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, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
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.
143 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
144 related applications. Previously an "H" was used where available info
145 says that "M" should be, so change to match.
147 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
148 as arguments, so an implementation trying to copy these into a local
149 buffer was taking a taint-enforcement trap. Fix by using dynamically
150 created buffers. Similar fix for radius expansion condition.
152 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
153 reasonable, eg. to count headers. Fix by using dynamically created
154 buffers rather than a local. Do similar fixes for ACL actions "dcc",
155 "log_reject_target", "malware" and "spam"; the arguments are expanded
156 so could be handling tainted values.
158 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
159 broken the (no-op) support for this sendmail command. Restore it
160 to doing nothing, silently, and returning good status.
162 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
163 record path was given (or the default used) without a leading directory
164 path, an error occurred on trying to open it. Use the transport's working
167 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
168 transport. Previously it was the name on the DNS A-record; use instead
169 the head of the CNAME chain leading there (if there is one). This seems
170 to align better with RFC 6125.
172 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
173 smtp_accept_max_per_host allocated resources which were not released
174 when the limit was exceeded. This eventually crashed the daemon. Fix
175 by adding a release action in that path.
177 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
178 expanded; previously using tainted values was rejected. Fix by using
179 dynamically-created buffers.
181 JH/09 Relax restrictions on ACL verify condition needing access to message
182 headers. Previously they were only permitted in data and non-smtp ACLs;
183 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
184 not_blind, header_sender and header_names_ascii verification.
186 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
187 Previously a macro used one argument twice; when called with the
188 argument as an expression having side-effects, incorrect operation
189 resulted. Use an inlineable function.
191 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
192 held open for a verify callout. Previously this wan not accounted for
193 and a corrupt onward SMTP conversation resulted.
195 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
196 Previously they were embedded naively in the constructed address; when
197 needed, strip the quoting and quote the entire local_part.
198 Also make the inbound_srs expansion condition handle quoting.
200 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
201 excluded, not matching the documentation.
203 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
204 was given for the sqlite_dbfile a trap resulted.
206 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
207 "name" argument resulted in a trap. There is no reason to disallow such;
208 this was a coding error.
210 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
211 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
212 spent suspended, ignoring the POSIX definition. Previously we assumed
213 it did and a constant offset from real time could be used as a correction.
214 Change to using the same clock source for the start-of-message and the
215 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
216 exists, just to get a clock slightly more aligned to reality.
218 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
219 RFC says it is optional some validators care. The missing char was not
220 intended but triggered by a line-wrap alignment. Discovery and fix by
221 Guillaume Outters, hacked on by JH.
223 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
224 name being quoted was tainted a trap would be taken. Fix by using
225 dynamically created buffers. The routine could have been called by a
226 rewrite with the "h" flag, by using the "-F" command-line option, or
227 by using a "name=" option on a control=submission ACL modifier.
229 JH/19 SPF: change the Authentication-Results expansion component to give
230 smtp.helo when the sender domain is empty. Previously it gave
233 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
234 not in 127.0.0.0/8 to help in spotting list domains taken over by a
235 domain-parking registrar.
237 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
238 Previously when a whitespace character was specified it was not inserted
239 after removing the newline.
241 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
242 the domain part of the recipient address. This overrides any tls_sni
243 option set, which was previously used.
245 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
248 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
249 is_tainted() had an off-by-one error in the overenthusiastic direction.
250 Find and fix by Gavan. Although NetBSD is not a supported platform for
251 4.94 this bug could affect other platforms.
253 PP/01 Fix default prime selection to be consistent.
254 One path used ike23 still, instead of exim.dev.20160529.3; now both
255 execution flows will use the same DH primes (currently
256 exim.dev.20160529.3).
258 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
259 option in smtp transport, to match the documentation. Previously
260 verification was not being done in some cases where it should have been.
262 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
263 than one server was defined and depending on the platform memory layout
264 details, an internal consistency trap could be hit while walking the list
267 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
268 files. The value is used by the authresults expansion item. Previously
269 if this was used in a router or transport, a crash could result.
271 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
272 have a sufficient incidence of EINTR returns from select that an
273 interaction with other operations done by the main daemon loop exposed
274 a bug in the error-handling. This was benign apart from the log
277 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
278 for consistency with delivered "=>" and failed "**" lines. While we're
279 there, handle PRX and TFO.
281 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
282 applied. This resulted, if any header-line rewrite rules were configured,
283 in a panic-log triggerable by sending a message with a long address in
284 a header. Fix by increasing the arbitrary limit to larger than a single
285 (dewrapped) 5322 header line maximum size.
287 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
288 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
289 RFC 6648 which deprecates X- options in protocols as a general practice.
290 Changeover between the implementations is handled by the mechanisms
293 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
294 a remote which uses callout/hold. Previously the local was queued.
296 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
299 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
300 log line. The asterisk is now consistently appended to the remote IP
301 (and port, if given), and will also be provided on defer and fail log
302 lines. Previously it could be placed on the local IP if that was being
303 logged, and was only provided on delivery lines.
305 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
307 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
308 authenticator client_send option. Previously the next char, after a pair
309 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
310 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
311 documentation. There is still no way to get a leading ^ immediately
312 after a NUL (ie. for the password of a PLAIN method authenticator.
314 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
315 files. For bad sizes read, delete the record and whine to paniclog.
317 JH/38 When logging an AUTH failure, as server, do not include sensitive
318 information. Previously, the credentials would be included if given
319 as part of the AUTH command line and an ACL denied authentication.
321 JH/39 Bug 2691: fix $local_part_data. When the matching list element
322 referred to a file, bad data was returned. This likely also affected
325 JH/40 The gsasl authenticator now supports caching of the salted password
326 generated by the client-side implementation. This required the addition
327 of a new variable: $auth4.
329 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
330 left undeleted; the attempt to re-create it then failed - resulting in
331 the usual "SIGHUP tp have daemon reload configuration" to not work.
332 This affected any platform not supporting "abstract" Unix-domain
333 sockets (i.e. not Linux).
335 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
336 recipients" response to RCPT in a later response, with a 250. The
337 previous coding assumed this would not happen, and under PIPELINING
338 would result in both lost and duplicate recipients for a message.
340 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
341 Previously the weighting was incorrectly applied. Similar fix for socks
342 proxies. Found and fixed by Heiko Schlichting.
344 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
345 not handle sub-lists included using the +namedlist syntax. While
346 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
347 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
350 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
351 Previously the permanent pool was used, so the sore could not be freed.
352 This meant a connection with many messages would use continually-growing
355 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
356 per-pool so as not to waste too much space. Previously a constant size
357 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
358 DOS attacks harder. The cost is wasted memory use in the larger blocks.
360 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
361 same space repeatedly, and a smaller process footprint.
363 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
364 for GnuTLS. Previously the name started "%s" which, while not a bug,
365 looked as if if might be one.
367 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
368 "continued connection") make the $proxy_* variables available. Previously
369 the information was not passed across the exec() call for subsequent
370 transport executions. This also mean that the log lines for the
371 messages can show the proxy information.
373 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
374 receive time. With modern systems the difference is significant.
375 The historical behaviour can be restored by disabling (a new) log_selector
376 "queue_time_exclusive".
378 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
379 content) were specifically tested for. Now, also cover expansions
380 resulting in ACL names and inline ACL content.
382 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
383 making it unusable in complex expressions.
385 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
386 fail with a taint-check on the spoolfile name, and leave the message
389 HS/01 Enforce absolute PID file path name.
391 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
393 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
395 PP/02 Bug 2643: Correct TLS DH constants.
396 A missing NUL termination in our code-generation tool had led to some
397 incorrect Diffie-Hellman constants in the Exim source.
398 Reported by kylon94, code-gen tool fix by Simon Arlott.
400 PP/03 Impose security length checks on various command-line options.
401 Fixes CVE-2020-SPRSS reported by Qualys.
403 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
404 better. Reported by Qualys.
406 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
407 providing a particularly obnoxious sender full name.
410 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
412 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
415 PP/08 Change default for recipients_max from unlimited to 50,000.
417 PP/09 Fix security issue with too many recipients on a message (to remove a
418 known security problem if someone does set recipients_max to unlimited,
419 or if local additions add to the recipient list).
420 Fixes CVE-2020-RCPTL reported by Qualys.
422 PP/10 Fix security issue in SMTP verb option parsing
423 Fixes CVE-2020-EXOPT reported by Qualys.
425 PP/11 Fix security issue in BDAT state confusion.
426 Ensure we reset known-good where we know we need to not be reading BDAT
427 data, as a general case fix, and move the places where we switch to BDAT
428 mode until after various protocol state checks.
429 Fixes CVE-2020-BDATA reported by Qualys.
431 HS/03 Die on "/../" in msglog file names
433 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
434 the Exim runtime user are allowed to create files.
436 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
439 QS/03 When reading the output from interpreted forward files we do not
440 pass the pipe between the parent and the interpreting process to
441 executed child processes (if any).
443 QS/04 Always die if requested from internal logging, even is logging is
446 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
447 the API; compilation noo longer completes with DMARC support included.
448 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
449 been reported on other platforms.
451 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
453 JH/56 Make the majority of info read from config files readonly, for defence-in-
454 depth against exploits. Suggestion by Qualys.
455 Not supported on Solaris 10.
457 JH/57 Fix control=fakreject for a custom message containing tainted data.
458 Previously this resulted in a log complaint, due to a re-expansion present
459 since fakereject was originally introduced.
461 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
462 modifier was given, a loop resulted.
464 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
465 pipelined SMTP command followed the BDAT LAST then it would be
466 incorrectly treated as part of the message body, causing a verification
469 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
470 require looser validation rules than those for 821-level addresses,
471 which only permit IP addresses.
477 JH/01 Avoid costly startup code when not strictly needed. This reduces time
478 for some exim process initialisations. It does mean that the logging
479 of TLS configuration problems is only done for the daemon startup.
481 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
483 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
484 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
487 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
488 DKIM signing is being done. Previously a transport_filter would always
489 disable CHUNKING, falling back to traditional DATA.
491 JH/05 Regard command-line recipients as tainted.
493 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
495 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
496 PAM library frees one of the arguments given to it, despite the
497 documentation. Therefore a plain malloc must be used.
499 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
500 on-stack buffers were used, resulting in a taint trap when DSN information
501 copied from a received message was written into the buffer.
503 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
504 the ordering of its ARC headers. This caused a crash.
506 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
507 a new record was being constructed with information from the peer, a trap
510 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
511 installation would get error messages from DMARC verify, when it hit the
512 nonexistent file indicated by the default. Distros wanting DMARC enabled
513 should both provide the file and set the option.
514 Also enforce no DMARC verification for command-line sourced messages.
516 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
517 could, depending on the platform, hang at the STARTTLS response.
519 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
520 message on a connection. Previously if one message had ARC headers and
521 the following one did not, a crash could result when adding an
522 Authentication-Results: header.
524 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
525 Exim main code and Exim-related utities. The introduction of taint
526 tracking also did many adjustments to string handling. Since then, eximon
527 frequently terminated with an assert failure.
529 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
530 check for 452 responses. This slightly helps the inefficieny of doing
531 a large alias-expansion into a recipient-limited target. The max_rcpt
532 transport option still applies (and at the current default, will override
533 the new feature). The check is done for either cause of synch, and forces
534 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
535 connection. The new facility is not tunable at this time.
537 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
538 library live data was being used, so the results became garbage. Make
539 copies while it is still usable.
541 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
542 on delivery deferred (==) and failed (**) lines (if a delivery was
543 attemtped). Previously it was only on completion (=>) lines.
545 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
546 for the expansion of the server_scram_iter and server_scram_salt options.
548 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
549 are now specifically given a NO_DATA response without hitting the system
550 resolver. The library goes on to do the now-standard TXT lookup.
551 Use of dnsdb lookups is not affected.
553 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
554 only retrieve the errormessage once. Previously two calls to dlerror()
555 were used, and the second one (for mainlog/paniclog) retrieved null
558 JH/20 Taint checking: disallow use of tainted data for
559 - the appendfile transport file and directory options
560 - the pipe transport command
561 - the autoreply transport file, log and once options
562 - file names used by the redirect router (including filter files)
564 - paths used by single-key lookups
565 Previously this was permitted.
567 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
568 adjusted the size of a major service buffer; this failed because the
569 buffer was in use at the time. Change to a compile-time increase in the
570 buffer size, when this authenticator is compiled into exim.
572 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
573 previous fast-mode was untenable in the face of glibs using mmap to
574 support larger malloc requests.
576 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
577 New values supported, if defined on system where compiled:
578 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
579 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
581 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
582 running a limited number of proceses in parallel, a benefit is gained. The
583 amount varies with the platform hardware and load. The use of the option
584 queue_run_in_order means we cannot do this, as ordering becomes
587 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
588 had introduced a string-copy (for ensuring NUL-termination) which was not
589 appropriate for that case, which can include embedded NUL bytes in the
590 block of data. Investigation showed the copy to actually be needless, the
591 data being length-specified.
593 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
594 done during a receiving connection, and both used TLS, global info was
595 used rather than per-connection info for tracking the state of data
596 queued for transmission. This could result in a connection hang.
598 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
599 Previously, when delivering serveral messages down a single connection
600 only the first would provide a SIZE. This was due to the size information
601 not being properly tracked.
603 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
604 TAI (at 37 seconds currently), pretend to be in UTC for time-related
605 expansion and logging. Previously, spurious values such as a future
606 minute could be seen.
608 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
609 it could crash from a null-deref. This could also affect the
610 ${addresses: } operator and ${readsock } item.
612 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
613 message following a mime one, the variable was not reset.
615 JH/30 When an pipelined-connect fails at the first response, assume incorrect
616 cached capability (perhaps the peer reneged?) and immediately retry in
619 JH/31 Fix spurious detection of timeout while writing to transport filter.
621 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
622 an attempt to copy the string was made before checking it.
624 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
625 taint of the lookup key was maintained; we now regard the presence in the
626 filesystem as sufficient validation.
628 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
629 argument is supplied.
631 JH/35 The dsearch lookup now requires that the directory is an absolute path.
632 Previously this was not checked, and nonempty relative paths made an
633 access under Exim's current working directory.
635 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
636 Previously no event was raised.
638 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
639 parameter supplied by the sender MAIL FROM command. Previously it was
640 ignored, and only the check_spool_space option value for the required
643 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
644 the size of the signing public-key. Previously it was instead giving
645 the size of the signature hash.
647 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
648 the default. See the (new) dkim_verify_min_keysizes option.
650 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
651 and an ACL use a lookup for checking either the local_part or domain,
652 stale data could be accessed. Ensure that variable references are
653 dropped between messages.
655 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
656 by the client was not checked as pointing within response data before
657 being used. A malicious client could thus cause an out-of-bounds read and
658 possibly gain authentication. Fix by adding the check.
