6605 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6606 for addresses that have a lot of PTR records. This alleviates a problem; it
6607 does not fully solve it.
6609 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6610 buffer now truncates the list and logs the incident, which is the same
6611 action as happens when Exim is looking up a host name and its aliases.
6612 Previously in this situation something unpredictable would happen;
6613 sometimes it was "internal error: store_reset failed".
6615 18. If a server dropped the connection unexpectedly when an Exim client was
6616 using GnuTLS and trying to read a response, the client delivery process
6617 crashed while trying to generate an error log message.
6619 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6620 string, for example:
6622 warn message = H1: something\nH2: something
6624 the text was added as a single header line from Exim's point of view
6625 though it ended up OK in the delivered message. However, searching for the
6626 second and subsequent header lines using $h_h2: did not work. This has been
6627 fixed. Similarly, if a system filter added multiple headers in this way,
6628 the routers could not see them.
6630 20. Expanded the error message when iplsearch is called with an invalid key to
6631 suggest using net-iplsearch in a host list.
6633 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6634 ACLs are no longer actually imposed (and a message to that effect is
6637 22. If a "gecos" field in a passwd entry contained escaped characters, in
6638 particular, if it contained a \" sequence, Exim got it wrong when building
6639 a From: or a Sender: header from that name. A second bug also caused
6640 incorrect handling when an unquoted " was present following a character
6641 that needed quoting.
6643 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6644 was not being matched caselessly.
6646 24. Arranged for all hyphens in the exim.8 source to be escaped with
6649 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6650 properly. Recipient callout cache records were still being keyed to include
6651 the sender, even when use_sender was set false. This led to far more
6652 callouts that were necessary. The sender is no longer included in the key
6653 when use_sender is false.
6655 26. Added "control = submission" modifier to ACLs.
6657 27. Added the ${base62d: operator to decode base 62 numbers.
6659 28. dnsdb lookups can now access SRV records.
6661 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6662 the configuration file.
6664 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6665 (-v) mode. This makes the output for a verbose queue run more intelligible.
6667 31. Added a use_postmaster feature to recipient callouts.
6669 32. Added the $body_zerocount variable, containing the number of binary zero
6670 bytes in the message body.
6672 33. The time of last modification of the "new" subdirectory is now used as the
6673 "mailbox time last read" when there is a quota error for a maildir
6676 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6678 35. Added +ignore_unknown as a special item in host lists.
6680 36. Code for decoding IPv6 addresses in host lists is now included, even if
6681 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6682 address was recognized as an IP address, but was then not correctly decoded
6683 into binary, causing unexpected and incorrect effects when compared with
6690 1. Very minor rewording of debugging text in manualroute to say "list of
6691 hosts" instead of "hostlist".
6693 2. If verify=header_syntax was set, and a header line with an unqualified
6694 address (no domain) and a large number of spaces between the end of the
6695 name and the colon was received, the reception process suffered a buffer
6696 overflow, and (when I tested it) crashed. This was caused by some obsolete
6697 code that should have been removed. The fix is to remove it!
6699 3. When running in the test harness, delay a bit after writing a bounce
6700 message to get a bit more predictability in the log output.
6702 4. Added a call to search_tidyup() just before forking a reception process. In
6703 theory, someone could use a lookup in the expansion of smtp_accept_max_
6704 per_host which, without the tidyup, could leave open a database connection.
6706 5. Added the variables $recipient_data and $sender_data which get set from a
6707 lookup success in an ACL "recipients" or "senders" condition, or a router
6708 "senders" option, similar to $domain_data and $local_part_data.
6710 6. Moved the writing of debug_print from before to after the "senders" test
6713 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6714 problems for message scanning, either using a data ACL, or using
6715 local_scan() because the Received: header was not generated till after they
6716 were called (in order to set the time as the time of reception completion).
6717 I have revised the way this works. The header is now generated after the
6718 body is received, but before the ACL or local_scan() are called. After they
6719 are run, the timestamp in the header is updated.
6725 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6726 before starting a queue runner without re-exec. This happened only when
6727 deliver_drop_privilege was set or when the Exim user was set to root. The
6728 effect of the bug was that timeouts during subsequent deliveries caused
6729 crashes instead of being properly handled. The handler is now left at its
6730 default (and expected) setting.
6732 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6733 message, again when deliver_drop_privilege is set or Exim is run as root.
6734 The bug described in (1) was not present in this case, but the tidying up
6735 of the other signals was missing. I have made the two cases consistent.
6737 3. The ignore_target_hosts setting on a manualroute router was being ignored
6738 for hosts that were looked up using the /MX notation.
6740 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6743 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6744 operated on the sender address. After changing the $sender_address to <>
6745 for the sender address verify, Exim was re-instated it as the original
6746 (before rewriting) address, but remembering that it had rewritten it, so it
6747 wasn't rewriting it again. This bug also had the effect of breaking the
6748 sender address verification caching when the sender address was rewritten.
6750 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6751 This has been changed so that if the ip literal address matches
6752 ignore_target_hosts, the router declines.
6754 7. Added expansion conditions match_domain, match_address, and match_local_
6755 part (NOT match_host).
6757 8. The placeholder for the Received: header didn't have a length field set.
6759 9. Added code to Exim itself and to exim_lock to test for a specific race
6760 condition that could lead to file corruption when using MBX delivery. The
6761 issue is with the lockfile that is created in /tmp. If this file is removed
6762 after a process has opened it but before that process has acquired a lock,
6763 there is the potential for a second process to recreate the file and also
6764 acquire a lock. This could lead to two Exim processes writing to the file
6765 at the same time. The added code performs the same test as UW imapd; it
6766 checks after acquiring the lock that its file descriptor still refers to
6767 the same named file.
6769 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6770 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6773 11. Added the smtp_active_hostname option. If used, this will typically be made
6774 to depend on the incoming interface address. Because $interface_address is
6775 not set up until the daemon has forked a reception process, error responses
6776 that can happen earlier (such as "too many connections") no longer contain
6779 12. If an expansion in a condition on a "warn" statement fails because a lookup
6780 defers, the "warn" statement is abandoned, and the next ACL statement is
6781 processed. Previously this caused the whole ACL to be aborted.
6783 13. Added the iplsearch lookup type.
6785 14. Added ident_timeout as a log selector.
6787 15. Added tls_certificate_verified as a log selector.
6789 16. Added a global option tls_require_ciphers (compare the smtp transport
6790 option of the same name). This controls incoming TLS connections.
6792 17. I finally figured out how to make tls_require_ciphers do a similar thing
6793 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6794 before starting the TLS session.
6796 18. Tabs are now shown as \t in -bP output.
6798 19. If the log selector return_path_on_delivery was set, Exim crashed when
6799 bouncing a message because it had too many Received: header lines.
6801 20. If two routers both had headers_remove settings, and the first one included
6802 a superfluous trailing colon, the final name in the first list and the
6803 first name in the second list were incorrectly joined into one item (with a
6804 colon in the middle).
6810 1. Added -C and -D options to the exinext utility, mainly to make it easier
6811 to include in the automated testing, but these could be helpful when
6812 multiple configurations are in use.
6814 2. The exinext utility was not formatting the output nicely when there was
6815 an alternate port involved in the retry record key, nor when there was a
6816 message id as well (for retries that were specific to a specific message
6817 and a specific host). It was also confused by IPv6 addresses, because of
6818 the additional colons they contain. I have fixed the IPv4 problem, and
6819 patched it up to do a reasonable job for IPv6.