660 JH/42 Internationalisation: change the default for downconversion in the smtp
661 transport to be "if needed". Previously it was "as previously set" for
662 the message, which usually meant "if needed" for message-submission but
663 "no" for everything else. However, MTAs have been seen using SMTPUTF8
664 even when the envelope addresses did not need it, resulting in forwarding
665 failures to non-supporting MTAs. A downconvert in such cases will be
666 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
667 transport. The change does mean that addresses needing conversion will
668 be converted when previously a delivery failure would occur.
670 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
671 response was received it would be used unchecked in a fail-DSN, violating
672 standards on line-length limits. Truncate if needed.
674 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
675 doesn't work. (Additionally add a new main config option to configure the
682 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
683 side, to decode a TLS 1.3 packet capture.
685 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
686 Previously the default library behaviour applied, sending two, each in
689 JH/03 Debug output for ACL now gives the config file name and line number for
692 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
694 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
696 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
697 buffer overrun for (non-chunking) other transports.
699 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
700 TLS1.3, means that a server rejecting a client certificate is not visible
701 to the client until the first read of encrypted data (typically the
702 response to EHLO). Add detection for that case and treat it as a failed
703 TLS connection attempt, so that the normal retry-in-clear can work (if
704 suitably configured).
706 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
707 and/or domain. Found and fixed by Jason Betts.
709 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
710 configuration). If a CNAME target was not a wellformed name pattern, a
713 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
714 the OS reports them interleaved with other addresses.
716 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
717 used both for input and for a verify callout, both encrypted, SMTP
718 responses being sent by the server could be lost. This resulted in
719 dropped connections and sometimes bounces generated by a peer sending
722 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
723 string. Previously it was possible to cause undefined behaviour in a
724 library routine (usually a crash). Found by "zerons".
726 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
729 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
730 API was removed, so update to use the newer ones.
732 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
733 any timeout set, is taking a long time. Previously we would hang on to a
734 rotated logfile "forever" if the input was arriving with long gaps
735 (a previous attempt to fix addressed lack, for a long time, of initial
738 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
739 shared (NFS) environment. The length of the tempfile name is now
740 4 + 16 ("hdr.$message_exim_id") which might break on file
741 systems which restrict the file name length to lower values.
744 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
745 shared (NFS) environment.
747 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
748 did for all versions <4.90). Notably -M, -m, --invert, -I may be
751 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
752 on some platforms for bit 31.
754 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
755 to changes apparently associated with TLS1.3 handling some of the APIs
756 previously used were either nonfunctional or inappropriate. Strings
757 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
758 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
759 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
760 This affects log line X= elements, the $tls_{in,out}_cipher variables,
761 and the use of specific cipher names in the encrypted= ACL condition.
763 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
765 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
766 verification result was not updated unless hosts_require_ocsp applied.
768 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
769 queue_list_requires_admin set to false, non-admin users were denied the
772 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
773 directory-of-certs mode. Previously they were advertised despite the
776 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
777 A single TCP connection by a client will now hold a TLS connection open
778 for multiple message deliveries, by default. Previously the default was to
781 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
782 default. If built with the facility, DANE will be used. The facility
783 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
785 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
786 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
787 must be defined and you must still, unless you define DISABLE_TLS, manage
788 the the include-dir and library-file requirements that go with that
789 choice. Non-TLS builds are still supported.
791 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
794 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
795 default. If the platform supports and has the facility enabled, it will
796 be requested on all coneections.
798 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
799 controlled by the build-time option SUPPORT_PIPE_CONNECT.
801 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
803 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
804 success-DSN messages. Previously the From: header was always the default
805 one for these; the option was ignored.
807 JH/28 Fix the timeout on smtp response to apply to the whole response.
808 Previously it was reset for every read, so a teergrubing peer sending
809 single bytes within the time limit could extend the connection for a
810 long time. Credit to Qualsys Security Advisory Team for the discovery.
812 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
813 delivery address, which leaked information of the results of local
814 forwarding. Change to the original envelope recipient address, per
817 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
818 requested. Previously not bounce was generated and a log entry of
819 error ignored was made.
821 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
823 JH/32 Introduce a general tainting mechanism for values read from the input
824 channel, and values derived from them. Refuse to expand any tainted
825 values, to catch one form of exploit.
827 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
828 was unused and the unexpanded text used for the test. Found and
829 fixed by Ruben Jenster.
831 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
832 an attempt to use a TLS library read routine dereffed a nul pointer,
835 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
836 them in our smtp response.
838 JH/36 Have the general router option retry_use_local_part default to true when
839 any of the restrictive preconditions are set (to anything). Previously it
840 was only for check_local user. The change removes one item of manual
841 configuration which is required for proper retries when a remote router
842 handles a subset of addresses for a domain.
844 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
845 link count into consideration.
847 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
848 caused the extension of big_buffer, the following lines were ignored.
850 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
851 accordance with RFC 2308. Previously there was no expiry, so a longlived
852 receive process (eg. due to ACL delays) versus a short SOA value could
855 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
857 JH/39 Promote DMARC support to mainline.
859 JH/40 Bug 2452: Add a References: header to DSNs.
861 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
862 parameters. The relevant library call is documented as "Deprecated: This
863 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
864 3.6.0, DH parameters are negotiated following RFC7919."
866 HS/06 Change the default of dnssec_request_domains to "*"
868 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
869 carried on and emitted a BDAT command, even when PIPELINING was not
872 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
873 buffer was used for the filename, resulting in a trap when tainted
874 arguments (eg. $domain) were used.
876 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
877 recommended to avoid a possible server-load attack. The feature can be
878 re-enabled via the openssl_options main cofiguration option.
880 JH/45 local_scan API: documented the current smtp_printf() call. This changed
881 for version 4.90 - adding a "more data" boolean to the arguments.
882 Bumped the ABI version number also, this having been missed previously;
883 release versions 4.90 to 4.92.3 inclusive were effectively broken in
884 respect of usage of smtp_printf() by either local_scan code or libraries
885 accessed via the ${dlfunc } expansion item. Both will need coding
886 adjustment for any calls to smtp_printf() to match the new function
887 signature; a FALSE value for the new argument is always safe.
889 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
890 the file-offset (which the Linux syscall does, and exim expects); this
891 resulted in an indefinite loop.
893 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
894 to do ARC verification. The Authentication-Results: header line added
895 by the configuration then had no ARC item.
897 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
898 an incorrect mode was used for reading the result, resulting in it being
905 JH/01 Remove code calling the customisable local_scan function, unless a new
906 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
908 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
909 non-signal-safe functions being used.
911 JH/03 Bug 2269: When presented with a received message having a stupidly large
912 number of DKIM-Signature headers, disable DKIM verification to avoid
913 a resource-consumption attack. The limit is set at twenty.
915 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
916 report of oldest_pass in ${authres } in consequence, and separate out
917 some descriptions of reasons for verification fail.
919 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
920 files in the spool were present and unlocked. A queue-runner could spot
921 them, resulting in a duplicate delivery. Fix that by doing the unlock
922 after the unlink. Investigation by Tim Stewart. Take the opportunity to
923 add more error-checking on spoolfile handling while that code is being
926 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
927 No known attacks, no CVE, this is defensive hardening.
929 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
930 a queue-runner could start a delivery while other operations were ongoing.
931 Cutthrough delivery was a common victim, resulting in duplicate delivery.
932 Found and investigated by Tim Stewart. Fix by using the open message data
933 file handle rather than opening another, and not locally closing it (which
934 releases a lock) for that case, while creating the temporary .eml format
935 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
937 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
938 $sender_verify_failure/$recipient_verify_failure to "random".
940 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
943 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
944 Previously this would segfault.
946 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
949 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
950 like zero, since the resolver should be doing this for us, But we need one
951 as a CNAME but no MX presence gets the CNAME returned; we need to check
952 that doesn't point to an MX to declare it "no MX returned" rather than
953 "error, loop". A new main option is added so the older capability of
954 following some limited number of chain links is maintained.
956 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
958 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
959 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
960 now-deprecated earlier definitions used only specified the range up to TLS
961 1.2 (in the older-version library docs).
963 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
965 JH/15 Rework TLS client-side context management. Stop using a global, and
966 explicitly pass a context around. This enables future use of TLS for
967 connections to service-daemons (eg. malware scanning) while a client smtp
968 connection is using TLS; with cutthrough connections this is quite likely.
970 JH/16 Fix ARC verification to do AS checks in reverse order.
972 JH/17 Support a "tls" option on the ${readsocket } expansion item.
974 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
975 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
976 Previously the "utf8" would be re-prepended for every additional message.
978 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
979 Previously thery were accepted, resulting in issues when attempting to
980 forward messages to a non-supporting MTA.
982 PP/02 Let -n work with printing macros too, not just options.
984 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
985 one parent address was copied, and bogus data was used at delivery-logging
986 time. Either a crash (after delivery) or bogus log data could result.
987 Discovery and analysis by Tim Stewart.
989 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
990 Previously if the string ended mid-character, we did not insert the
991 promised '?' replacement.
993 PP/04 Documentation: current string operators work on bytes, not codepoints.
995 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
996 should pack well giving a smaller memory footprint so better caching and
997 therefore performance. Group the declarations where this can't be done so
998 that the byte-sized flag variables are not interspersed among pointer
999 variables, giving a better chance of good packing by the compiler.
1001 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1002 non-null, to avoid issues with sites running BATV. Previously reports were
1003 sent with an empty envelope sender so looked like bounces.
1005 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1006 The ignore_error flag wasn't being returned from the filter subprocess so
1007 was not set for later routers. Investigation and fix by Matthias Kurz.
1009 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1010 and a msg:complete for the whole, when a message is manually removed using
1011 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1013 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1014 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1015 hope that is portable enough.
1017 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1018 requiring a known-CA anchor certificate; make it now rely entirely on the
1019 TLSA as an anchor. Checking the name on the leaf cert against the name
1020 on the A-record for the host is still done for TA (but not for EE mode).
1022 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1023 included in delivery lines for non-proxied connections, when compiled with
1024 SUPPORT_SOCKS and running with proxy logging enabled.
1026 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1027 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1028 move the existing event to fire before the normal logging of message
1029 failure so that custom logging is bracketed by normal logging.
1031 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1032 msg:fail:internal event. Developement by Matthias Kurz.
1034 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1035 far too small for todays use of crypto signatures stored there. Go all
1036 the way to the max DNS message size of 64kB, even though this might be
1037 overmuch for IOT constrained device use.
1039 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1040 copy a string over itself. The library routine is documented as not
1041 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1043 JH/32 For main options check_spool_space and check_inode_space, where the
1044 platform supports 64b integers, support more than the previous 2^31 kB
1045 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1046 the previous G, M, k.
1048 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1049 $authenticated_fail_id variable on authentication failure. Previously
1052 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1053 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1054 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1055 GNUTLS_SEC_PARAM_MEDIUM.
1057 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1058 side. Previously we would continue as if no SNI had been received.
1060 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1061 "<" character, which should be a list-separator specification, we walked
1062 off past the nul-terimation.
1064 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1065 causes) even when the retry time is not yet met. Previously they were
1066 not, meaning that when (say) an account was over-quota and temp-rejecting,
1067 and multiple senders' messages were queued, only one sender would get
1068 notified on each configured delay_warning cycle.
1070 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1072 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1073 cipher-suites, an error can be left on the stack even for a succeeding
1074 accept; this results in impossible error messages when a later operation
1077 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1078 return error codes indicating retry. Under TLS1.3 this becomes required.
1080 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1081 it only wrote the new authenticators, resulting in a lack of tracking of
1082 peer changes of ESMTP extensions until the next cache flush.
1084 JH/41 Fix the loop reading a message header line to check for integer overflow,
1085 and more-often against header_maxsize. Previously a crafted message could
1086 induce a crash of the recive process; now the message is cleanly rejected.
1088 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1089 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1095 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1096 When redis_servers is set to a list of > 1 element, and the Redis servers
1097 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1098 case of MOVED into a DEFER case instead, thus moving the query onto the
1099 next server in the list. For a cluster of N elements, all N servers must
1100 be defined in redis_servers.
1102 GF/02 Catch and remove uninitialized value warning in exiqsumm
1103 Check for existence of @ARGV before looking at $ARGV[0]
1105 JH/01 Replace the store_release() internal interface with store_newblock(),
1106 which internalises the check required to safely use the old one, plus
1107 the allocate and data copy operations duplicated in both (!) of the
1108 extant use locations.
1110 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1111 modifier. This matches the restriction on the commandline.
1113 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1114 Previously only the last row was returned.
1116 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1117 we assumed that tags in the header were well-formed, and parsed the
1118 element content after inspecting only the first char of the tag.
1119 Assumptions at that stage could crash the receive process on malformed
1122 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1123 While running the DKIM ACL we operate on the Permanent memory pool so that
1124 variables created with "set" persist to the DATA ACL. Also (at any time)
1125 DNS lookups that fail create cache records using the Permanent pool. But
1126 expansions release any allocations made on the current pool - so a dnsdb
1127 lookup expansion done in the DKIM ACL releases the memory used for the
1128 DNS negative-cache, and bad things result. Solution is to switch to the
1129 Main pool for expansions.
1130 While we're in that code, add checks on the DNS cache during store_reset,
1131 active in the testsuite.
1132 Problem spotted, and debugging aided, by Wolfgang Breyha.
1134 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1135 When none of the hosts presented to a transport match an already-open
1136 connection, close it and proceed with the list. Previously we would
1137 queue the message. Spotted by Lena with Yahoo, probably involving
1140 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1141 Previously a spurious "250 OK id=" response was appended to the proper
1144 JH/08 The "support for" informational output now, which built with Content
1145 Scanning support, has a line for the malware scanner interfaces compiled
1146 in. Interface can be individually included or not at build time.
1148 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1149 by the template makefile "src/EDITME". The "STREAM" support for an older
1150 ClamAV interface method is removed.
1152 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1153 rows affected is given instead).
1155 JH/11 The runtime Berkeley DB library version is now additionally output by
1156 "exim -d -bV". Previously only the compile-time version was shown.
1158 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1159 SMTP connection. Previously, when one had more recipients than the
1160 first, an abortive onward connection was made. Move to full support for
1161 multiple onward connections in sequence, handling cutthrough connection
1162 for all multi-message initiating connections.
1164 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1165 routers. Previously, a multi-recipient message would fail to match the
1166 onward-connection opened for the first recipient, and cause its closure.
1168 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1169 a timeout on read on a GnuTLS initiating connection, resulting in the
1170 initiating connection being dropped. This mattered most when the callout
1171 was marked defer_ok. Fix to keep the two timeout-detection methods
1174 JH/15 Relax results from ACL control request to enable cutthrough, in
1175 unsupported situations, from error to silently (except under debug)
1176 ignoring. This covers use with PRDR, frozen messages, queue-only and
1179 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1181 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1182 metadata, resulting in a crash in free().
1184 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1185 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1186 Broken also in d185889f4, with init system revamp.
1188 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1189 Previously we abruptly closed the connection after reading a malware-
1190 found indication; now we go on to read the "scan ok" response line,
1193 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1194 ACL. Previously, a crash would result.
1196 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1197 macro text after a replacement (previously it was only once per line) and
1198 by skipping builtin macros when searching for an uppercase lead character.
1200 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1201 for the build is renamed.
1203 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1204 was allocated for every new TLS startup, meaning one per message. Fix
1205 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1207 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1208 reported the original. Fix to report (as far as possible) the ACL
1209 result replacing the original.
1211 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1212 OpenSSL. Certificate information is loaded for every new TLS startup,
1213 and the resources needed to be freed.
1215 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1217 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1220 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1221 DATA response info to the (existing) per-recipient response info for
1222 the "C=" log element. It can have useful tracking info from the
1223 destination system. Patch from Simon Arlott.
1225 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1226 length value. Previously this would segfault.
1228 HS/02 Support Avast multiline protoocol, this allows passing flags to
1229 newer versions of the scanner.
1231 JH/28 Ensure that variables possibly set during message acceptance are marked
1232 dead before release of memory in the daemon loop. This stops complaints
1233 about them when the debug_store option is enabled. Discovered specifically
1234 for sender_rate_period, but applies to a whole set of variables.
1235 Do the same for the queue-runner and queue-list loops, for variables set
1236 from spool message files. Do the same for the SMTP per-message loop, for
1237 certain variables indirectly set in ACL operations.
1239 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1240 as a multi-recipient message from a mailinglist manager). The coding had
1241 an arbitrary cutoff number of characters while checking for more input;
1242 enforced by writing a NUL into the buffer. This corrupted long / fast
1243 input. The problem was exposed more widely when more pipelineing of SMTP
1244 responses was introduced, and one Exim system was feeding another.
1245 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1246 receiving system, and refused recipients seen by the sending system
1247 (propating to people being dropped from mailing lists).
1248 Discovered and pinpointed by David Carter.
1250 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1251 replaced by the ${authresults } expansion.
1253 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1255 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1256 allows proper process termination in container environments.
1258 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1259 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1261 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1262 and "err_perm", deprecated since 4.83 when the RFC-defined words
1263 "temperror" and "permerror" were introduced.
1265 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1266 transport-filters or DKIM-signing. The restriction was lost in the
1267 consolidation of verify-callout and delivery SMTP handling.
1268 Extend the restriction to also cover ARC-signing.
1270 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1271 in defer=pass mode supply a 450 to the initiator. Previously the message
1274 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1275 tls_require_ciphers is used as before.
1277 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1278 "pass_unscanned". Only tmpfails from the scanner are written to
1279 the paniclog, as they may require admin intervention (permission
1280 denied, license issues). Other scanner errors (like decompression
1281 bombs) do not cause a paniclog entry.
1283 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1284 Previously it was possible to log spurious information in receive log
1287 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1288 triggered odd behaviour from Outlook Express clients.
1290 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1293 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1294 since the IETF WG has not yet settled on that versus the original
1295 "bare" representation.
1297 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1298 Previously the millisecond value corrupted the output.
1299 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1300 corrupted the output.
1306 JH/01 Rework error string handling in TLS interface so that the caller in
1307 more cases is responsible for logging. This permits library-sourced
1308 string to be attached to addresses during delivery, and collapses
1309 pairs of long lines into single ones.
1311 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1312 during configuration. Wildcards are allowed and expanded.
1314 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1315 This permits better logging.
1317 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1318 do not close it down and have the child transport start it up again on
1319 the passed-on TCP connection. Instead, proxy the child (and any
1320 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1321 affected: the continued delivery log lines do not have any DNSSEC, TLS
1322 Certificate or OCSP information. TLS cipher information is still logged.
1324 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1325 identical IP addresses on different listening ports. Will also affect
1328 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1329 add noisy ifdef guards to special-case this sillyness.
1330 Patch from Bernd Kuhls.
1332 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1333 than 255 are no longer allowed.
1335 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1336 Disable the session-cache too, which might reduce our load. Since we
1337 currrectly use a new context for every connection, both as server and
1338 client, there is no benefit for these.
1339 GnuTLS appears to not support tickets server-side by default (we don't
1340 call gnutls_session_ticket_enable_server()) but client side is enabled
1341 by default on recent versions (3.1.3 +) unless the PFS priority string
1344 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1345 <https://reproducible-builds.org/specs/source-date-epoch/>.
1347 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1348 the check for any unsuccessful recipients did not notice the limit, and
1349 erroneously found still-pending ones.
1351 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1352 MSG_MORE. Only in-clear (not on TLS connections).
1354 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1355 transport-filter is involved we can buffer the headers in memory for
1356 creating the signature, and read the spool data file once for the
1357 signature and again for transmission.
1359 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1360 4.77 as the kernel support then wasn't solid, having issues in 64bit
1361 mode. Now, it's been long enough. Add support for FreeBSD also.
1363 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1364 case where the routing stage had gathered several addresses to send to
1365 a host before calling the transport for the first, we previously failed
1366 to close down TLS in the old transport process before passing the TCP
1367 connection to the new process. The new one sent a STARTTLS command
1368 which naturally failed, giving a failed delivery and bloating the retry
1369 database. Investigation and fix prototype from Wolfgang Breyha.
1371 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1372 false-negatives in the check that the sender had not preempted a response
1373 or prompt from Exim (running as a server), due to that code's lack of
1374 awareness of the SMTP input buffering.
1376 PP/04 Add commandline_checks_require_admin option.
1377 Exim drops privileges sanely, various checks such as -be aren't a
1378 security problem, as long as you trust local users with access to their
1379 own account. When invoked by services which pass untrusted data to
1380 Exim, this might be an issue. Set this option in main configuration
1381 AND make fixes to the calling application, such as using `--` to stop
1384 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1385 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1386 responses to those, into a single TLS record each way (this usually means
1387 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1390 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1391 affects you only if you're dancing at the edge of the param size limits.
1392 If you are, and this message makes sense to you, then: raise the
1393 configured limit or use OpenSSL 1.1. Nothing we can do for older
1396 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1397 cmdline element to get the documented default one. Previously it was
1400 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1401 are unsuable for non-message-specific reasons - eg. connection timeout,
1402 banner-time rejection.
1404 JH/16 Fix logging of delivery remote port, when specified by router, under
1407 PP/06 Repair manualroute's ability to take options in any order, even if one
1408 is the name of a transport.
1411 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1413 JH/17 Change the list-building routines interface to use the expanding-string
1414 triplet model, for better allocation and copying behaviour.
1416 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1417 Previously it was constructed the first time a possibly-matching string
1418 was met in the configuration file input during startup; now it is done
1421 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1422 compatible one, to avoid the (poorly documented) possibility of a config
1423 file in the working directory redirecting the DB files, possibly correpting
1424 some existing file. CVE-2017-10140 assigned for BDB.
1426 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1427 cache-hot. Previously, although the result was properly cached, the
1428 initial verify call returned a defer.
1430 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1431 the main verify for receipient in uncached-mode.
1433 JH/22 Retire historical build files to an "unsupported" subdir. These are
1434 defined as "ones for which we have no current evidence of testing".
1436 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1437 if present. Previously it was ignored.
1439 JH/24 Start using specified-initialisers in C structure init coding. This is
1440 a C99 feature (it's 2017, so now considered safe).
1442 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1443 if was a fixed-sized field and bitmask ops via macros; it is now more
1446 PP/07 GitHub PR 56: Apply MariaDB build fix.
1447 Patch provided by Jaroslav Škarvada.
1449 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1450 during Coverity cleanups [4.87 JH/47]
1451 Diagnosis and fix provided by Michael Fischer v. Mollard.
1453 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1454 the right size to place the terminating semicolon on its own folded
1455 line, the header hash was calculated to an incorrect value thanks to
1456 the (relaxed) space the fold became.
1458 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1459 and confused the parent.
1461 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1462 which could crash as a result. This could lead to undeliverable messages.
1464 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1467 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1468 matters on fast-turnover and PID-randomising systems, which were getting
1469 out-of-order delivery.
1471 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1472 a possibly-overlapping copy. The symptom was that "Remote host closed
1473 connection in response to HELO" was logged instead of the actual 4xx
1476 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1477 Previously only that bufferd was discarded, resulting in SYMTP command
1480 JH/32 DKIM: when a message has multiple signatures matching an identity given
1481 in dkim_verify_signers, run the dkim acl once for each. Previously only
1482 one run was done. Bug 2189.
1484 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1485 "panic the current process" to "deliberately defer". The panic log is
1486 still written with the problem list name; the mail and reject logs now
1487 get a temp-reject line for the message that was being handled, saying
1488 something like "domains check lookup or other defer". The SMTP 451
1489 message is still "Temporary local problem".
1491 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1492 A crafted sequence of BDAT commands could result in in-use memory beeing
1493 freed. CVE-2017-16943.
1495 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1496 from SMTP input. Previously it was always done; now only done for DATA
1497 and not BDAT commands. CVE-2017-16944.
1499 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1500 to the message (such as an overlong header line). Previously this was
1501 not done and we did not exit BDAT mode. Followon from the previous item
1502 though a different problem.
1508 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1509 than -2003 did; needs libidn2 in addition to libidn.
1511 JH/02 The path option on a pipe transport is now expanded before use.
1513 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1514 Patch provided by "Björn", documentation fix added too.
1516 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1517 missing a wire-to-host endian conversion.
1519 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1520 close after a BDAT command line could be taken as a following command,
1521 giving a synch failure. Fix by only checking for synch immediately
1522 before acknowledging the chunk.
1524 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1525 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1526 macro. Patches provided by Josh Soref.
1528 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1529 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1530 by IANA as a service extension. However, John Klensin suggests that we
1533 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1534 the dkim code may be unix-mode line endings rather than smtp wire-format
1535 CRLF, so prepend a CR to any bare LF.
1537 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1538 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1540 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1541 the first were themselves being wrongly included in the feed into dkim
1542 processing; with most chunk sizes in use this resulted in an incorrect
1543 body hash calculated value.
1545 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1546 DKIM signature block, for verification. Although advised against by
1547 standards it is specifically not ruled illegal.
1549 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1551 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1552 missing a body hash (the bh= tag).
1554 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1555 It seems that HAProxy sends the Proxy Protocol information in clear and
1556 only then does a TLS startup, so do the same.
1558 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1559 TCP connections (such as for Spamd) unless the daemon successfully set
1560 Fast Open mode on its listening sockets. This fixes breakage seen on
1561 too-old kernels or those not configured for Fast Open, at the cost of
1562 requiring both directions being enabled for TFO, and TFO never being used
1563 by non-daemon-related Exim processes.
1565 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1566 endings, at least on the first header line. Try to canonify any that get
1567 past that check, despite the cost.
1569 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1570 now limited to an arbitrary five deep, while parsing addresses with the
1571 strip_excess_angle_brackets option enabled.
1573 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1574 instead leave the unprompted TLS handshake in socket buffer for the
1575 TLS library to consume.
1577 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1579 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1581 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1582 region in one operation, for speed, and this leaves assigned pointers
1583 dangling. Add checks run only under the testsuite which checks all
1584 variables at a store-reset and panics on a dangling pointer; add code
1585 explicitly nulling out all the variables discovered. Fixes one known
1586 bug: a transport crash, where a dangling pointer for $sending_ip_address
1587 originally assigned in a verify callout, is re-used.
1589 PP/06 Drop '.' from @INC in various Perl scripts.
1591 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1593 PP/08 Reduce a number of compilation warnings under clang; building with
1594 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1595 should be warning-free.
1597 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1599 HS/01 Fix portability problems introduced by PP/08 for platforms where
1600 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1602 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1603 chunk. This allows us to accept broken chunked messages. We need a more
1604 general solution here.
1606 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1607 already-broken messages in the queue.
1609 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1611 JH/19 Fix reference counting bug in routing-generated-address tracking.
1617 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1618 supports it and a size is available (ie. the sending peer gave us one).
1620 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1621 years of being deprecated). The replacements are the ACLs
1622 acl_smtp_mime and acl_not_smtp_mime.
1624 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1625 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1626 or even an in-clear connection were permitted. Now, if the host lookup
1627 was dnssec and dane was requested then the host is only used if the
1628 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1629 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1630 if one fails this test.
1631 This means that a poorly-configured remote DNS will make it incommunicado;
1632 but it protects against a DNS-interception attack on it.
1634 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1635 non-noisy when a race steals the message being considered.
1637 JH/05 If main configuration option tls_certificate is unset, generate a
1638 self-signed certificate for inbound TLS connections.
1640 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1641 in rewrites and routers.
1643 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1644 and logged a warning sing 4.83; now they are a configuration file error.
1646 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1647 (lacking @domain). Apply the same qualification processing as RCPT.
1649 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1651 JH/10 Support ${sha256:} applied to a string (as well as the previous
1654 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1655 a cutthrough deliver is pending, as we always want to make a connection.
1656 This also avoids re-routing the message when later placing the cutthrough
1657 connection after a verify cache hit.
1658 Do not update it with the verify result either.
1660 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1661 when routing results in more than one destination address.
1663 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1664 signing (which inhibits the cutthrough capability). Previously only
1665 the presence of an option was tested; now an expansion evaluating as
1666 empty is permissible (obviously it should depend only on data available
1667 when the cutthrough connection is made).
1669 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1670 the relevant preceding SMTP command did not note the pipelining mode.
1672 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1673 Previously they were not counted.
1675 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1676 as one having no matching records. Previously we deferred the message
1677 that needed the lookup.
1679 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1680 distinguished as "(=".
1682 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1683 for missing MX records. Previously it only worked for missing A records.
1685 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1687 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1688 after the data-go-ahead and data-ack. Patch from Jason Betts.
1690 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1691 even for a "none" policy. Patch from Tony Meyer.
1693 JH/22 Fix continued use of a connection for further deliveries. If a port was
1694 specified by a router, it must also match for the delivery to be
1697 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1698 When one of the recipients of a message was unsuitable for the connection
1699 (has no matching addresses), we lost track of needing to mark it
1700 deferred. As a result mail would be lost.
1702 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1704 JH/25 Decoding ACL controls is now done using a binary search; the source code
1705 takes up less space and should be simpler to maintain. Merge the ACL
1706 condition decode tables also, with similar effect.
1708 JH/26 Fix problem with one_time used on a redirect router which returned the
1709 parent address unchanged. A retry would see the parent address marked as
1710 delivered, so not attempt the (identical) child. As a result mail would
1713 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1714 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1715 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1718 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1719 Main config options check_{log,spool}_{inodes,space} are now
1720 100 inodes, 10MB unless set otherwise in the configuration.