6821 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6822 delivery, the log line now contains "pipelined" if PIPELINING was used.
6824 4. An SMTP transport process used to panic and die if the bind() call to set
6825 an explicit outgoing interface failed. This has been changed; it is now
6826 treated in the same way as a connect() failure.
6828 5. A reference to $sender_host_name in the part of a conditional expansion
6829 that was being skipped was still causing a DNS lookup. This no longer
6832 6. The def: expansion condition was not recognizing references to header lines
6833 that used bh_ and bheader_.
6835 7. Added the _cache feature to named lists.
6837 8. The code for checking quota_filecount in the appendfile transport was
6838 allowing one more file than it should have been.
6840 9. For compatibility with Sendmail, the command line option
6848 and sets the incoming protocol and host name (for trusted callers). The
6849 host name and its colon can be omitted when only the protocol is to be set.
6850 Note the Exim already has two private options, -pd and -ps, that refer to
6851 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6852 "s", but I don't think that's a major issue.
6854 10. A number of refactoring changes to the code, none of which should affect
6857 (a) The number of logging options was getting close to filling up the
6858 32-bit word that was used as a bit map. I have split them into two classes:
6859 those that are passed in the argument to log_write(), and those that are
6860 only ever tested independently outside of that function. These are now in
6861 separate 32-bit words, so there is plenty of room for expansion again.
6862 There is no change in the user interface or the logging behaviour.
6864 (b) When building, for example, log lines, the code previously used a
6865 macro that called string_cat() twice, in order to add two strings. This is
6866 not really sufficiently general. Furthermore, there was one instance where
6867 it was actually wrong because one of the argument was used twice, and in
6868 one call a function was used. (As it happened, calling the function twice
6869 did not affect the overall behaviour.) The macro has been replaced by a
6870 function that can join an arbitrary number of extra strings onto a growing
6873 (c) The code for expansion conditions now uses a table and a binary chop
6874 instead of a serial search (which was left over from when there were very
6875 few conditions). Also, it now recognizes conditions like "pam" even when
6876 the relevant support is not compiled in: a suitably worded error message is
6877 given if an attempt is made to use such a condition.
6879 11. Added ${time_interval:xxxxx}.
6881 12. A bug was causing one of the ddress fields not to be passed back correctly
6882 from remote delivery subprocesses. The field in question was not being
6883 subsequently used, so this caused to problems in practice.
6885 13. Added new log selectors queue_time and deliver_time.
6887 14. Might have fixed a bug in maildirsizefile handling that threw up
6888 "unexpected character" debug warnings, and recalculated the data
6889 unnecessarily. In any case, I expanded the warning message to give more
6892 15. Added the message "Restricted characters in address" to the statements in
6893 the default ACL that block characters like @ and % in local parts.
6895 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6896 Three changes have been made:
6898 (a) There was a serious bug; a negative response to MAIL caused the whole
6899 recipient domain to be cached as invalid, thereby blocking all messages
6900 to all local parts at the same domain, from all senders. This bug has
6901 been fixed. The domain is no longer cached after a negative response to
6902 MAIL if the sender used is not empty.
6904 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6907 (c) A new callout option, "use_sender" has been added for people who want
6908 the modified behaviour.
6914 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6917 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6918 indeed breaks things for older releases.
6920 3. Added additional logging to the case where there is a problem reading data
6921 from a filter that is running in a subprocess using a pipe, in order to
6922 try to track down a specific problem.
6924 4. Testing facility fudge: when running in the test harness and attempting
6925 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6926 getting "No route to host". Convert this to a timeout.
6928 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6931 6. Some OS don't have socklen_t but use size_t instead. This affects the
6932 fifth argument of getsockopt() amongst other things. This is now
6933 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6934 can be set for individual OS. I have set it for SunOS5, OSF1, and
6935 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6936 some earlier ones do not.
6938 7. Change 4.30/15 was not doing the test caselessly.
6940 8. The standard form for an IPv6 address literal was being rejected by address
6941 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6942 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6943 this, as well as the form without the "IPv6" on the front (but only when
6944 address literals are enabled, of course).
6946 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6948 10. Exim crashed if a message with an empty sender address specified by -f
6949 encountered a router with an errors_to setting. This could be provoked only
6950 by a command such as
6954 where an empty string was supplied; "<>" did not hit this bug.
6956 11. Installed PCRE release 4.5.
6958 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6959 remained set. It is now erased.
6961 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6962 times from message ids (which are base 36 rather than the normal 62).
6964 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6965 were being counted as actual protocol errors, and logged if the log
6966 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6967 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6968 and DATA following a set of rejected RCPTs do not count as protocol errors.
6969 In other words, Exim assumes they were pipelined, though this may not
6970 actually be the case. Of course, in all cases the client gets an
6971 appropriate error code.
6973 15. If a lookup fails in an ACL condition, a message about the failure may
6974 be available; it is used if testing the ACL cannot continue, because most
6975 such messages specify what the cause of the deferral is. However, some
6976 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6977 that caused an old message to be retained and used if a later statement
6978 caused a defer, replacing the real cause of the deferral.
6980 16. If an IP address had so many PTR records that the DNS lookup buffer
6981 was not large enough to hold them, Exim could crash while trying to process
6982 the truncated data. It now detects and logs this case.
6984 17. Further to 4.21/58, another change has been made: if (and only if) the
6985 first line of a message (the first header line) ends with CRLF, a bare LF
6986 in a subsequent header line has a space inserted after it, so as not to
6987 terminate the header.
6989 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6990 unnecessarily, used atoi() instead of strtol(), and didn't check the
6991 termination when getting file sizes from file names by regex.
6993 19. Completely re-implemented the support for maildirsize files, in the light
6994 of a number of problems with the previous contributed implementation
6995 (4.30/29). In particular:
6997 . If the quota is zero, the maildirsize file is maintained, but no quota is
7000 . If the maildir directory does not exist, it is created before any attempt
7001 to write a maildirsize file.
7003 . The quota value in the file is just a cache; if the quota is changed in
7004 the transport, the new value overrides.
7006 . A regular expression is available for excluding directories from the
7009 20. The autoreply transport checks the characters in options that define the
7010 message's headers; it allows continued headers, but it was checking with
7011 isspace() after an embedded newline instead of explicitly looking for a
7014 21. If all the "regular" hosts to which an address was routed had passed their
7015 expiry times, and had not reached their retry times, the address was
7016 bounced, even if fallback hosts were defined. Now Exim should go on to try
7019 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7020 equivalent code in the SMTP transport. Some hosts send humungous responses
7021 to HELO/EHLO, more than 1024 it seems.
7023 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7024 gives compiler warnings in some environments. I've now done it "properly",
7027 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7028 (because of problems with the built-in one) was declared to return uschar *
7029 instead of char *, causing compiler failure.
7031 25. Fixed a file descriptor leak when processing alias/forward files.
7033 26. Fixed a minor format string issue in dbfn.c.
7035 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7037 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7038 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7041 29. When a sender address is verified, it is cached, to save repeating the test
7042 when there is more than one recipient in a message. However, when the
7043 verification involves a callout, it is possible for different callout
7044 options to be set for different recipients. It is too complicated to keep
7045 track of this in the cache, so now Exim always runs a verification when a
7046 callout is required, relying on the callout cache for the optimization.
7047 The overhead is duplication of the address routing, but this should not be
7050 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7051 to be verified with callout=postmaster, and the main callout worked but the
7052 postmaster check failed, the verification correctly failed. However, if a
7053 subsequent RCPT command asked for sender verification *without* the
7054 postmaster check, incorrect caching caused this verification also to fail,
7057 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7058 it was not caching the DNS options (qualify_single, search_parents) that
7059 were used when the lookup failed. A subsequent lookup with different
7060 options therefore always gave the same answer, though there were cases
7061 where it should not have. (Example: a "domains = !$mx_any" option on a
7062 dnslookup router: the "domains" option is always processed without any
7063 widening, but the router might have qualify_single set.) Now Exim uses the
7064 cached value only when the same options are set.
7066 32. Added John Jetmore's "exipick" utility to the distribution.
7068 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7069 than a timeout (e.g. a certificate is required, and is not provided), an
7070 Exim server now closes the connection immediately. Previously it waited for
7071 the client to close - but if the client is SSL, it seems that they each
7072 wait for each other, leading to a delay before one of them times out.
7074 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7075 maintained 0.8.x compatibility because I don't think many are using it, and
7076 it is clearly obsolete.
7078 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7081 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7082 client certificate was expired. A simple patch fixes this, though I don't
7083 understand the full logic of why the verify callback is called multiple
7086 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7087 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7088 which causes problems with some clients (such as the Certicom SSL Plus
7089 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7090 disables the coutermeasure allowing Eudora to connect."
7092 38. Exim was not checking that a write() to a log file succeeded. This could
7093 lead to Bad Things if a log got too big, in particular if it hit a file
7094 size limit. Exim now panics and dies if it cannot write to a log file, just
7095 as it does if it cannot open a log file.
7097 39. Modified OS/Makefile-Linux so that it now contains
7099 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7101 The two -D definitions ensure that Exim is compiled with large file
7102 support, which makes it possible to handle log files that are bigger than
7105 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7106 instance) a domain was checked against a named list that involved a lookup,
7107 causing $domain_data to be set, then another domain was checked against the
7108 same list, then the first domain was re-checked, the value of $domain_data
7109 after the final check could be wrong. In particular, if the second check
7110 failed, it could be set empty. This bug probably also applied to
7113 41. The strip_trailing_dot option was not being applied to the address given
7114 with the -f command-line option.
7116 42. The code for reading a message's header from the spool was incrementing
7117 $received_count, but never initializing it. This meant that the value was
7118 incorrect (doubled) while delivering a message in the same process in which
7119 it was received. In the most common configuration of Exim, this never
7120 happens - a fresh exec is done - but it can happen when
7121 deliver_drop_privilege is set.
7123 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7124 it now includes up to 150 characters of the unexpected data in the log
7127 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7128 and building data strings. The size of both of these buffers was 10 000
7129 bytes - far larger than anybody would *ever* want, thought I. Needless to
7130 say, somebody hit the limit. I have increased the maximum line length to
7131 20 000 and the maximum data length of concatenated lines to 100 000. I have
7132 also fixed two bugs, because there was no checking on these buffers. Tsk,
7133 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7134 buffer is too small.
7136 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7137 lsearch lookups. Now it does.
7139 46. When parsing a route_list item in a manualroute router, a fixed-length
7140 buffer was used for the list of hosts. I made this 1024 bytes long,
7141 thinking that nobody would ever have a list of hosts that long. Wrong.
7142 Somebody had a whole pile of complicated expansion conditions, and the
7143 string was silently truncated, leading to an expansion error. It turns out
7144 that it is easier to change to an unlimited length (owing to other changes
7145 that have happened since this code was originally written) than to build
7146 structure for giving a limitation error. The length of the item that
7147 expands into the list of hosts is now unlimited.
7149 47. The lsearch lookup could not handle data where the length of text line was
7150 more than 4095 characters. Such lines were truncated, leading to shortened
7151 data being returned. It should now handle lines of any length.
7153 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7154 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7157 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7158 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7159 no longer happens in comment lines. A list of replacements is now placed
7160 at the head of all of the source files, except those whose only change is
7161 to replace PERL_COMMAND in the very first #! line.
7163 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7164 messages on the queue, with a bottom-up merge sort, using code contributed
7165 by Michael Haardt. This should make operations like -bp somewhat faster on
7166 large queues. It won't affect queue runners, except when queue_run_in_order
7169 51. Installed eximstats 1.31 in the distribution.
7171 52. Added support for SRV lookups to the dnslookup router.
7173 53. If an ACL referred to $message_body or $message_body_end, the value was not
7174 reset for any messages that followed in the same SMTP session.
7176 54. The store-handling optimization for building very long strings was not
7177 differentiating between the different store pools. I don't think this
7178 actually made any difference in practice, but I've tidied it.
7180 55. While running the routers to verify a sender address, $sender_address
7181 was still set to the sender address. This is wrong, because when routing to
7182 send a bounce to the sender, it would be empty. Therefore, I have changed
7183 it so that, while verifying a sender address, $sender_address is set to <>.
7184 (There is no change to what happens when verifying a recipient address.)
7186 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7187 target A or AAAA records (if not already returned) without resetting the
7188 qualify_single or search_parents options of the DNS resolver. These are
7189 inappropriate in this case because the targets of MX and SRV records must
7190 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7191 target that, when qualified, matched something in the local domain. These
7192 two options are now turned off when doing these lookups.
7194 57. It seems that at least some releases of Reiserfs (which does not have the
7195 concept of a fixed number of inodes) returns zero and not -1 for the
7196 number of available inodes. This interacted badly with check_spool_inodes,
7197 which assumed that -1 was the "no such thing" setting. What I have done is
7198 to check that the total number of inodes is greater than zero before doing
7199 the test of how many are available.
7201 58. When a "warn" ACL statement has a log_message modifier, the message is
7202 remembered, and not repeated. This is to avoid a lot of repetition when a
7203 message has many recipients that cause the same warning to be written.
7204 However, Exim was preserving the list of already written lines for an
7205 entire SMTP session, which doesn't seem right. The memory is now reset if a
7206 new message is started.
7208 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7209 debugging output unless log_rewrite was also set.
7211 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7212 of (int)(handle) when we know that handle contains (void *)(-1).
7214 61. The Exim daemon panic-logs an error return when it closes the incoming
7215 connection. However "connection reset by peer" seems to be common, and
7216 isn't really an error worthy of noting specially, so that particular error
7219 62. When Exim is trying to find all the local interfaces, it used to panic and
7220 die if the ioctl to get the interface flags failed. However, it seems that
7221 on at least one OS (Solaris 9) it is possible to have an interface that is
7222 included in the list of interfaces, but for which you get a failure error
7223 for this call. This happens when the interface is not "plumbed" into a
7224 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7225 failure of the "get flags" call assumes that the interface is down.
7227 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7228 makes life easier for people who are doing arithmetic on fields extracted
7229 from dates, where you often get leading zeros that should not be
7230 interpreted as octal.
7232 64. Added qualify_domain to the redirect router, to override the global
7235 65. If a pathologically long header line contained very many addresses (the
7236 report of this problem mentioned 10 000) and each of them was rewritten,
7237 Exim could use up a very large amount of memory. (It kept on making new
7238 copies of the header line as it rewrote, and never released the old ones.)