1722 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1723 Previously it only applied to the main-section connection policy
1726 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1728 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1729 by me. Added RFC7919 DH primes as an alternative.
1731 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1732 are not in the system include path.
1734 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1735 GnuTLS, when a session startup failed (eg because the client disconnected)
1736 Exim did stdio operations after fclose. This was exposed by a recent
1737 change which nulled out the file handle after the fclose.
1739 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1740 signed directly by the cert-signing cert, rather than an intermediate
1741 OCSP-signing cert. This is the model used by LetsEncrypt.
1743 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1745 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1746 an incoming connection.
1748 HS/02 Bug 1802: Do not half-close the connection after sending a request
1751 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1752 fallback to "prime256v1".
1754 JH/34 SECURITY: Use proper copy of DATA command in error message.
1755 Could leak key material. Remotely exploitable. CVE-2016-9963.
1761 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1762 and 3.4.4 - once the server is enabled to respond to an OCSP request
1763 it does even when not requested, resulting in a stapling non-aware
1764 client dropping the TLS connection.
1766 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1767 support variable-length bit vectors. No functional change.
1769 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1770 The I= interface field on outgoing lines is now after the H= remote
1771 host field, same as incoming lines. There is a separate
1772 outgoing_interface log selector which allows you to disable the
1775 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1776 If not running log_selector +smtp_connection the mainlog would be held
1777 open indefinitely after a "too many connections" event, including to a
1778 deleted file after a log rotate. Leave the per net connection logging
1779 leaving it open for efficiency as that will be quickly detected by the
1780 check on the next write.
1782 HS/01 Bug 1671: Fix post transport crash.
1783 Processing the wait-<transport> messages could crash the delivery
1784 process if the message IDs didn't exist for some reason. When
1785 using 'split_spool_directory=yes' the construction of the spool
1786 file name failed already, exposing the same netto behaviour.
1788 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1789 mime_regex ACL conditions.
1791 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1792 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1793 to HELO, local diagnostic string.
1795 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1796 log line to debug. Even when configured with a tls authenticator many
1797 client connections are expected to not authenticate in this way, so
1798 an authenticate fail is not an error.
1800 HS/02 Add the Exim version string to the process info. This way exiwhat
1801 gives some more detail about the running daemon.
1803 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1804 matter for fast-change records such as DNSBLs.
1806 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1807 retry record, even if constant. There may be multiple transports with
1808 different interface settings and the retry behaviour needs to be kept
1811 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1813 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1815 JH/10 Bug 840: fix log_defer_output option of pipe transport
1817 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1818 hosts_randomize. This matters a lot when combined with mua_wrapper.
1820 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1821 ${quote_pgsql:<string>} operator.
1823 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1825 JH/14 Change header file name for freeradius-client. Relevant if compiling
1826 with Radius support; from the Gentoo tree and checked under Fedora.
1828 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1830 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1831 option transport_filter, meaning no filtering.
1833 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1835 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1836 defaults to "*" (all hosts). The variable is now available when not built
1837 with TLS, default unset, mainly to enable keeping the testsuite sane.
1838 If a server certificate is not supplied (via tls_certificate) an error is
1839 logged, and clients will find TLS connections fail on startup. Presumably
1840 they will retry in-clear.
1841 Packagers of Exim are strongly encouraged to create a server certificate
1842 at installation time.
1844 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1845 with the $config_file variable.
1847 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1848 in transport context, after the attempt, and per-recipient. The latter type
1849 is per host attempted. The event data is the error message, and the errno
1850 information encodes the lookup type (A vs. MX) used for the (first) host,
1851 and the trailing two digits of the smtp 4xx response.
1853 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1854 to write to mainlog (or rejectlog, paniclog) in the window between file
1855 creation and permissions/ownership being changed. Particularly affects
1856 installations where exicyclog is run as root, rather than exim user;
1857 result is that the running daemon panics and dies.
1859 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1861 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1862 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1863 "pri" and "weight". Note that the previous implicit priority given by the
1864 list order is no longer honoured.
1866 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1867 for DKIM processing.
1869 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1870 by defining SUPPORT_SOCKS.
1872 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1873 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1874 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1875 variables are renamed to proxy_{local,external}_{address,port}.
1877 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1878 for a build by defining SUPPORT_I18N
1880 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1881 of the query string, and make ${quote_redis:} do that quoting.
1883 JH/29 Move Events support from Experimental to mainline, enabled by default
1884 and removable for a build by defining DISABLE_EVENT.
1886 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1888 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1889 cached by the daemon.
1891 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1892 by defining LOOKUP_REDIS. The libhiredis library is required.
1894 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1895 keys are given for lookup.
1897 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1898 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1899 only supported when built with TLS support. The PolarSSL SHA routines
1900 are still used when the TLS library is too old for convenient support.
1902 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1903 openssl_options), for security. OpenSSL forces this from version 1.1.0
1904 server-side so match that on older versions.
1906 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1907 allocation for $value could be released as the expansion processing
1908 concluded, but leaving the global pointer active for it.
1910 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1911 and to use the domains and local_parts ACL conditions.
1913 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1914 incorrectly not doubled on cutthrough transmission, hence seen as a
1915 body-termination at the receiving system - resulting in truncated mails.
1916 Commonly the sender saw a TCP-level error, and retransmitted the message
1917 via the normal store-and-forward channel. This could result in duplicates
1918 received - but deduplicating mailstores were liable to retain only the
1919 initial truncated version.
1921 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1923 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1925 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1926 we're in there, support oversigning also; bug 1309.
1928 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1930 HS/04 Add support for keep_environment and add_environment options.
1932 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1933 either intentional arithmetic overflow during PRNG, or testing config-
1936 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1937 delivery resulted in actual delivery. Cancel cutthrough before DATA
1940 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1941 rejects a recipient: pass the reject to the originator.
1943 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1944 Many were false-positives and ignorable, but it's worth fixing the
1947 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1948 for the new environment-manipulation done at startup. Move the routines
1949 from being local to tls.c to being global via the os.c file.
1951 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1952 an extract embedded as result-arg for a map, the first arg for extract
1953 is unavailable so we cannot tell if this is a numbered or keyed
1954 extraction. Accept either.
1960 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1963 JH/02 The smtp transport option "multi_domain" is now expanded.
1965 JH/03 The smtp transport now requests PRDR by default, if the server offers
1968 JH/04 Certificate name checking on server certificates, when exim is a client,
1969 is now done by default. The transport option tls_verify_cert_hostnames
1970 can be used to disable this per-host. The build option
1971 EXPERIMENTAL_CERTNAMES is withdrawn.
1973 JH/05 The value of the tls_verify_certificates smtp transport and main options
1974 default to the word "system" to access the system default CA bundle.
1975 For GnuTLS, only version 3.0.20 or later.
1977 JH/06 Verification of the server certificate for a TLS connection is now tried
1978 (but not required) by default. The verification status is now logged by
1979 default, for both outbound TLS and client-certificate supplying inbound
1982 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1985 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1986 Status Notification (bounce) messages are now MIME format per RFC 3464.
1987 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1988 under the control of the dsn_advertise_hosts option, and routers may
1989 have a dsn_lasthop option.
1991 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1992 default, modifiable by a malware= option. The list separator for
1993 the options can now be changed in the usual way. Bug 68.
1995 JH/10 The smtp_receive_timeout main option is now expanded before use.
1997 JH/11 The incoming_interface log option now also enables logging of the
1998 local interface on delivery outgoing connections.
2000 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2001 if the interface and destination host and port all match.
2003 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2006 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2007 Patch from Andrew Lewis.
2009 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2010 now supports optional time-restrictions, weighting, and priority
2011 modifiers per server. Patch originally by <rommer@active.by>.
2013 JH/16 The spamd_address main option now supports a mixed list of local
2014 and remote servers. Remote servers can be IPv6 addresses, and
2015 specify a port-range.
2017 JH/17 Bug 68: The spamd_address main option now supports an optional
2018 timeout value per server.
2020 JH/18 Bug 1581: Router and transport options headers_add/remove can
2021 now have the list separator specified.
2023 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2026 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2029 JH/21 Support for the A6 type of dns record is withdrawn.
2031 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2032 rather than the verbs used.
2034 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2035 from 255 to 1024 chars.
2037 JH/24 Verification callouts now attempt to use TLS by default.
2039 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2040 are generic router options now. The defaults didn't change.
2042 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2043 Original patch from Alexander Shikoff, worked over by JH.
2045 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2046 files if ZCAT_COMMAND is not executable.
2048 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2050 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2052 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2053 Normally benign, it bites when the pair was led to by a CNAME;
2054 modern usage is to not canonicalize the domain to a CNAME target
2055 (and we were inconsistent anyway for A-only vs AAAA+A).
2057 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2059 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2060 when evaluating $sender_host_dnssec.
2062 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2063 $sender_helo_dnssec variable.
2065 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2067 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2069 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2071 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2072 documented as working, but never had. Support all but $spam_report.
2074 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2075 added for tls authenticator.
2077 HS/03 Add perl_taintmode main config option
2083 TL/01 When running the test suite, the README says that variables such as
2084 no_msglog_check are global and can be placed anywhere in a specific
2085 test's script, however it was observed that placement needed to be near
2086 the beginning for it to behave that way. Changed the runtest perl
2087 script to read through the entire script once to detect and set these
2088 variables, reset to the beginning of the script, and then run through
2089 the script parsing/test process like normal.
2091 TL/02 The BSD's have an arc4random API. One of the functions to induce
2092 adding randomness was arc4random_stir(), but it has been removed in
2093 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2094 function when detected.
2096 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2097 cause callback expansion.
2099 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2100 syntax errors in an expansion can be treated as a string instead of
2101 logging or causing an error, due to the internal use of bool_lax
2102 instead of bool when processing it.
2104 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2105 server certificates when making smtp deliveries.
2107 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2109 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2111 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2113 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2114 Merged patch from Sebastian Wiedenroth.
2116 JH/05 Fix results-pipe from transport process. Several recipients, combined
2117 with certificate use, exposed issues where response data items split
2118 over buffer boundaries were not parsed properly. This eventually
2119 resulted in duplicates being sent. This issue only became common enough
2120 to notice due to the introduction of connection certificate information,
2121 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2123 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2124 size buffer was used, resulting in syntax errors when an expansion
2127 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2128 version 3.3.6 or later.
2130 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2131 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2132 both become "event_action", the variables become $event_name, $event_data
2133 and $event_defer_errno. There is a new variable $verify_mode, usable in
2134 routers, transports and related events. The tls:cert event is now also
2135 raised for inbound connections, if the main configuration event_action
2138 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2139 early OCSP support, but no stapling (appears to be less than 1.0.0).
2141 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2142 server certificate names available under the smtp transport option
2143 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2146 JH/10 Time-related extraction expansions from certificates now use the main
2147 option "timezone" setting for output formatting, and are consistent
2148 between OpenSSL and GnuTLS compilations. Bug 1541.
2150 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2151 encoded parameter in the incoming message. Bug 1558.
2153 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2154 include certificate info, eximon was claiming there were spoolfile
2157 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2159 JH/14 Log delivery-related information more consistently, using the sequence
2160 "H=<name> [<ip>]" wherever possible.
2162 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2163 are problematic for Debian distribution, omit them from the release
2166 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2168 JH/16 Fix string representation of time values on 64bit time_t architectures.
2171 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2172 output list separator was used.
2177 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2178 checkers that were complaining about end of non-void function with no
2181 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2182 This was a regression introduced in 4.83 by another bugfix.
2184 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2186 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2187 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2193 TF/01 Correctly close the server side of TLS when forking for delivery.
2195 When a message was received over SMTP with TLS, Exim failed to clear up
2196 the incoming connection properly after forking off the child process to
2197 deliver the message. In some situations the subsequent outgoing
2198 delivery connection happened to have the same fd number as the incoming
2199 connection previously had. Exim would try to use TLS and fail, logging
2200 a "Bad file descriptor" error.
2202 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2203 utilities have not been installed.
2205 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2206 temporary space as the ACL may create new global variables.
2208 TL/01 LDAP support uses per connection or global context settings, depending
2209 upon the detected version of the libraries at build time.
2211 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2212 to extract and use the src ip:port in logging and expansions as if it
2213 were a direct connection from the outside internet. PPv2 support was
2214 updated based on HAProxy spec change in May 2014.
2216 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2218 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2219 Properly escape header and check for NULL return.
2221 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2224 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2226 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2227 characters in header names, implemented as a verify condition.
2228 Contributed by Michael Fischer v. Mollard.
2230 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2231 results permerror and temperror. Previous values are deprecated but
2232 still accepted. In a future release, err_perm and err_temp will be
2233 completely removed, which will be a backward incompatibility if the
2234 ACL tests for either of these two old results. Patch contributed by
2235 user bes-internal on the mailing list.
2237 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2239 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2240 selectors, in both main and reject logs.
2242 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2245 JH/07 Add malware type "sock" for talking to simple daemon.
2247 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2249 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2250 routers/transports under cutthrough routing.
2252 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2253 numbers. Touch up "bool" conditional to keep the same definition.
2255 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2257 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2259 JH/12 Expand items in router/transport headers_add or headers_remove lists
2260 individually rather than the list as a whole. Bug 1452.
2262 Required for reasonable handling of multiple headers_ options when
2263 they may be empty; requires that headers_remove items with embedded
2264 colons must have them doubled (or the list-separator changed).
2266 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2267 view the policy declared in the DMARC record. Currently, $dmarc_status
2268 is a combined value of both the record presence and the result of the
2271 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2273 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2274 dnslookup router and the smtp transport (applying to the forward
2277 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2278 of ldap servers used for a specific lookup. Patch provided by Heiko
2281 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2282 New variable $lookup_dnssec_authenticated for observability.
2284 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2285 Patch submitted by Lars Timman.
2287 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2289 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2290 Requires trusted mode and valid format message id, aborts otherwise.
2291 Patch contributed by Heiko Schlichting.
2293 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2294 certextract with support for various fields. Bug 1358.
2296 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2297 is requested by default, modifiable by smtp transport option
2300 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2301 operate on certificate variables to give certificate fingerprints
2302 Also new ${sha256:cert_variable}.
2304 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2306 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2307 Christian Aistleitner.
2309 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2311 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2312 file. Patch from Wolfgang Breyha.
2314 JH/25 Expand the coverage of the delivery $host and $host_address to
2315 client authenticators run in verify callout. Bug 1476.
2317 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2318 align with daemon_smtp_ports. Bug 72.
2320 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2321 support and error reporting did not work properly.
2323 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2324 and is readable. Patch from Andrew Colin Kissa.
2326 TL/14 Enhance documentation of ${run expansion and how it parses the
2327 commandline after expansion, particularly in the case when an
2328 unquoted variable expansion results in an empty value.