7239 At the expense of a bit more processing, the header rewriting function has
7240 been changed so that it no longer eats memory in this way.
7242 66. The generation of the Received: header has been moved from the time that a
7243 message starts to be received, to the time that it finishes. The timestamp
7244 in the Received: header should now be very close to that of the <= log
7245 line. There are two side-effects of this change:
7247 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7248 logged header lines no longer include the local Received: line, because
7249 it has not yet been created. The same applies to a copy of the message
7250 that is returned to a non-SMTP sender when a message is rejected.
7252 (b) When a filter file is tested using -bf, no additional Received: header
7253 is added to the test message. After some thought, I decided that this
7256 This change does not affect the value of $received_for. It is still set
7257 after address rewriting, but before local_scan() is called.
7259 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7261 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7262 gave an unhelpful panic error message, and a defer error. I have managed to
7263 change this behaviour so that it now rejects any supplied certificate,
7264 which seems right, as the list of acceptable certificates is empty.
7266 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7267 gave an unhelpful defer error. I have not managed to make this reject any
7268 supplied certificates, but the error message it gives is "no certificate
7269 supplied", which is not helpful.
7271 70. exigrep's output now also includes lines that are not associated with any
7272 message, but which match the given pattern. Implemented by a patch from
7273 Martin Sluka, which also tidied up the Perl a bit.
7275 71. Recipient callout verification, like sender verification, was using <> in
7276 the MAIL FROM command. This isn't really the right thing, since the actual
7277 sender may affect whether the remote host accepts the recipient or not. I
7278 have changed it to use the actual sender in the callout; this means that
7279 the cache record is now keyed on a recipient/sender pair, not just the
7280 recipient address. There doesn't seem to be a real danger of callout loops,
7281 since a callout by the remote host to check the sender would use <>.
7282 [SEE ABOVE: changed after hitting problems.]
7284 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7285 temporary errors. However, in the case of such a code being given after
7286 the end of a data transmission (i.e. after ".") Exim was failing to write
7287 a retry record for the message. (Yes, there was some broken host that was
7288 actually sending 8xx at this point.)
7290 73. An unknown lookup type in a host list could cause Exim to panic-die when
7291 the list was checked. (An example that provoked this was putting <; in the
7292 middle of a list instead of at the start.) If this happened during a DATA
7293 ACL check, a -D file could be left lying around. This kind of configuration
7294 error no longer causes Exim to die; instead it causes a defer error. The
7295 incident is still logged to the main and panic logs.
7297 74. Buglet left over from Exim 3 conversion. The message "too many messages
7298 in one connection" was written to the rejectlog but not the mainlog, except
7299 when address rewriting (yes!) was being logged.
7301 75. Added write_rejectlog option.
7303 76. When a system filter was run not as root (that is, when system_filter_user
7304 was set), the values of the $n variables were not being returned to the
7305 main process; thus, they were not subsequently available in the $sn
7308 77. Added +return_path_on_delivery log selector.
7310 78. A connection timeout was being treated differently from recipients deferred
7311 when testing hosts_max_try with a message that was older than the host's
7312 retry timeout. (The host should not be counted, thus allowing all hosts to
7313 be tried at least once before bouncing.) This may have been the cause of an
7314 occasionally reported bug whereby a message would remain on the queue
7315 longer than the retry timeout, but would be bounced if a delivery was
7316 forced. I say "may" because I never totally pinned down the problem;
7317 setting up timeout/retry tests is difficult. See also the next item.
7319 79. The ultimate address timeout was not being applied to errors that involved
7320 a combination of host plus message (for example, a timeout on a MAIL
7321 command). When an address resolved to a number of possible hosts, and they
7322 were not all tried for each delivery (e.g. because of hosts_max_try), a
7323 message could remain on the queue longer than the retry timeout.
7325 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7328 81. Fixed an obscure SMTP outgoing bug which required at least the following
7329 conditions: (a) there was another message waiting for the same server;
7330 (b) the server returned 5xx to all RCPT commands in the first message so
7331 that the message was not completed; (c) the server dropped the connection
7332 or gave a negative response to the RSET that Exim sends to abort the
7333 transaction. The observed case was a dropped connection after DATA that had
7334 been sent in pipelining mode. That is, the server had advertised PIPELINING
7335 but was not implementing it correctly. The effect of the bug was incorrect
7336 behaviour, such as trying another host, and this could lead to a crash.
7342 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7343 and daemon.c were passed as pointers to ints; they should have been
7344 pointers to socklen_t variables (which are typically unsigned ints).
7346 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7349 3. Fixed a really odd bug that affected only the testing scheme; patching a
7350 certain fixed string in the binary changed the value of another string that
7351 happened to be identical to the end of the original first string.
7353 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7354 name", it returns that address as the IP address. On some operating
7355 systems (e.g. Solaris), it also passes back the IP address string as the
7356 "host name". However, on others (e.g. Linux), it passes back an empty
7357 string. Exim wasn't checking for this, and was changing the host name to an
7358 empty string, assuming it had been canonicalized.
7360 5. Although rare, it is permitted to have more than one PTR record for a given
7361 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7362 all the names associated with an address, because they do in Solaris.
7363 However, it seems that they do not in Linux for data that comes from the
7364 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7365 I found this out when I moved to a new Linux workstation and tried to run
7366 the Exim test suite.
7368 To get round this problem I have changed the code so that it now does its
7369 own call to the DNS to look up PTR records when searching for a host name.
7370 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7371 addresses that are only in /etc/hosts are still found.
7373 This behaviour is, however, controlled by an option called host_lookup_
7374 order, which defaults to "bydns:byaddr". If people want to use the other
7375 order, or indeed, just use one or the other means of lookup, they can
7376 specify it in this variable.
7378 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7379 some operating systems, this comes back from gethostbyaddr() as an empty
7380 string, and this is what Exim used to test for. However, it seems that in
7381 other systems, "." is yielded. Exim now tests for this case too.
7383 7. The values of check_spool_space and check_log_space are now held internally
7384 as a number of kilobytes instead of an absolute number of bytes. If a
7385 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7386 kilobyte. This means that much larger values can be stored.
7388 8. Exim monitor: an attempt to get the action menu when not actually pointing
7389 at a message produces an empty menu entitled "No message selected". This
7390 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7391 no entries in it ("Shell widget menu has zero width and/or height"). So I
7392 have added a single, blank menu entry in this case.
7394 9. Added ${quote_local_part.
7396 10. MIME decoding is now applied to the contents of Subject: header lines when
7399 11. Now that a reference to $sender_host_address automatically causes a reverse
7400 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7401 host lookup before query-style lookups in lists that might use this
7402 variable. This has therefore been abolished, and the "net-" prefix is no
7403 longer necessary for query-style lookups.
7405 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7406 have been a typo for LFLAGS, so it has been changed.
7408 13. The install script calls Exim with "-C /dev/null" in order to find the
7409 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7410 to be output. However, since Exim outputs its version number before the
7411 error, it didn't break the script. It just looked ugly. I fixed this by
7412 always allowing "-C /dev/null" if the caller is root.
7414 14. Ignore overlarge ACL variable number when reading spool file - insurance
7415 against a later release with more variables having written the file.
7417 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7418 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7419 this, as well as the form without the "IPv6" on the front.
7421 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7422 OS/Makefile-Darwin file.
7424 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7425 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7427 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7428 variable. However, some OS use ints for inodes. I've added cast to long int
7429 to get rid of the compiler warning.
7431 19. I had forgotten to lock out "/../" in configuration file names when
7432 ALT_CONFIG_PREFIX was set.
7434 20. Routers used for verification do not need to specify transports. However,
7435 if such a router generated a host list, and callout was configured, Exim
7436 crashed, because it could not find a port number from the (non-existent)
7437 transport. It now assumes port 25 in this circumstance.
7439 21. Added the -t option to exigrep.
7441 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7442 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7443 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7444 them is set, LOOKUP_LSEARCH is forced.
7446 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7448 24. Added sender and host information to the "rejected by local_scan()" log
7449 line; previously there was no indication of these.
7451 25. Added .include_if_exists.
7453 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7454 Linux. It turns out that not all file systems support this. Apparently some
7455 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7456 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7457 not supported on the file descriptor, is now ignored when Exim is trying to
7458 sync a directory. This applies only to Linux.
7460 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7462 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7463 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7466 29. Added support for maildirsize files from supplied patch (modified a bit).
7468 30. The use of :fail: followed by an empty string could lead Exim to respond to
7469 sender verification failures with (e.g.):
7471 550 Verification failed for <xxx>
7472 550 Sender verify failed
7474 where the first response line was missing the '-' that indicates it is not
7475 the final line of the response.
7477 31. The loop for finding the name of the user that called Exim had a hardwired
7478 limit of 10; it now uses the value of finduser_retries, which is used for
7479 all other user lookups.
7481 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7484 33. Exim was neglecting to zero errno before one call of strtol() when
7485 expanding a string and expecting an integer value. On some systems this
7486 resulted in spurious "integer overflow" errors. Also, it was casting the
7487 result into an int without checking.
7489 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7490 did not work. The code looks as if it has *never* worked, though it appears
7491 to have been documented since at least release 1.62. I have made it work.
7493 35. The "timeout_DNS" error in retry rules, also documented since at least
7494 1.62, also never worked. As it isn't clear exactly what this means, and
7495 clearly it isn't a major issue, I have abolished the feature by treating it
7496 as "timeout", and writing a warning to the main and panic logs.
7498 36. The display of retry rules for -brt wasn't always showing the error code
7501 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7502 timeout_connect_A, timeout_connect_MX.
7504 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7505 to the empty sender.
7507 39. The daemon was not analysing the content of -oX till after it had closed
7508 stderr and disconnected from the controlling terminal. This meant that any
7509 syntax errors were only noted on the panic log, and the return code from
7510 the command was 0. By re-arranging the code a little, I've made the
7511 decoding happen first, so such errors now appear on stderr, and the return
7512 code is 1. However, the actual setting up of the sockets still happens in
7513 the disconnected process, so errors there are still only recorded on the
7516 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7517 connections (as happens on some IP stacks) was logged at start up time as
7518 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7519 it from "IPv6 and IPv4", which means that two separate sockets are being
7522 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7523 says whether AF_INET or AF_INET6 was passed as an argument.
7525 42. Exiwhat output was messed up when time zones were included in log
7528 43. Exiwhat now gives more information about the daemon's listening ports,
7529 and whether -tls-on-connect was used.
7531 44. The "port" option of the smtp transport is now expanded.
7533 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7534 silently ignored. Now an error message is written to the main and panic
7537 46. There's a new ACL modifier called "logwrite" which writes to a log file
7538 as soon as it is encountered.
7540 47. Added $local_user_uid and $local_user_gid at routing time.
7542 48. Exim crashed when trying to verify a sender address that was being
7545 49. Exim was recognizing only a space character after ".include". It now also
7546 recognizes a tab character.
7548 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7549 extracting the relevant information from the specification. The man page no
7550 longer contains scrambled data for the -d option, and I've added a section
7551 at the front about calling Exim under different names.
7553 51. Added "extra_headers" argument to the "mail" command in filter files.
7555 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7558 53. Installed eximstats 1.29.
7560 54. Added transport_filter_timeout as a generic transport option.
7562 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7563 Cc: header lines. This was required by RFC 822, but it not required by RFC
7566 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7567 incoming messages that did not have them. Now it does so only if the
7568 message originates locally, that is, if there is no associated remote host
7569 address. When Resent- header lines are present, this applies to the Resent-
7570 lines rather than the non-Resent- lines.
7572 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7573 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7575 58. Messages for configuration errors now include the name of the main
7576 configuration file - useful now that there may be more than one file in a
7577 list (.included file names were always shown).
7579 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7580 for those rare installations that do not start the daemon as root or run it
7581 setuid root. I've cut out the call to initgroups() if the daemon is not
7584 60. The Exim user and group can now be bound into the binary as text strings
7585 that are looked up at the start of Exim's processing.
7587 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7589 62. Added $mailstore_basename variable.
7591 63. Installed patch to sieve.c from Michael Haardt.
7593 64. When Exim failed to open the panic log after failing to open the main log,
7594 the original message it was trying to log was written to stderr and debug
7595 output, but if they were not available (the usual case in production), it
7596 was lost. Now it is written to syslog before the two lines that record the
7597 failures to open the logs.
7599 65. Users' Exim filters run in subprocesses under the user's uid. It is
7600 possible for a "deliver" command or an alias in a "personal" command to
7601 provoke an address rewrite. If logging of address rewriting is configured,
7602 this fails because the process is not running as root or exim. There may be
7603 a better way of dealing with this, but for the moment (because 4.30 needs
7604 to be released), I have disabled address rewrite logging when running a
7605 filter in a non-root, non-exim process.
7611 1. The buildconfig auxiliary program wasn't quoting the value set for
7612 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7613 not defined. This bug was masked in 4.22 by the effect that was fixed in
7616 2. Some messages that were rejected after a message id was allocated were
7617 shown as "incomplete" by exigrep. It no longer does this for messages that
7618 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7620 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7621 have allow_domain_literals set, the ID did not get logged in the <= line.
7622 Domain literals are now always recognized in Message-ID: header lines.
7624 4. The first argument for a ${extract expansion item is the key name or field
7625 number. Leading and trailing spaces in this item were not being ignored,
7626 causing some misleading effects.
7628 5. When deliver_drop_privilege was set, single queue runner processes started
7629 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7630 same command in the process it spins off) were not dropping privilege.
7632 6. When the daemon running as "exim" started a queue runner, it always
7633 re-executed Exim in the spun-off process. This is a waste of effort when
7634 deliver_drop_privilege is set. The new process now just calls the
7635 queue-runner function directly.
7641 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7644 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7645 ignored. Though the use of -oP was forcing the writing of a pid file, it
7646 was always written to the default place.
7648 3. If the message "no IP address found for host xxxx" is generated during
7649 incoming verification, it is now followed by identification of the incoming
7650 connection (so you can more easily find what provoked it).
7652 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7654 5. Added some features to "harden" Exim a bit more against certain attacks:
7656 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7657 be put in Local/Makefile. This is like the never_users runtime option,
7658 but it cannot be overridden. The default setting is "root".
7660 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7661 prefix string with which any file named in a -C command line option
7664 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7665 is retained for -C and -D only if the caller of Exim is root. Without
7666 it, the exim user may also use -C and -D and retain privilege.