2330 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2332 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2333 by renaming away from it. Addresses GH issue 15, reported by
2336 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2338 TL/15 SECURITY: prevent double expansion in math comparison functions
2339 (can expand unsanitized data). Not remotely exploitable.
2346 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2348 PP/02 Make -n do something, by making it not do something.
2349 When combined with -bP, the name of an option is not output.
2351 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2354 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2355 $sender_host_name and config options to manage this, and basic check
2358 PP/05 DSCP support for outbound connections and control modifier for inbound.
2360 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2361 (Only plugin which currently uses this is kerberos4, which nobody should
2362 be using, but we should make it available and other future plugins might
2363 conceivably use it, even though it would break NAT; stuff *should* be
2364 using channel bindings instead).
2366 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2367 name; added for Sendmail compatibility; requires admin caller.
2368 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2369 just ignore it); requires trusted caller.
2370 Also parse but ignore: -Ac -Am -X<logfile>
2373 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2375 TL/02 Add +smtp_confirmation as a default logging option.
2377 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2378 Patch by Magnus Holmgren from 2007-02-20.
2380 TL/04 Bugzilla 1281 - Spec typo.
2381 Bugzilla 1283 - Spec typo.
2382 Bugzilla 1290 - Spec grammar fixes.
2384 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2386 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2388 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2389 mailing list by by Dmitry Isaikin.
2391 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2393 JH/02 Support "G" suffix to numbers in ${if comparisons.
2395 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2397 NM/01 Bugzilla 1197 - Spec typo
2398 Bugzilla 1196 - Spec examples corrections
2400 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2402 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2403 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2406 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2407 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2409 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2410 "acl {{name}{arg}...}", and optional args on acl condition
2413 JH/05 Permit multiple router/transport headers_add/remove lines.
2415 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2417 JH/07 Avoid using a waiting database for a single-message-only transport.
2418 Performance patch from Paul Fisher. Bugzilla 1262.
2420 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2423 JH/09 Add $headers_added variable, with content from use of ACL modifier
2424 add_header (but not yet added to the message). Bugzilla 199.
2426 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2427 Pulled from Bugzilla 817 by Wolfgang Breyha.
2429 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2431 (nb: this is the same fix as in Exim 4.80.1)
2433 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2436 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2439 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2441 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2442 advertise SMTP AUTH mechanism to us, instead of a generic
2443 protocol violation error. Also, make Exim more robust to bad
2444 data from the Dovecot auth socket.
2446 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2448 When a queue runner is handling a message, Exim first routes the
2449 recipient addresses, during which it prunes them based on the retry
2450 hints database. After that it attempts to deliver the message to
2451 any remaining recipients. It then updates the hints database using
2454 So if a recipient address works intermittently, it can get repeatedly
2455 deferred at routing time. The retry hints record remains fresh so the
2456 address never reaches the final cutoff time.
2458 This is a fairly common occurrence when a user is bumping up against
2459 their storage quota. Exim had some logic in its local delivery code
2460 to deal with this. However it did not apply to per-recipient defers
2461 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2463 This change adds a proper retry rule check during routing so that the
2464 final cutoff time is checked against the message's age. We only do
2465 this check if there is an address retry record and there is not a
2466 domain retry record; this implies that previous attempts to handle
2467 the address had the retry_use_local_parts option turned on. We use
2468 this as an approximation for the destination being like a local
2469 delivery, as in LMTP.
2471 I suspect this new check makes the old local delivery cutoff check
2472 redundant, but I have not verified this so I left the code in place.
2474 TF/02 Correct gecos expansion when From: is a prefix of the username.
2476 Test 0254 submits a message to Exim with the header
2480 When I ran the test suite under the user fanf2, Exim expanded
2481 the header to contain my full name, whereas it should have added
2482 a Resent-Sender: header. It erroneously treats any prefix of the
2483 username as equal to the username.
2485 This change corrects that bug.
2487 GF/01 DCC debug and logging tidyup
2488 Error conditions log to paniclog rather than rejectlog.
2489 Debug lines prefixed by "DCC: " to remove any ambiguity.
2491 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2493 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2494 Bug spotted by Jeremy Harris; was flawed since initial commit.
2495 Would have resulted in OCSP responses post-SNI triggering an Exim
2496 NULL dereference and crash.
2498 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2500 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2501 Bug detection, analysis and fix by Samuel Thibault.
2502 Bugzilla 1331, Debian bug #698092.
2504 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2506 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2507 Server implementation by Todd Lyons, client by JH.
2508 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2509 config variable "prdr_enable" controls whether the server
2510 advertises the facility. If the client requests PRDR a new
2511 acl_data_smtp_prdr ACL is called once for each recipient, after
2512 the body content is received and before the acl_smtp_data ACL.
2513 The client is controlled by both of: a hosts_try_prdr option
2514 on the smtp transport, and the server advertisement.
2515 Default client logging of deliveries and rejections involving
2516 PRDR are flagged with the string "PRDR".
2518 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2519 fclose(). Diagnosis by Todd Lyons.
2521 PP/17 Update configure.default to handle IPv6 localhost better.
2522 Patch by Alain Williams (plus minor tweaks).
2525 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2526 This is now consistent with GnuTLS, and is now documented: the
2527 previous undocumented portable approach to treating the option as
2528 unset was to force an expansion failure. That still works, and
2529 an empty string is now equivalent.
2531 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2532 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2533 not performing validation itself.
2535 PP/20 Added force_command boolean option to pipe transport.
2536 Patch from Nick Koston, of cPanel Inc.
2538 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2541 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2543 PP/21 Fix eximon continuous updating with timestamped log-files.
2544 Broken in a format-string cleanup in 4.80, missed when I repaired the
2545 other false fix of the same issue.
2546 Report and fix from Heiko Schlichting.
2549 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2550 Report from Prashanth Katuri.
2552 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2553 It's SecureTransport, so affects any MacOS clients which use the
2554 system-integrated TLS libraries, including email clients.
2556 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2557 using a MIME ACL for non-SMTP local injection.
2558 Report and assistance in diagnosis by Warren Baker.
2560 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2562 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2564 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2565 last id that failed so it may be referenced in subsequent ACL's.
2567 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2570 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2571 ldap library initialization, allowing self-signed CA's to be
2572 used. Also properly sets require_cert option later in code by
2573 using NULL (global ldap config) instead of ldap handle (per
2574 session). Bug diagnosis and testing by alxgomz.
2576 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2577 the src/util/ subdirectory.
2579 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2580 renamed to Transport Post Delivery Action by Jeremy Harris, as
2583 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2584 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2585 redis_servers = needs to be configured which will be used by the redis
2586 lookup. Patch from Warren Baker, of The Packet Hub.
2588 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2590 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2591 hostname or reverse DNS when processing a host list. Used suggestions
2592 from multiple comments on this bug.
2594 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2596 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2597 Missed a few lines, added it to make the runtest require no keyboard
2600 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2601 contains upper case chars. Make router use caseful_local_part.
2603 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2604 support when GnuTLS has been built with p11-kit.
2610 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2612 This, or similar/improved, will also be change PP/11 of 4.82.
2618 PP/01 Handle short writes when writing local log-files.
2619 In practice, only affects FreeBSD (8 onwards).
2620 Bugzilla 1053, with thanks to Dmitry Isaikin.
2622 NM/01 Bugzilla 949 - Documentation tweak
2624 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2627 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2629 PP/02 Implemented gsasl authenticator.
2631 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2633 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2634 `pkg-config foo` for cflags/libs.
2636 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2637 with rest of GSASL and with heimdal_gssapi.
2639 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2640 `pkg-config foo` for cflags/libs for the TLS implementation.
2642 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2643 properties get this fed in as external SSF. A number of robustness
2644 and debugging improvements to the cyrus_sasl authenticator.
2646 PP/08 cyrus_sasl server now expands the server_realm option.
2648 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2649 Patch by Jeremy Harris.
2651 PP/10 Added dbmjz lookup type.
2653 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2655 PP/12 MAIL args handles TAB as well as SP, for better interop with
2656 non-compliant senders.
2657 Analysis and variant patch by Todd Lyons.
2659 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2660 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2661 Patch from Dirk Mueller <dmueller@suse.com>
2663 PP/13 tls_peerdn now print-escaped for spool files.
2664 Observed some $tls_peerdn in wild which contained \n, which resulted
2665 in spool file corruption.
2667 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2668 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2669 or write after TLS renegotiation, which otherwise led to messages
2672 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2673 as a tracking header (ie: a signed header comes before the signature).
2674 Patch from Wolfgang Breyha.
2676 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2677 comma-sep list; embedded commas doubled.
2679 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2681 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2683 Report and patch from Dmitry Banschikov.
2685 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2686 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2687 failures. We appear to now support TLS1.1+ with Exim.
2689 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2690 lets Exim select keys and certificates based upon TLS SNI from client.
2691 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2692 before an outbound SMTP session. New log_selector, +tls_sni.
2694 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2695 NULL dereference. Report and patch from Alun Jones.
2697 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2698 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2699 Not seeing resolver debug output on NetBSD, but suspect this is a
2700 resolver implementation change.
2702 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2703 Left warnings. Added "eximon gdb" invocation mode.
2705 PP/21 Defaulting "accept_8bitmime" to true, not false.
2707 PP/22 Added -bw for inetd wait mode support.
2709 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2710 locate the relevant includes and libraries. Made this the default.
2712 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2713 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2715 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2716 This may cause build issues on older platforms.
2718 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2719 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2720 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2721 Added SNI support via GnuTLS too.
2722 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2724 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2726 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2727 Applied second patch from Janne, implementing suggestion to default
2728 multiple-strings-in-record handling to match SPF spec.
2730 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2732 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2733 read-only, out of scope).
2734 Patch from Wolfgang Breyha, report from Stuart Northfield.
2736 PP/29 Fix three issues highlighted by clang analyser static analysis.
2737 Only crash-plausible issue would require the Cambridge-specific
2738 iplookup router and a misconfiguration.
2739 Report from Marcin Mirosław.
2741 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2743 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2744 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2745 As part of this, removing so much warning spew let me fix some minor
2746 real issues in debug logging.
2748 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2749 assignment on my part. Fixed.
2751 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2752 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2753 Janne Snabb (who went above and beyond: thank you).
2755 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2756 string otherwise requires a connection and a bunch more work and it's
2757 relatively easy to get wrong. Should also expose TLS library linkage
2760 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2761 64-bit ${eval} (JH/03).
2763 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2764 GNU libc to support some of the 64-bit stuff, should not lead to
2765 conflicts. Defined before os.h is pulled in, so if a given platform
2766 needs to override this, it can.
2768 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2769 protection layer was required, which is not implemented.
2770 Bugzilla 1254, patch from Wolfgang Breyha.
2772 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2773 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2774 tls_dhparam take prime identifiers. Also unbreak combination of
2775 OpenSSL+DH_params+TLSSNI.
2777 PP/39 Disable SSLv2 by default in OpenSSL support.
2783 PP/01 Solaris build fix for Oracle's LDAP libraries.
2784 Bugzilla 1109, patch from Stephen Usher.
2786 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2788 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2791 TF/02 Fix a couple more cases where we did not log the error message
2792 when unlink() failed. See also change 4.74-TF/03.
2794 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2795 lock up or crash if it happened to be inside a call to libc when it
2796 got a SIGUSR1 from exiwhat.
2798 The SIGUSR1 handler appends the current process status to the process
2799 log which is later printed by exiwhat. It used to use the general
2800 purpose logging code to do this, but several functions it calls are
2801 not safe for signals.
2803 The new output code in the SIGUSR1 handler is specific to the process
2804 log, and simple enough that it's easy to inspect for signal safety.
2805 Removing some special cases also simplifies the general logging code.
2806 Removing the spurious timestamps from the process log simplifies
2809 TF/04 Improved ratelimit ACL condition.
2811 The /noupdate option has been deprecated in favour of /readonly which
2812 has clearer semantics. The /leaky, /strict, and /readonly update modes
2813 are mutually exclusive. The update mode is no longer included in the
2814 database key; it just determines when the database is updated. (This
2815 means that when you upgrade Exim will forget old rate measurements.)
2817 Exim now checks that the per_* options are used with an update mode that
2818 makes sense for the current ACL. For example, when Exim is processing a
2819 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2820 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2821 must specify per_mail/readonly. If you omit the update mode it defaults to
2822 /leaky where that makes sense (as before) or /readonly where required.
2824 The /noupdate option is now undocumented but still supported for
2825 backwards compatibility. It is equivalent to /readonly except that in
2826 ACLs where /readonly is required you may specify /leaky/noupdate or
2827 /strict/noupdate which are treated the same as /readonly.
2829 A useful new feature is the /count= option. This is a generalization
2830 of the per_byte option, so that you can measure the throughput of other
2831 aggregate values. For example, the per_byte option is now equivalent
2832 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2834 The per_rcpt option has been generalized using the /count= mechanism
2835 (though it's more complicated than the per_byte equivalence). When it is
2836 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2837 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2838 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2839 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2840 Note that using per_rcpt with a non-readonly update mode in more than
2841 one ACL will cause the recipients to be double-counted. (The per_mail
2842 and per_byte options don't have this problem.)
2844 The handling of very low rates has changed slightly. If the computed rate
2845 is less than the event's count (usually one) then this event is the first
2846 after a long gap. In this case the rate is set to the same as this event's
2847 count, so that the first message of a spam run is counted properly.
2849 The major new feature is a mechanism for counting the rate of unique
2850 events. The new per_addr option counts the number of different
2851 recipients that someone has sent messages to in the last time period. It
2852 behaves like per_rcpt if all the recipient addresses are different, but
2853 duplicate recipient addresses do not increase the measured rate. Like
2854 the /count= option this is a general mechanism, so the per_addr option
2855 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2856 example, measure the rate that a client uses different sender addresses
2857 with the options per_mail/unique=$sender_address. There are further
2858 details in the main documentation.
2860 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2862 TF/06 Removed a few PCRE remnants.
2864 TF/07 Automatically extract Exim's version number from tags in the git
2865 repository when doing development or release builds.
2867 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2868 Bugzilla 879. Patch from Paul Fisher.
2870 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2871 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2874 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2876 PP/05 Variable $av_failed, true if the AV scanner deferred.
2877 Bugzilla 1078. Patch from John Horne.
2879 PP/06 Stop make process more reliably on build failure.
2880 Bugzilla 1087. Patch from Heiko Schlittermann.
2882 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2883 Bugzilla 1089. Patch from Heiko Schlittermann.
2885 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2886 Bugzilla 1131. Patch from Holger Weiß.
2888 PP/09 Handle IPv6 addresses with SPF.
2889 Bugzilla 860. Patch from Wolfgang Breyha.
2891 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2893 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2896 PP/11 match_* no longer expand right-hand-side by default.
2897 New compile-time build option, EXPAND_LISTMATCH_RHS.
2898 New expansion conditions, "inlist", "inlisti".
2900 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2902 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2904 PP/14 fix log_write() format string regression from TF/03.
2905 Bugzilla 1152. Patch from Dmitry Isaikin.
2911 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2913 PP/02 Harmonised TLS library version reporting; only show if debugging.
2914 Layout now matches that introduced for other libraries in 4.74 PP/03.
2916 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2918 PP/04 New "dns_use_edns0" global option.
2920 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2923 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2924 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2926 TK/01 Updated PolarSSL code to 0.14.2.
2927 Bugzilla 1097. Patch from Andreas Metzler.
2929 PP/07 Catch divide-by-zero in ${eval:...}.
2930 Fixes bugzilla 1102.
2932 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2935 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2936 format-string attack -- SECURITY: remote arbitrary code execution.
2938 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2939 time unintentionally subject to list matching rules, letting the header
2940 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2941 arbitrary string expansion). This allowed for information disclosure.
2943 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2944 INT_MIN/-1 -- value coerced to INT_MAX.
2950 NM/01 Workaround for PCRE version dependency in version reporting
2953 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2954 This fixes portability to compilers other than gcc, notably
2955 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2957 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2958 makefiles for portability to HP-UX and POSIX correctness.
2960 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2961 Also via indented variable definition in the Makefile.
2962 (Debugging by Oliver Heesakkers).
2964 PP/02 Restore caching of spamd results with expanded spamd_address.
2965 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2967 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2968 Improves build reliability. Fix from: Frank Elsner
2970 NM/02 Fix wide character breakage in the rfc2047 coding
2971 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2973 NM/03 Allow underscore in dnslist lookups
2974 Fixes bug 1026. Patch from Graeme Fowler
2976 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2977 Code patches from Adam Ciarcinski of NetBSD.
2979 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2982 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2983 is logged, to avoid truncation. Patch from John Horne.
2985 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2986 Patch from Jakob Hirsch.
2988 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2989 SQL string expansion failure details.
2990 Patch from Andrey Oktyabrski.
2992 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2993 Patch from Simon Arlott.
2995 PP/09 New lookups functionality failed to compile on old gcc which rejects
2996 extern declarations in function scope.
2997 Patch from Oliver Fleischmann
2999 PP/10 Use sig_atomic_t for flags set from signal handlers.
3000 Check getgroups() return and improve debugging.
3001 Fixed developed for diagnosis in bug 927 (which turned out to be
3004 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3005 Patch from Mark Zealey.
3007 PP/12 Bugzilla 1056: Improved spamd server selection.
3008 Patch from Mark Zealey.
3010 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3011 Based on patch from Heiko Schlittermann.
3013 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3014 Patch from Uwe Doering, sign-off by Michael Haardt.
3016 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3017 variable declaration deep within a block. Bug and patch from
3020 PP/15 lookups-Makefile IRIX compatibility coercion.
3022 PP/16 Make DISABLE_DKIM build knob functional.
3024 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3025 Patch by Simon Arlott
3027 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3028 variable argument macros. Our copy now differs from upstream.
3034 TF/01 Failure to get a lock on a hints database can have serious
3035 consequences so log it to the panic log.
3037 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3038 controlled using the smtp_confirmation log selector.
3040 TF/03 Include the error message when we fail to unlink a spool file.
3042 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3043 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3044 for maintaining out-of-tree patches for some time.
3046 PP/01 Bugzilla 139: Documentation and portability issues.
3047 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3048 Handle per-OS dynamic-module compilation flags.
3050 PP/02 Let /dev/null have normal permissions.
3051 The 4.73 fixes were a little too stringent and complained about the
3052 permissions on /dev/null. Exempt it from some checks.
3053 Reported by Andreas M. Kirchwitz.
3055 PP/03 Report version information for many libraries, including
3056 Exim version information for dynamically loaded libraries. Created
3057 version.h, now support a version extension string for distributors
3058 who patch heavily. Dynamic module ABI change.
3060 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3061 privilege escalation vulnerability whereby the Exim run-time user
3062 can cause root to append content of the attacker's choosing to
3065 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3068 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3069 If dropping privileges for untrusted macros, we disabled normal logging
3070 on the basis that it would fail; for the Exim run-time user, this is not
3071 the case, and it resulted in successful deliveries going unlogged.
3072 Fixed. Reported by Andreas Metzler.
3078 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3079 only prepend for the Resent-* case. Fixes regression introduced in
3080 Exim 4.70 by NM/22 for Bugzilla 607.
3082 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3083 increasing numbers of administrators be bitten by this.
3085 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3087 PP/03 Bugzilla 994: added openssl_options main configuration option.
3089 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3091 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3093 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3095 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3096 without a peer certificate, leading to a segfault because of an
3097 assumption that peers always have certificates. Be a little more
3098 paranoid. Problem reported by Martin Tscholak.
3100 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3101 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3102 NB: ClamAV planning to remove STREAM in "middle of 2010".
3103 CL also introduces -bmalware, various -d+acl logging additions and
3104 more caution in buffer sizes.
3106 PP/09 Implemented reverse_ip expansion operator.
3108 PP/10 Bugzilla 937: provide a "debug" ACL control.
3110 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3112 PP/12 Bugzilla 973: Implement --version.
3114 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3116 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3118 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3120 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3121 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3122 ignore trailing whitespace.
3124 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3126 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3129 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3130 Notification from Dr Andrew Aitchison.
3132 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3133 ExtendedDetectionInfo response format.
3134 Notification from John Horne.
3136 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3139 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3140 XSL and documented dependency on system catalogs, with examples of how
3143 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3146 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3147 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3148 configuration file which is writeable by the Exim user or group.
3150 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3151 of configuration files to cover files specified with the -C option if
3152 they are going to be used with root privileges, not just the default
3155 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3156 option (effectively making it always true).
3158 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3159 files to be used while preserving root privileges.
3161 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3162 that rogue child processes cannot use them.
3164 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3165 run-time user, instead of root.
3167 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3168 Exim run-time user without dropping privileges.
3170 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3171 result string, instead of calling string_vformat() twice with the same
3174 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3175 for other users. Others should always drop root privileges if they use
3176 -C on the command line, even for a whitelisted configure file.
3178 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3180 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3186 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3187 $data_path, and $header_path variables; fixed documentation bugs and
3190 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3191 exipick to access non-standard spools, including the "frozen" queue
3194 NM/01 Bugzilla 965: Support mysql stored procedures.
3195 Patch from Alain Williams
3197 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3199 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3200 Patch from Andreas Metzler
3202 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3203 Patch from Kirill Miazine
3205 NM/05 Bugzilla 671: Added umask to procmail example.
3207 JJ/03 installed exipick 20100323.0, fixing doc bug
3209 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3210 directory. Notification and patch from Dan Rosenberg.
3212 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3214 TK/02 Improve log output when DKIM signing operation fails.
3216 MH/01 Treat the transport option dkim_domain as a colon separated
3217 list, not as a single string, and sign the message with each element,
3218 omitting multiple occurences of the same signer.
3220 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3221 Bugzilla 985, 986. Patch by Simon Arlott
3223 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3224 Patch by Simon Arlott
3226 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3227 MBX locking. Notification from Dan Rosenberg.
3233 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3235 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3237 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3239 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3241 NM/04 Bugzilla 915: Flag broken perl installation during build.
3247 TK/01 Added patch by Johannes Berg that expands the main option
3248 "spamd_address" if it starts with a dollar sign.
3250 TK/02 Write list of recipients to X-Envelope-Sender header when building
3251 the mbox-format spool file for content scanning (suggested by Jakob
3254 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3255 (http://www.dcc-servers.net/) support via dccifd. Activated by
3256 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3258 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3259 by Mark Daniel Reidel <mr@df.eu>.
3261 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3262 When building exim an external PCRE library is now needed -
3263 PCRE is a system library on the majority of modern systems.
3264 See entry on PCRE_LIBS in EDITME file.
3266 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3267 conversation. Added nologin parameter to request.
3268 Patch contributed by Kirill Miazine.
3270 TF/01 Do not log submission mode rewrites if they do not change the address.
3272 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3274 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3275 log files in place. Contributed by Roberto Lima.
3277 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3279 TF/03 Bugzilla 615: When checking the local_parts router precondition
3280 after a local_part_suffix or local_part_prefix option, Exim now
3281 does not use the address's named list lookup cache, since this
3282 contains cached lookups for the whole local part.
3284 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3285 Robert Millan. Documentation is in experimental-spec.txt.
3287 TF/04 Bugzilla 668: Fix parallel build (make -j).
3289 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3291 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3292 Patch provided by Jan Srzednicki.
3294 TF/05 Leading white space used to be stripped from $spam_report which
3295 wrecked the formatting. Now it is preserved.
3297 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3298 that they are available at delivery time.
3300 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3302 TF/08 TLS error reporting now respects the incoming_interface and
3303 incoming_port log selectors.
3305 TF/09 Produce a more useful error message if an SMTP transport's hosts
3306 setting expands to an empty string.
3308 NM/06 Bugzilla 744: EXPN did not work under TLS.
3309 Patch provided by Phil Pennock.
3311 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3312 Patch provided by Richard Godbee.
3314 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3315 acl_smtp_notquit, added index entry.
3317 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3318 Patch provided by Eugene Bujak.
3320 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3321 accept(). Patch provided by Maxim Dounin.
3323 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3324 Patch provided by Phil Pennock.
3326 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3328 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3329 Patch provided by Brad "anomie" Jorsch.
3331 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3332 Patch provided by Dean Brooks.
3334 TK/05 Add native DKIM support (does not depend on external libraries).
3336 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3337 Patch provided by Graeme Fowler.
3339 NM/16 Bugzilla 851: Documentation example syntax fix.
3341 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3343 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3346 NM/19 Bugzilla 745: TLS version reporting.
3347 Patch provided by Phil Pennock.
3349 NM/20 Bugzilla 167: bool: condition support.
3350 Patch provided by Phil Pennock.
3352 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3353 clients. Patch provided by Phil Pennock.
3355 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3356 Patch provided by Brad "anomie" Jorsch.
3358 NM/23 Bugzilla 687: Fix misparses in eximstats.
3359 Patch provided by Heiko Schlittermann.
3361 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3362 Patch provided by Heiko Schlittermann.
3364 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3365 plus update to original patch.
3367 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3369 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3370 Patch provided by David Brownlee.
3372 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3374 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3376 NM/30 Bugzilla 888: TLS documentation bugfixes.
3378 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3380 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3381 Unlike the original bugzilla I have changed all shell scripts in src tree.
3383 NM/33 Bugzilla 898: Transport filter timeout fix.
3384 Patch by Todd Rinaldo.
3386 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3387 Patch by Serge Demonchaux.
3389 NM/35 Bugzilla 39: Base64 decode bug fixes.
3390 Patch by Jakob Hirsch.
3392 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3394 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3396 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3398 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3404 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3405 ALT-N's libdkim that I have put here:
3406 http://duncanthrax.net/exim-experimental/
3408 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3409 were called 'true' and it seems that C99 defines that as a reserved
3410 keyword to be used with 'bool' variable types. That means you could
3411 not include C99-style headers which use bools without triggering
3412 build errors in sieve.c.
3414 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3415 as mailq or other aliases. Changed the --help handling significantly
3416 to do whats expected. exim_usage() emits usage/help information.
3418 SC/01 Added the -bylocaldomain option to eximstats.
3420 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3422 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3424 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3430 PH/01 Another patch from the Sieve maintainer.
3432 PH/02 When an IPv6 address is converted to a string for single-key lookup
3433 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3434 dots are used instead of colons so that keys in lsearch files need not
3435 contain colons. This was done some time before quoting was made available
3436 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3437 (notated using the quote facility) so as to distinguish them from IPv4
3438 keys. This meant that lookups for IP addresses in host lists did not work
3439 for iplsearch lookups.
3441 This has been fixed by arranging for IPv6 addresses to be expressed with
3442 colons if the lookup type is iplsearch. This is not incompatible, because
3443 previously such lookups could never work.
3445 The situation is now rather anomalous, since one *can* have colons in
3446 ordinary lsearch keys. However, making the change in all cases is
3447 incompatible and would probably break a number of configurations.
3449 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3452 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3453 conversion specification without a maximum field width, thereby enabling
3454 a rogue spamd server to cause a buffer overflow. While nobody in their
3455 right mind would setup Exim to query an untrusted spamd server, an
3456 attacker that gains access to a server running spamd could potentially
3457 exploit this vulnerability to run arbitrary code as the Exim user.
3459 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3460 $primary_hostname instead of what libspf2 thinks the hosts name is.
3462 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3463 a directory entry by the name of the lookup key. Previously, if a
3464 symlink pointed to a non-existing file or a file in a directory that
3465 Exim lacked permissions to read, a lookup for a key matching that
3466 symlink would fail. Now it is enough that a matching directory entry
3467 exists, symlink or not. (Bugzilla 503.)
3469 PH/03 The body_linecount and body_zerocount variables are now exported in the
3472 PH/04 Added the $dnslist_matched variable.
3474 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3475 This means they are set thereafter only if the connection becomes
3478 PH/06 Added the client_condition to authenticators so that some can be skipped
3479 by clients under certain conditions.
3481 PH/07 The error message for a badly-placed control=no_multiline_responses left
3482 "_responses" off the end of the name.
3484 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3486 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3487 (without spaces) instead of just copying the configuration text.
3489 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3491 PH/11 Added $max_received_linelength.
3493 PH/12 Added +ignore_defer and +include_defer to host lists.
3495 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3496 way in which PCRE > 7.0 is built.
3498 PH/14 Implemented queue_only_load_latch.
3500 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3501 MAIL command. The effect was to mangle the value on 64-bit systems.
3503 PH/16 Another patch from the Sieve maintainer.
3505 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3507 PH/18 If a system quota error occurred while trying to create the file for
3508 a maildir delivery, the message "Mailbox is full" was not appended to the
3509 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3510 only to a quota excession during the actual writing of the file.
3512 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3513 characters?) which causes problems in log lines. The DN values are now
3514 passed through string_printing() before being added to log lines.
3516 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3517 and InterBase are left for another time.)
3519 PH/21 Added message_body_newlines option.
3521 PH/22 Guard against possible overflow in moan_check_errorcopy().
3523 PH/23 POSIX allows open() to be a macro; guard against that.
3525 PH/24 If the recipient of an error message contained an @ in the local part
3526 (suitably quoted, of course), incorrect values were put in $domain and
3527 $local_part during the evaluation of errors_copy.
3533 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3534 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3537 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3538 issue a MAIL command.
3540 PH/02 In an ACL statement such as
3542 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3544 if a client was not listed at all, or was listed with a value other than
3545 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3546 the condition was not true (as it should be), so access was not denied.