7668 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7669 command line option is disabled.
7671 6. Macro names set by the -D option must start with an upper case letter, just
7672 like macro names defined in the configuration file.
7674 7. Added "dereference=" facility to LDAP.
7676 8. Two instances of the typo "uknown" in the source files are fixed.
7678 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7679 the Configure-Makefile script screwed up while processing it.
7681 10. Incorporated PCRE 4.4.
7683 11. The SMTP synchronization check was not operating right at the start of an
7684 SMTP session. For example, it could not catch a HELO sent before the client
7685 waited for the greeting. There is now a check for outstanding input at the
7686 point when the greeting is written. Because of the duplex, asynchronous
7687 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7688 way, but not yet received, when the check is performed.
7690 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7691 on TCP/IP sockets, because this apparently causes some broken clients to
7694 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7695 unchanged) from the Cygwin maintainer.
7697 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7698 was supported instead of testing for mbx. Effectively a typo.
7700 15. The spa authenticator server code was not checking that the input it
7701 received was valid base64.
7703 16. The debug output line for the "set" modifier in ACLs was not showing the
7704 name of the variable that was being set.
7706 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7708 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7709 lookup on its own. Something else was needed to trigger the lookup. For
7710 example, a match in host_lookup or the need for a host name in a host list.
7711 Now, if $sender_host_name is referenced and the host name has not yet been
7712 looked up, a lookup is performed. If the lookup fails, the variable remains
7713 empty, and $host_lookup_failed is set to "1".
7715 19. Added "eqi" as a case-independent comparison operator.
7717 20. The saslauthd authentication condition could segfault if neither service
7718 nor realm was specified.
7720 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7721 error message that was logged was misleading, and incoming SMTP
7722 connections were dropped. The message is now more accurate, and temporary
7723 errors are given to SMTP connections.
7725 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7726 (except RSET) with a 421 error, until QUIT is received. However, it was
7727 failing to send a response to QUIT.
7729 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7730 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7731 ACL is now run after the helo_try_verify_hosts code.
7733 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7734 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7735 case-independent, so other case variants are also recognized). Apparently
7736 some systems use these upper case variants.
7738 25. If more than two messages were waiting for the same host, and a transport
7739 filter was specified for the transport, Exim sent two messages over the
7740 same TCP/IP connection, and then failed with "socket operation on non-
7741 socket" when it tried to send the third.
7743 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7745 27. The extern definition of crypt16() in expand.c was not being excluded when
7746 the OS had its own crypt16() function.
7748 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7749 as a preferred synonym for return_size_limit, both as an option and as an
7752 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7754 30. Changed the default configuration ACL to relax the local part checking rule
7755 for addresses that are not in any local domains. For these addresses,
7756 slashes and pipe symbols are allowed within local parts, but the sequence
7757 /../ is explicitly forbidden.
7759 31. SPA server authentication was not clearing the challenge buffer before
7762 32. log_message in a "warn" ACL statement was writing to the reject log as
7763 well as to the main log, which contradicts the documentation and doesn't
7764 seem right (because no rejection is happening). So I have stopped it.
7766 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7767 However, I am unable to do any testing of this.
7769 34. Fixed an infelicity in the appendfile transport. When checking directories
7770 for a mailbox, to see if any needed to be created, it was accidentally
7771 using path names with one or more superfluous leading slashes; tracing
7772 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7774 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7775 contents are added to the log line that is written for every discarded
7776 recipient. (Previously a log_message setting was ignored.)
7778 36. The ${quote: operator now quotes the string if it is empty.
7780 37. The install script runs exim in order to find its version number. If for
7781 some reason other than non-existence or emptiness, which it checks, it
7782 could not run './exim', it was installing it with an empty version number,
7783 i.e. as "exim-". This error state is now caught, and the installation is
7786 38. An argument was missing from the function that creates an error message
7787 when Exim fails to connect to the socket for saslauthd authentication.
7788 This could cause Exim to crash, or give a corrupted message.
7790 39. Added isip, isip4, and isip6 to ${if conditions.
7792 40. The ACL variables $acl_xx are now saved with the message, and can be
7793 accessed later in routers, transports, and filters.
7795 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7796 strings in the file are not string-expanded.
7798 42. If a MAIL command specified a SIZE value that was too large to fit into an
7799 int variable, the check against message_size_limit failed. Such values are
7800 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7801 day this will have to be increased, but I don't think I want to be around
7802 when emails are that large.
7809 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7810 iconv() is not standard in FreeBSD.
7812 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7813 IPv6 enabled. The observed symptom was a segmentation fault on return from
7814 the function os_common_find_running_interfaces() in src/os.c.
7816 3. In the check_special_case() function in daemon.c I had used "errno" as an
7817 argument name, which causes warnings on some systems. This was basically a
7818 typo, since it was named "eno" in the comments!
7820 4. The code that waits for the clock to tick (at a resolution of some fraction
7821 of a second) so as to ensure message-id uniqueness was always waiting for
7822 at least one whole tick, when it could have waited for less. [This is
7823 almost certainly not relevant at current processor speeds, where it is
7824 unlikely to ever wait at all. But we try to future-proof.]
7826 5. The function that sleeps for a time interval that includes fractions of a
7827 second contained a race. It did not block SIGALRM between setting the
7828 timer, and suspending (a couple of lines later). If the interval was short
7829 and the sigsuspend() was delayed until after it had expired, the suspension
7830 never ended. On busy systems this could lead to processes getting stuck for
7833 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7834 process, before it forks any delivery processes. The open lookup caching
7835 mechanism meant that the open file or database connection was passed into
7836 the delivery process. The problem was that delivery processes always tidy
7837 up cached lookup data. This could cause a problem for the next delivery
7838 process started by the queue runner, because the external queue runner
7839 process does not know about the closure. So the next delivery process
7840 still has data in the lookup cache. In the case of a file lookup, there was
7841 no problem because closing a file descriptor in a subprocess doesn't affect
7842 the parent. However, if the lookup was caching a connection to a database,
7843 the connection was closed, and the second delivery process was likely to
7844 see errors such as "PGSQL: query failed: server closed the connection
7845 unexpectedly". The problem has been fixed by closing all cached lookups
7846 in a queue runner before running a delivery process.
7848 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7849 seem to have the "const" qualifier which it has on other OS. I've
7852 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7853 *of the same type* (client or server) with the same public name that an
7854 error should be diagnosed.
7856 9. When Exim looked up a host name for an IP address, but failed to find the
7857 original IP address when looking up the host name (a safety check), it
7858 output the message "<ip address> does not match any IP for NULL", which was
7859 confusing, to say the least. The bug was that the host name should have
7860 appeared instead of "NULL".
7862 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7863 user that is built into the binary, and the -C or -D options is used, root
7864 privilege is dropped before the configuration file is read. In addition,
7865 logging is switched to stderr instead of the normal log files. If the
7866 configuration then re-defines the Exim user, the unprivileged environment
7867 is probably not what is expected, so Exim logs a panic warning message (but
7870 However, if deliver_drop_privilege is set, the unprivileged state may well
7871 be exactly what is intended, so the warning has been cut out in that case,
7872 and Exim is allowed to try to write to its normal log files.
7878 1. smtp_return_error_details was not giving details for temporary sender
7879 or receiver verification errors.