3547 The bug was that the ! inversion was incorrectly passed on to the second
3548 item. This has been fixed.
3550 PH/03 Added additional dnslists conditions == and =& which are different from
3551 = and & when the dns lookup returns more than one IP address.
3553 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3554 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3556 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3557 FSYNC, which compiles an option called disable_fsync that allows for
3558 bypassing fsync(). The documentation is heavily laced with warnings.
3560 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3562 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3563 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3564 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3565 including adding "make clean"; (3) Added -fPIC when compiling the test
3566 dynamically loaded module, to get rid of a warning.
3568 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3569 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3570 = 0s. The bug is otherwise harmless.
3572 PH/07 There was a bug in the dovecot authenticator such that the value of
3573 $auth1 could be overwritten, and so not correctly preserved, after a
3574 successful authentication. This usually meant that the value preserved by
3575 the server_setid option was incorrect.
3577 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3579 PH/09 Installed PCRE release 7.0.
3581 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3582 run for batched SMTP input. It is now run at the start of every message
3583 in the batch. While fixing this I discovered that the process information
3584 (output by running exiwhat) was not always getting set for -bs and -bS
3585 input. This is fixed, and it now also says "batched" for BSMTP.
3587 PH/11 Added control=no_pipelining.
3589 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3590 patch, slightly modified), and move the expansion of helo_data till after
3591 the connection is made in the smtp transport (so it can use these
3594 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3596 PH/14 Added log_selector = +pid.
3598 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3600 PH/16 Add ${if forany and ${if forall.
3602 PH/17 Added dsn_from option to vary the From: line in DSNs.
3604 PH/18 Flush SMTP output before performing a callout, unless control =
3605 no_callout_flush is set.
3607 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3608 was true (the default) a successful delivery failed to delete the retry
3609 item, thus causing premature timeout of the address. The bug is now
3612 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3614 PH/21 Long custom messages for fakedefer and fakereject are now split up
3615 into multiline responses in the same way that messages for "deny" and
3616 other ACL rejections are.
3618 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3619 with slight modification.
3621 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3622 draft, changing the syntax and factoring some duplicate code".
3624 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3625 for deliveries of the second and subsequent messages over the same SMTP
3628 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3629 ${reduce, with only minor "tidies".
3631 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3633 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3634 expansion side effects.
3636 PH/27 When a message times out after an over-quota error from an Exim-imposed
3637 quota, the bounce message says "mailbox is full". This message was not
3638 being given when it was a system quota that was exceeded. It now should
3641 MH/03 Made $recipients available in local_scan(). local_scan() already has
3642 better access to the recipient list through recipients_list[], but
3643 $recipients can be useful in postmaster-provided expansion strings.
3645 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3646 in the case of a MAIL command with additional options following the
3647 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3648 were accidentally chopped off.
3650 PH/29 SMTP synchronization checks are implemented when a command is read -
3651 there is a check that no more input is waiting when there shouldn't be
3652 any. However, for some commands, a delay in an ACL can mean that it is
3653 some time before the response is written. In this time, more input might
3654 arrive, invalidly. So now there are extra checks after an ACL has run for
3655 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3656 pipelining has not been advertised.
3658 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3660 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3661 returned in the SMTP response when smtp_return_error_details was false.
3662 This has been fixed.
3664 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3665 instead of the C I/O that was originally supplied, because problems were
3666 reported on Solaris.
3668 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3669 Exim which did not show up earlier: it was assuming that a call to
3670 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3671 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3672 was a macro that became an assignment, so it seemed to work. This has
3673 changed to a proper function call with a void return, hence the compile
3674 error. Exim's code has been fixed.
3676 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3679 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3681 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3683 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3684 because it turns out to be too big or there is a timeout) there is no
3685 "Completed" line in the log. When some messages of this type were
3686 selected by exigrep, they were listed as "not completed". Others were
3687 picked up by some special patterns. I have improved the selection
3688 criteria to be more general.
3690 PH/38 The host_find_failed option in the manualroute router can now be set
3691 to "ignore", to completely ignore a host whose IP address cannot be
3692 found. If all hosts are ignored, the behaviour is controlled by the new
3693 host_all_ignored option.
3695 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3696 homing or because of multiple MX records with /mx) generated more than
3697 one IP address, and the following item turned out to be the local host,
3698 all the secondary addresses of the first item were incorrectly removed
3699 from the list, along with the local host and any following hosts (which
3700 is what is supposed to happen).
3702 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3703 whoever called Exim into the -H file. In the case of the daemon it was
3704 behaving confusingly. When first started, it used values for whoever
3705 started the daemon, but after a SIGHUP it used the Exim user (because it
3706 calls itself on a restart). I have changed the code so that it now always
3709 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3710 message are rejected with the same error (e.g. no authentication or bad
3711 sender address), and a DATA command is nevertheless sent (as can happen
3712 with PIPELINING or a stupid MUA), the error message that was given to the
3713 RCPT commands is included in the rejection of the DATA command. This is
3714 intended to be helpful for MUAs that show only the final error to their
3717 PH/42 Another patch from the Sieve maintainer.
3719 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3720 Eximstats - Fixed some broken HTML links and added missing column headers
3722 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3723 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3725 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3727 PH/43 Yet another patch from the Sieve maintainer.
3729 PH/44 I found a way to check for a TCP/IP connection going away before sending
3730 the response to the final '.' that terminates a message, but only in the
3731 case where the client has not sent further data following the '.'
3732 (unfortunately, this is allowed). However, in many cases there won't be
3733 any further data because there won't be any more messages to send. A call
3734 to select() can be used: if it shows that the input is "ready", there is
3735 either input waiting, or the socket has been closed. An attempt to read
3736 the next input character can distinguish the two cases. Previously, Exim
3737 would have sent an OK response which the client would never have see.
3738 This could lead to message repetition. This fix should cure that, at
3739 least in a lot of common cases.
3741 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3742 advertised in response to EHLO.
3748 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3749 fixed by 4.65/MH/01 (is this a record?) are fixed:
3751 (i) An empty string was always treated as zero by the numeric comparison
3752 operators. This behaviour has been restored.
3754 (ii) It is documented that the numeric comparison operators always treat
3755 their arguments as decimal numbers. This was broken in that numbers
3756 starting with 0 were being interpreted as octal.
3758 While fixing these problems I realized that there was another issue that
3759 hadn't been noticed. Values of message_size_limit (both the global option
3760 and the transport option) were treated as octal if they started with 0.
3761 The documentation was vague. These values are now always treated as
3762 decimal, and I will make that clear in the documentation.
3768 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3769 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3772 MH/01 Don't check that the operands of numeric comparison operators are
3773 integers when their expansion is in "skipping" mode (fixes bug
3774 introduced by 4.64-PH/07).
3776 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3777 child addresses, Exim now panics and dies. Previously, because the count
3778 is held in a short int, deliveries were likely to be lost. As such a
3779 large number of recipients for a single message is ridiculous
3780 (performance will be very, very poor), I have chosen to impose a limit
3781 rather than extend the field.
3787 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3788 leftover -K file (the existence of which was triggered by #402).
3789 While we were at it, introduced process PID as part of the -K
3790 filename. This should rule out race conditions when creating
3793 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3794 processing considerably. Previous code took too long for large mails,
3795 triggering a timeout which in turn triggers #401.
3797 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3798 in the DK code in transports.c. sendfile() is not really portable,
3799 hence the _LINUX specificness.
3801 TF/01 In the add_headers option to the mail command in an Exim filter,
3802 there was a bug that Exim would claim a syntax error in any
3803 header after the first one which had an odd number of characters
3806 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3807 callout verification, Exim cached a "reject" for the entire domain. This
3808 is correct for most verifications, but it is not correct for a recipient
3809 verification with use_sender or use_postmaster set, because in that case
3810 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3811 case of MAIL FROM:<> rejection from other early rejections (e.g.
3812 rejection of HELO). When verifying a recipient using a non-null MAIL
3813 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3814 Whatever the result of the callout, the value of the domain cache is
3815 left unchanged (for any other kind of callout, getting as far as trying
3816 RCPT means that the domain itself is ok).
3818 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3821 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3822 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3823 session, the daemon ignores ECONNECT errors and logs others; it now
3824 ignores EPIPE as well.
3826 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3827 (quoted-printable decoding).
3829 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3830 later the small subsequent patch to fix an introduced bug.
3832 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3834 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3836 PH/08 An error is now given if message_size_limit is specified negative.
3838 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3839 to be given (somewhat) arbitrary names.
3841 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3844 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3845 miscellaneous code fixes
3847 PH/10 Added the log_reject_target ACL modifier to specify where to log
3850 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3851 hostname. This is wrong, because it relates to the incoming message (and
3852 probably the interface on which it is arriving) and not to the outgoing
3853 callout (which could be using a different interface). This has been
3854 changed to use the value of the helo_data option from the smtp transport
3855 instead - this is what is used when a message is actually being sent. If
3856 there is no remote transport (possible with a router that sets up host
3857 addresses), $smtp_active_hostname is used.
3859 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3860 tweaks were necessary in order to get it to work (see also 21 below):
3861 (a) The code assumed that strncpy() returns a negative number on buffer
3862 overflow, which isn't the case. Replaced with Exim's string_format()
3864 (b) There were several signed/unsigned issues. I just did the minimum
3865 hacking in of casts. There is scope for a larger refactoring.
3866 (c) The code used strcasecmp() which is not a standard C function.
3867 Replaced with Exim's strcmpic() function.
3868 (d) The code set only $1; it now sets $auth1 as well.
3869 (e) A simple test gave the error "authentication client didn't specify
3870 service in request". It would seem that Dovecot has changed its
3871 interface. Fortunately there's a specification; I followed it and
3872 changed what the client sends and it appears to be working now.
3874 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3877 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3878 address A is aliased to B and C, where B exists and C does not. Without
3879 -v the output is "A verified" because verification stops after a
3880 successful redirection if more than one address is generated. However,
3881 with -v the child addresses are also verified. Exim was outputting "A
3882 failed to verify" and then showing the successful verification for C,
3883 with its parentage. It now outputs "B failed to verify", showing B's
3884 parentage before showing the successful verification of C.
3886 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3887 look up a TXT record in a specific list after matching in a combined
3890 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3891 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3892 they consult the DNS. I had assumed they would set it the way they
3893 wanted; and indeed my experiments on Linux seem to show that in some
3894 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3895 To be on the safe side, however, I have now made the interface to
3896 host_find_byname() similar to host_find_bydns(), with an argument
3897 containing the DNS resolver options. The host_find_byname() function now
3898 sets these options at its start, just as host_find_bydns() does. The smtp
3899 transport options dns_qualify_single and dns_search_parents are passed to
3900 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3901 of host_find_byname() use the default settings of RES_DEFNAMES
3902 (qualify_single) but not RES_DNSRCH (search_parents).
3904 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3905 spool_read_header() do less string testing, by means of a preliminary
3906 switch on the second character of optional "-foo" lines. (This is
3907 overdue, caused by the large number of possibilities that now exist.
3908 Originally there were few.) While I was there, I also converted the
3909 str(n)cmp tests so they don't re-test the leading "-" and the first
3910 character, in the hope this might squeeze out yet more improvement.
3912 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3913 flag allowing group syntax was set by the header_syntax check but not
3914 turned off, possible causing trouble later; (2) The flag was not being
3915 set at all for the header_verify test, causing "group"-style headers to
3916 be rejected. I have now set it in this case, and also caused header_
3917 verify to ignore an empty address taken from a group. While doing this, I
3918 came across some other cases where the code for allowing group syntax
3919 while scanning a header line wasn't quite right (mostly, not resetting
3920 the flag correctly in the right place). These bugs could have caused
3921 trouble for malformed header lines. I hope it is now all correct.
3923 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3924 with the "reply" argument non-NULL. The code, however (which originally
3925 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3926 but it didn't always do it. This confused somebody who was copying the
3927 code for some other use. I have removed all the tests.
3929 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3930 feature that was used to support insecure browsers during the U.S. crypto
3931 embargo. It requires special client support, and Exim is probably the
3932 only MTA that supported it -- and would never use it because real RSA is
3933 always available. This code has been removed, because it had the bad
3934 effect of slowing Exim down by computing (never used) parameters for the
3935 RSA_EXPORT functionality.
3937 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3938 authenticator to fail if there's a tab character in the incoming data
3939 (there should never be unless someone is messing about, as it's supposed
3940 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3941 if the connection is using TLS or if the remote IP is the same as the
3942 local IP, and the "valid-client-cert option" if a client certificate has
3945 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3946 authenticators. This can be used for authorization after authentication
3947 succeeds. (In the case of plaintext, it servers for both authentication
3950 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3951 if any retry times were supplied.
3953 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3954 connection, where there is no client IP address to check. In this
3955 situation, the verify now always succeeds.
3957 PH/25 Applied John Jetmore's -Mset patch.
3959 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3961 PH/27 In a string expansion for a processed (not raw) header when multiple
3962 headers of the same name were present, leading whitespace was being
3963 removed from all of them, but trailing whitespace was being removed only
3964 from the last one. Now trailing whitespace is removed from each header
3965 before concatenation. Completely empty headers in a concatenation (as
3966 before) are ignored.
3968 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3969 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3971 PH/29 [Removed. This was a change that I later backed out, and forgot to
3972 correct the ChangeLog entry (that I had efficiently created) before
3973 committing the later change.]
3975 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3976 address errors (4xx response to RCPT) over the same connection as other
3977 messages routed to the same hosts. Such deliveries are always "forced",
3978 so retry times are not inspected. This resulted in far too many retries
3979 for the affected addresses. The effect occurred only when there were more
3980 hosts than the hosts_max_try setting in the smtp transport when it had
3981 the 4xx errors. Those hosts that it had tried were not added to the list
3982 of hosts for which the message was waiting, so if all were tried, there
3983 was no problem. Two fixes have been applied:
3985 (i) If there are any address or message errors in an SMTP delivery, none
3986 of the hosts (tried or untried) are now added to the list of hosts
3987 for which the message is waiting, so the message should not be a
3988 candidate for sending over the same connection that was used for a
3989 successful delivery of some other message. This seems entirely
3990 reasonable: after all the message is NOT "waiting for some host".
3991 This is so "obvious" that I'm not sure why it wasn't done
3992 previously. Hope I haven't missed anything, but it can't do any
3993 harm, as the worst effect is to miss an optimization.
3995 (ii) If, despite (i), such a delivery is accidentally attempted, the
3996 routing retry time is respected, so at least it doesn't keep
3997 hammering the server.
3999 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4000 in ${readsocket because some servers need this prod.
4002 PH/32 Added some extra debug output when updating a wait-xxx database.
4004 PH/33 The hint "could be header name not terminated by colon", which has been
4005 given for certain expansion errors for a long time, was not being given
4006 for the ${if def:h_colon_omitted{... case.
4008 PH/34 The spec says: "With one important exception, whenever a domain list is
4009 being scanned, $domain contains the subject domain." There was at least
4010 one case where this was not true.
4012 PH/35 The error "getsockname() failed: connection reset by peer" was being
4013 written to the panic log as well as the main log, but it isn't really
4014 panic-worthy as it just means the connection died rather early on. I have
4015 removed the panic log writing for the ECONNRESET error when getsockname()
4018 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4019 runs only) independently of the message's sender address. This meant
4020 that, if the 4xx error was in fact related to the sender, a different
4021 message to the same recipient with a different sender could confuse
4022 things. In particular, this can happen when sending to a greylisting
4023 server, but other circumstances could also provoke similar problems.
4024 I have changed the default so that the retry time for these errors is now
4025 based a combination of the sender and recipient addresses. This change
4026 can be overridden by setting address_retry_include_sender=false in the
4029 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4030 remote server are returned as part of bounce messages. This was not
4031 happening for LMTP over a pipe (the lmtp transport), but now it is the
4032 same for both kinds of LMTP.
4034 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4035 in header lines that were in fact CNAMEs. This is no longer the case.
4037 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4038 and queue runs started by the daemon processed all messages. This has
4039 been fixed so that -R and -S can now usefully be given with -q<time>.
4041 PH/40 Import PCRE release 6.7 (fixes some bugs).
4043 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4045 PH/42 Give an error if -q is specified more than once.
4047 PH/43 Renamed the variables $interface_address and $interface_port as
4048 $received_ip_address and $received_port, to make it clear that these
4049 values apply to message reception, and not to the outgoing interface when
4050 a message is delivered. (The old names remain recognized, of course.)
4052 PH/44 There was no timeout on the connect() call when using a Unix domain
4053 socket in the ${readsocket expansion. There now is.
4055 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4056 be meaningful with "accept".
4058 SC/01 Eximstats V1.43
4059 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4061 SC/02 Eximstats V1.44
4062 Use a glob alias rather than an array ref in the generated
4063 parser. This improves both readability and performance.
4065 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4066 Collect SpamAssassin and rejection statistics.
4067 Don't display local sender or destination tables unless
4068 there is data to show.
4069 Added average volumes into the top table text output.
4071 SC/04 Eximstats V1.46
4072 Collect data on the number of addresses (recipients)
4073 as well as the number of messages.
4075 SC/05 Eximstats V1.47
4076 Added 'Message too big' to the list of mail rejection
4077 reasons (thanks to Marco Gaiarin).
4079 SC/06 Eximstats V1.48
4080 Mainlog lines which have GMT offsets and are too short to
4081 have a flag are now skipped.
4083 SC/07 Eximstats V1.49 (Alain Williams)
4084 Added the -emptyok flag.
4086 SC/08 Eximstats V1.50
4087 Fixes for obtaining the IP address from reject messages.
4089 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4090 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4091 whitespace changes from 4.64-PH/27
4093 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4096 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4097 are found, allow negative numbers in numeric criteria)
4099 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4101 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4102 to match changes made in 4.64-PH/43
4104 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4106 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4107 "domain" is actually an IP address, and force a failure. This locks out
4108 those revolvers/nameservers that support "A-for-A" lookups, in
4109 contravention of the specifications.
4111 PH/48 When a host name was looked up from an IP address, and the subsequent
4112 forward lookup of the name timed out, the host name was left in
4113 $sender_host_name, contrary to the specification.
4115 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4116 restricted to single-key lookups, Exim was not diagnosing an error if
4117 * or *@ was used with a query-style lookup.
4119 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4121 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4122 long ago, but noone interested enough thought of it. Let's just say that
4123 the "1.1" means that there are some new functions that weren't there at
4124 some point in the past.
4126 PH/51 Error processing for expansion failure of helo_data from an smtp
4127 transport during callout processing was broken.
4129 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4130 tested/used via the -bh/-bhc/-bs options.
4132 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4133 bug, fixed in subsequent PCRE releases).
4135 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4136 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4138 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4144 SC/01 Use a glob alias rather than an array ref in eximstats generated
4145 parser. This improves both readability and performance.
4147 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4148 Don't display local sender or destination tables in eximstats unless
4149 there is data to show.
4150 Added average volumes into the eximstats top table text output.
4152 SC/03 Collect data on the number of addresses (recipients) as well
4153 as the number of messages in eximstats.
4155 TF/01 Correct an error in the documentation for the redirect router. Exim
4156 does (usually) call initgroups() when daemonizing.
4158 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4159 with consistent privilege compared to when running as a daemon.
4161 TF/03 Note in the spec that $authenticated_id is not set for local
4162 submissions from trusted users.
4164 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4165 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4167 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4168 by adding some example configuration directives to the default
4169 configuration file. A little bit of work is required to uncomment the
4170 directives and define how usernames and passwords are checked, but
4171 there is now a framework to start from.
4173 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4174 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4175 without this. I don't know how relevant this is to other LDAP libraries.
4177 PH/02 Add the verb name to the "unknown ACL verb" error.
4179 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4181 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4183 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4184 directory not expanded when it should be if an expanded home directory
4185 was set for the address (which is overridden by the transport).
4187 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4190 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4191 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4192 because it is too late at that time, and has no effect.
4194 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4195 security issue with \' (bugzilla #107). I could not use the
4196 PQescapeStringConn() function, because it needs a PGconn value as one of
4199 PH/08 When testing addresses using -bt, indicate those final addresses that
4200 are duplicates that would not cause an additional delivery. At least one
4201 person was confused, thinking that -bt output corresponded to deliveries.
4202 (Suppressing duplicates isn't a good idea as you lose the information
4203 about possibly different redirections that led to the duplicates.)
4205 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4206 systems where poll() doesn't work, in particular OS X.
4208 PH/10 Added more information to debugging output for retry time not reached.
4210 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4211 operations in malware.c.
4213 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4216 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4217 syslog_duplication set false, log lines that would normally be written
4218 both the the main log and to the reject log were not written to syslog at
4221 PH/14 In the default configuration, change the use of "message" in ACL warn
4222 statements to "add_header".
4224 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4225 not followed by a command (e.g. "seen endif").
4227 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4228 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4231 PH/17 Added extra conditions to the default value of delay_warning_condition
4235 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4236 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4237 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4240 The Auto-Submitted: and various List- headers are standardised, whereas I
4241 don't think Precedence: ever was.
4243 PH/18 Refactored debugging code in route_finduser() to show more information,
4244 in particular, the error code if getpwnam() issues one.
4246 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4247 This is apparently needed in addition to the PH/07 change above to avoid
4248 any possible encoding problems.
4250 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4251 but not after initializing Perl.
4253 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4254 output them only if debugging. By default they are written stderr,
4255 apparently, which is not desirable.
4257 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4260 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4263 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4265 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4266 authenticated or an ident call has been made. Suppress the default
4267 values for $authenticated_id and $authenticated_sender (but permit -oMai
4268 and -oMas) when testing with -bh.
4270 PH/24 Re-jigged the order of the tests in the default configuration so that the
4271 tests for valid domains and recipients precede the DNS black list and CSA
4272 tests, on the grounds that those ones are more expensive.
4274 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4275 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4276 This bug exists in every version of Exim that I still have, right back to
4279 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4280 However, an attempt to turn on case-sensitivity in a regex key by
4281 including (?-i) didn't work because the subject string was already
4282 lowercased, and the effects were non-intuitive. It turns out that a
4283 one-line patch can be used to allow (?-i) to work as expected.
4289 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4290 other effects) broke the use of negated acl sub-conditions.
4292 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4295 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4296 "Deny" causes Exim to reject the incoming connection with a 554 error.
4297 Unfortunately, if there is a major crisis, such as a disk failure,
4298 tcp-wrappers gives "deny", whereas what one would like would be some
4299 kind of temporary error. A kludge has been added to help with this.
4300 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4301 554 error is used if errno is still zero or contains ENOENT (which occurs
4302 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4305 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4307 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4308 errors. Otherwise a message that provokes a temporary error (when other
4309 messages do not) can cause a whole host to time out.
4311 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4312 addresses were routed directly to files or pipes from a redirect router.
4313 File deliveries just didn't batch; pipe deliveries might have suffered
4316 PH/06 A failure to get a lock for a hints database would erroneously always say
4317 "Failed to get write lock", even when it was really a read lock.
4319 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4320 of 0600. This has been changed to use the value of the lockfile_mode
4321 option (which defaults to 0600).
4323 PH/08 Applied small patch from the Sieve maintainer.
4325 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4326 folder from quota calculations, a direct delivery into this folder messed
4327 up the contents of the maildirsize file. This was because the regex was
4328 used only to exclude .Trash (or whatever) when the size of the mailbox
4329 was calculated. There was no check that a delivery was happening into an
4330 excluded directory. This bug has been fixed by ignoring all quota
4331 processing for deliveries into excluded directories.
4333 PH/10 Added the maildirfolder_create_regex option to appendfile.
4339 PH/01 The code for finding all the local interface addresses on a FreeBSD
4340 system running IPv6 was broken. This may well have applied to all BSD
4341 systems, as well as to others that have similar system calls. The broken
4342 code found IPv4 interfaces correctly, but gave incorrect values for the
4343 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4344 that it would not match correctly against @[] and not recognize the IPv6
4347 PH/02 The ipliteral router was not recognizing addresses of the form user@
4348 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4350 PH/03 Added disable_ipv6.
4352 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4353 decoded form, because it is most often used to construct To: headers
4354 lines in autoreplies, and the decoded form may well be syntactically
4355 invalid. However, $reply_address has leading white space removed, and all
4356 newlines turned into spaces so that the autoreply transport does not
4359 PH/05 If group was specified without a user on a router, and no group or user
4360 was specified on a transport, the group from the router was ignored.
4362 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4363 for visible compile-time settings that can be used to change these
4364 numbers, for those that want even more. Backwards compatibility with old
4365 spool files has been maintained. However, going back to a previous Exim
4366 release will lost any variables that are in spool files.
4368 PH/07 Two small changes when running in the test harness: increase delay when
4369 passing a TCP/IP connection to a new process, in case the original
4370 process has to generate a bounce, and remove special handling of
4371 127.0.0.2 (sic), which is no longer necessary.
4373 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4374 be the same on different OS.
4376 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4379 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4380 whether --show-vars was specified or not
4382 JJ/02 exipick: Added support for new ACL variable spool format introduced
4385 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4386 syntactically invalid From: or Reply-to: line, and a filter used this to
4387 generate an autoreply, and therefore failed to obtain an address for the
4388 autoreply, Exim could try to deliver to a non-existent relative file
4389 name, causing unrelated and misleading errors. What now happens is that
4390 it logs this as a hard delivery error, but does not attempt to create a
4393 PH/11 The exinext utility has a -C option for testing purposes, but although
4394 the given file was scanned by exinext itself; it wasn't being passed on
4395 when Exim was called.
4397 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4398 an end-of-file indication when reading a command response.
4400 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4401 compiled. In many other places in Exim, IPv6 addresses are always
4402 recognized, so I have changed this. It also means that IPv4 domain
4403 literals of the form [IPV4:n.n.n.n] are now always recognized.
4405 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4406 used if the router is not running as root, for example, when verifying at
4407 ACL time, or when using -bh. The debugging output from this situation was
4408 non-existent - all you got was a failure to exec. I have made two
4411 (a) Failures to set uid/gid, the current directory, or a process leader
4412 in a subprocess such as that created by queryprogram now generate
4413 suitable debugging output when -d is set.
4415 (b) The queryprogram router detects when it is not running as root,
4416 outputs suitable debugging information if -d is set, and then runs
4417 the subprocess without attempting to change uid/gid.
4419 PH/15 Minor change to Makefile for building test_host (undocumented testing
4422 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4423 additional section of a DNS packet that returns MX or SRV records.
4424 Instead, it always explicitly searches for A/AAAA records. This avoids
4425 major problems that occur when a DNS server includes only records of one
4426 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4427 fixed another bug: if SRV records were looked up and the corresponding
4428 address records were *not* found in the additional section, the port
4429 values from the SRV records were lost.
4431 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4432 using the correct key (the original address) when searching the retry
4433 rules in order to find which one to use for generating the retry hint.
4435 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4436 adding the default one. Similarly, if it contains a Reply-To: header, the
4437 errors_reply_to option, if set, is not used.
4439 PH/19 When calculating a retry time, Exim used to measure the "time since
4440 failure" by looking at the "first failed" field in the retry record. Now
4441 it does not use this if it is later than than the arrival time of the
4442 message. Instead it uses the arrival time. This makes for better
4443 behaviour in cases where some deliveries succeed, thus re-setting the
4444 "first failed" field. An example is a quota failure for a huge message
4445 when small messages continue to be delivered. Without this change, the
4446 "time since failure" will always be short, possible causing more frequent
4447 delivery attempts for the huge message than are intended.
4448 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4450 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4451 $1, $2, $3) because the numerical variables can be reset during some
4452 expansion items (e.g. "match"), thereby losing the authentication data.
4454 PH/21 Make -bV show the size of off_t variables so that the test suite can
4455 decide whether to run tests for quotas > 2G.
4457 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4458 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4459 filecount value is greater than 2G or if a quota value is greater than 2G
4460 on a system where the size of off_t is not greater than 4, a panic error
4463 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4464 never match. The debug and -bh output now contains an explicit error
4465 message indicating a malformed IPv4 address or mask.
4467 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4468 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4469 PH/23 above applies.
4471 PH/25 Do not write to syslog when running in the test harness. The only
4472 occasion when this arises is a failure to open the main or panic logs
4473 (for which there is an explicit test).
4475 PH/26 Added the /no_tell option to "control=freeze".
4477 PH/27 If a host name lookup failed very early in a connection, for example, if
4478 the IP address matched host_lookup and the reverse lookup yielded a name
4479 that did not have a forward lookup, an error message of the form "no IP
4480 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4481 could be logged. Now it outputs the IP address instead of "NULL".
4483 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4484 allows the sender and the authenticated sender to be set when
4485 submitting a message from within Exim. Since child_open_exim() is
4486 documented for local_scan(), the new function should be too.
4488 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4489 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4490 results in an empty string is now treated as unset.
4492 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4494 PH/31 Added sender_verify_fail logging option.
4496 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4497 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4498 tidied the source and removed it altogether.
4500 PH/33 When a queue run was abandoned because the load average was too high, a
4501 log line was always written; now it is written only if the queue_run log
4502 selector is set. In addition, the log line for abandonment now contains
4503 information about the queue run such as the pid. This is always present
4504 in "start" and "stop" lines but was omitted from the "abandon" line.
4506 PH/34 Omit spaces between a header name and the colon in