7881 2. Diagnose a configuration error if two authenticators have the same public
7884 3. Exim used not to create the message log file for a message until the first
7885 delivery attempt. This could be confusing when incoming messages were held
7886 for policy or load reasons. The message log file is now created at the time
7887 the message is received, and an initial "Received" line is written to it.
7889 4. The automatically generated man page for command line options had a minor
7890 bug that caused no ill effects; however, a more serious problem was that
7891 the procedure for building the man page automatically didn't always
7892 operate. Consequently, release 4.20 contains an out-of-date version. This
7893 shouldn't happen again.
7895 5. When building Exim with embedded Perl support, the script that builds the
7896 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7897 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7899 6. The freeze_tell option was not being used for messages that were frozen on
7900 arrival, either by an ACL or by local_scan().
7902 7. Added the smtp_incomplete_transaction log selector.
7904 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7905 was accepting AUTH without a new EHLO.
7907 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7908 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7911 10. Logging of TCP/IP connections (when configured) now happens in the main
7912 daemon process instead of the child process, so that the TCP/IP connection
7913 count is more accurate (but it can never be perfect).
7915 11. The use of "drop" in a nested ACL was not being handled correctly in the
7916 outer ACL. Now, if condition failure induced by the nested "drop" causes
7917 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7918 or "require"), the connection is dropped.
7920 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7921 that yield "discard" can now be used with an "accept" or a "discard" verb,
7922 but an error is generated for any others (because I can't see a useful way
7923 to define what should happen).
7925 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7926 are now processed in the same way as lines in the Exim configuration file.
7927 In particular, continuation lines are supported.
7929 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7931 15. Added -ti meaning -t -i.
7933 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7934 domains, and warn by logging if others are found.
7936 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7937 returned from GIFCONF when Exim is trying to find the list of interfaces on
7938 a host. The code in os.c has been modified to copy each ifreq to an aligned
7939 structure in all cases.
7941 Also, in some cases, the returned ifreq's were being copied to a 'struct
7942 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7943 means the last couple of bytes of an IPv6 address could be chopped if the
7944 ifreq contained only a normal sockaddr (14 bytes storage).
7946 18. Named domain lists were not supported in the hosts_treat_as_local option.
7947 An entry such as +xxxx was not recognized, and was treated as a literal
7950 19. Ensure that header lines added by a DATA ACL are included in the reject log
7951 if the ACL subsequently rejects the message.
7953 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7954 MD5 (which is deprecated).
7956 21. When testing a filter file using -bf, Exim was writing a message when it
7957 took the sender from a "From " line in the message, but it was not doing so
7958 when it took $return_path from a Return-Path: header line. It now does.
7960 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7961 with a valid header line field name (a series of printing characters
7962 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7964 23. Changed "disc" in the source to "disk" to conform to the documentation and
7965 the book and for uniformity.
7967 24. Ignore Sendmail's -Ooption=value command line item.
7969 25. When execve() failed while trying to run a command in a pipe transport,
7970 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7971 could be confused with a return value of 69 from the command itself. This
7972 has been changed to 127, the value the shell returns if it is asked to run
7973 a non-existent command. The wording for the related log line suggests a
7974 non-existent command as the problem.
7976 26. If received_header_text expands to an empty string, do not add a Received:
7977 header line to the message. (Well, it adds a token one on the spool, but
7978 marks it "old" so that it doesn't get used or transmitted.)
7980 27. Installed eximstats 1.28 (addition of -nt option).
7982 28. There was no check for failure on the call to getsockname() in the daemon
7983 code. This can fail if there is a shortage of resources on the system, with
7984 ENOMEM, for example. A temporary error is now given on failure.
7986 29. Contrary to the C standard, it seems that in some environments, the
7987 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7988 program. Exim now does this explicitly; it affects the formatting of
7989 timestamps using strftime().
7991 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7992 complaints. I've now initialized all the variables, to avoid this.
7994 32. Header lines added by a system filter were not being "seen" during
7995 transport-time rewrites.
7997 33. The info_callback() function passed to OpenSSL is set up with type void
7998 (*)(SSL *, int, int), as described somewhere. However, when calling the
7999 function (actually a macro) that sets it up, the type void(*)() is
8000 expected. I've put in a cast to prevent warnings from picky compilers.
8002 34. If a DNS black list lookup found a CNAME record, but there were no A
8003 records associated with the domain it pointed at, Exim crashed.
8005 35. If a DNS black list lookup returned more than one A record, Exim ignored
8006 all but the first. It now scans all returned addresses if a particular IP
8007 value is being sought. In this situation, the contents of the
8008 $dnslist_value variable are a list of all the addresses, separated by a
8011 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8012 to accept a host name and all its aliases if the forward lookup for any of
8013 them yielded the IP address of the incoming connection. Now it accepts only
8014 those names whose forward lookup yields the correct IP address. Any other
8015 names are discarded. This closes a loophole whereby a rogue DNS
8016 administrator could create reverse DNS records to break through a
8017 wildcarded host restriction in an ACL.
8019 37. If a user filter or a system filter that ran in a subprocess used any of
8020 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8021 the wrong values were passed to the pipe command ($thisaddress had the
8022 value of $0, $0 had the value of $1, etc). This bug was introduced by
8023 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8025 38. Improved the line breaking for long SMTP error messages from ACLs.
8026 Previously, if there was no break point between 40 and 75 characters, Exim
8027 left the rest of the message alone. Two changes have been made: (a) I've
8028 reduced the minimum length to 35 characters; (b) if it can't find a break
8029 point between 35 and 75 characters, it looks ahead and uses the first one
8030 that it finds. This may give the occasional overlong line, but at least the
8031 remaining text gets split now.
8033 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8034 file descriptors might be low, and that setting 1000 would always raise it.
8035 It turns out that in some environments, the limit is already over 1000 and
8036 that lowering it causes trouble. So now Exim takes care not to decrease it.
8038 40. When delivering a message, the value of $return_path is set to $sender_
8039 address at the start of routing (routers may change the value). By an
8040 oversight, this default was not being set up when an address was tested by
8041 -bt or -bv, which affected the outcome if any router or filter referred to
8044 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8045 log only when "message" or "log_message" are set. However, if one of the
8046 conditions was an address verification, or a call to a nested ACL, the
8047 messages generated by the underlying test were being passed through. This
8048 no longer happens. The underlying message is available in $acl_verify_
8049 message for both "message" and "log_message" expansions, so it can be
8050 passed through if needed.
8052 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8053 new expansion $bh_ to give the encoded byte string without charset
8054 translation. Translation happens only if iconv() is available; HAVE_ICONV
8055 indicates this at build time. HEADERS_CHARSET gives the charset to
8056 translate to; headers_charset can change it in the configuration, and
8057 "headers charset" can change it in an individual filter file.
8059 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8060 that creates RFC 2047 encoded "words" labels them as that charset instead
8061 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8062 expansion operator; (ii) when Exim creates a From: line for a local
8063 message; (iii) when a header line is rewritten to include a "phrase" part.
8065 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8066 buggy, causing it to skip the first lines of messages whose message ID
8067 ended in 'D'. This would not have bitten before Exim release 4.14, because
8068 message IDs were unlikely to end in 'D' before then. The effect was to have
8069 incorrect size information for certain domains.
8071 45. #include "config.h" was missing at the start of the crypt16.c module. This
8072 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8075 46. If there was a timeout during a "random" callout check, Exim treated it as
8076 a failure of the random address, and carried on sending RSET and the real
8077 address. If the delay was just some slowness somewhere, the response to the
8078 original RCPT would be taken as a response to RSET and so on, causing
8079 mayhem of various kinds.
8081 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8082 when I implemented it. It didn't allow for the fact that some option values
8083 may legitimately be negative (e.g. size_addition), and it didn't even do
8084 the right test for positive values.
8086 48. Domain names in DNS records are case-independent. Exim always looks them up
8087 in lower case. Some resolvers return domain names in exactly the case they
8088 appear in the zone file, that is, they may contain uppercase letters. Not
8089 all resolvers do this - some return always lower case. Exim was treating a
8090 change of case by a resolver as a change of domain, similar to a widening
8091 of a domain abbreviation. This triggered its re-routing code and so it was
8092 trying to route what was effectively the same domain again. This normally
8093 caused routing to fail (because the router wouldn't handle the domain
8094 twice). Now Exim checks for this case specially, and just changes the
8095 casing of the domain that it ultimately uses when it transmits the message
8098 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8101 50. If a filter generated a file delivery with a non-absolute name (possible if
8102 no home directory exists for the router), the forbid_file option was not
8105 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8106 the existing equality matching.
8108 52. Exim was using ints instead of ino_t variables in some places where it was
8109 dealing with inode numbers.
8111 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8112 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8113 called TMPDIR, and if it finds it is different, it changes its value.
8115 54. The smtp_printf() function is now made available to local_scan() so
8116 additional output lines can be written before returning. There is also an
8117 smtp_fflush() function to enable the detection of a dropped connection.
8118 The variables smtp_input and smtp_batched_input are exported to
8121 55. Changed the default runtime configuration: the message "Unknown user"
8122 has been removed from the ACL, and instead placed on the localuser router,
8123 using the cannot_route_message feature. This means that any verification
8124 failures that generate their own messages won't get overridden. Similarly,
8125 the "Unrouteable address" message that was in the ACL for unverifiable
8126 relay addresses has also been removed.
8128 56. Added hosts_avoid_esmtp to the smtp transport.
8130 57. The exicyclog script was not checking for the esoteric option
8131 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8132 will work only if exicyclog is run under the appropriate euid.
8134 58. Following a discussion on the list, the rules by which Exim recognises line
8135 endings on incoming messages have been changed. The -dropcr and drop_cr
8136 options are now no-ops, retained only for backwards compatibility. The
8137 following line terminators are recognized: LF CRLF CR. However, special
8138 processing applies to CR:
8140 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8141 nor a local message in the state where . is a terminator.
8143 (ii) If a bare CR is encountered in a header line, an extra space is added
8144 after the line terminator so as not to end the header. The reasoning
8145 behind this is that bare CRs in header lines are most likely either
8146 to be mistakes, or people trying to play silly games.
8148 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8149 was being incorrectly given as 18 bytes larger than it should have been.
8150 This is a VOB (very old bug).
8152 60. This may never have affected anything current, but just in case it has:
8153 When the local host is found other than at the start of a list of hosts,
8154 the local host, those with the same MX, and any that follow, are discarded.
8155 When the list in question was part of a longer list of hosts, the following
8156 hosts (not currently being processed) were also being discarded. This no
8157 longer happens. I'm not sure if this situation could ever has previously
8160 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8163 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8164 header. This is something that is recommended in a new Internet Draft, and
8165 is something that is documented as being done by Sendmail. There are two
8166 possible values. For messages generated by the autoreply transport, Exim
8169 Auto-Submitted: auto-replied
8171 whereas for all other generated messages (e.g. bounces) it adds
8173 Auto-Submitted: auto-generated
8175 63. The "personal" condition in filters now includes a test for the
8176 Auto-Submitted: header. If it contains the string "auto-" the message it
8177 not considered personal.
8179 64. Added rcpt_include_affixes as a generic transport option.
8181 65. Added queue_only_override (default true).
8183 66. Added the syslog_duplication option.
8185 67. If what should have been the first header line of a message consisted of
8186 a space followed by a colon, Exim was mis-interpreting it as a header line.
8187 It isn't of course - it is syntactically invalid and should therefore be
8188 treated as the start of the message body. The misbehaviour could have
8189 caused a number of strange effects, including loss of data in subsequent
8190 header lines, and spool format errors.
8192 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8193 client host had authenticated. This control can now be exercised by an ACL
8194 for more flexibility.
8196 69. By default, callouts do not happen when testing with -bh. There is now a
8197 variant, -bhc, which does actually run the callout code, including
8198 consulting and updating the callout cache.
8200 70. Added support for saslauthd authentication, courtesy of Alexander
8203 71. If statvfs() failed on the spool or log directories while checking their
8204 size for availability, Exim confusingly gave the error "space shortage".
8205 Furthermore, in debugging mode it crashed with a floating point exception.
8206 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8207 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8208 serious problem, Exim now writes to the main and panic logs when this
8209 happens, with details of the failure. It then refuses to accept the
8210 incoming message, giving the message "spool directory problem" or "log
8211 directory problem" with a 421 code for SMTP messages.
8213 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8214 0, 1, and 2 exist, because some OS complain for execs without them (see
8215 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8216 descriptors. However, the code omitted to check that the open succeeded,
8217 causing mysterious errors if for some reason the permissions on /dev/null
8218 got screwed. Now Exim writes a message to the main and panic logs, and
8219 bombs out if it can't open /dev/null.
8221 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8222 interact so that it is all more flexible. It is supposed to remain
8223 backwards compatible. Also added extra_local_interfaces.
8225 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8226 to bomb out with an assertion failure - to the client this appears as a
8227 connection drop. This problem occurs in the part of the code that was taken
8228 from the Samba project. Fortunately, the assertion is in a very simple
8229 function, so I have fixed this by reproducing the function inline in the
8230 one place where it is called, and arranging for authentication to fail
8231 instead of killing the process with assert().
8233 75. The SPA client code was not working when the server requested OEM rather
8234 than Unicode encoding.
8236 76. Added code to make require_files with a specific uid setting more usable in
8237 the case where statting the file as root fails - usually a non-root-mounted
8238 NFS file system. When this happens and the failure is EACCES, Exim now
8239 forks a subprocess and does the per-uid checking as the relevant uid.
8241 77. Added process_log_path.
8243 78. If log_file_path was not explicitly set, a setting of check_log_space or
8244 check_log_inodes was ignored.
8246 79. If a space check for the spool or log partitions fails, the incident is now
8247 logged. Of course, in the latter case the data may get lost...
8249 80. Added the %p formatting code to string_format() so that it can be used to
8250 print addresses in debug_print(). Adjusted all the address printing in the
8251 debugging in store.c to use %p rather than %d.
8253 81. There was a concern that a line of code in smtp_in.c could overflow a
8254 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8255 initially expressed, the concern was not well-founded, because trailing
8256 spaces are removed early. However, if the trailing spaces were followed by
8257 a NULL, they did not get removed, so the overflow was possible. Two fixes
8260 (a) I re-wrote the offending code in a cleaner fashion.
8261 (b) If an incoming SMTP command contains a NULL character, it is rejected
8264 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8265 runs initgroups(), so that if the Exim user is in any additional groups,
8266 they will be used during message reception.
8272 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob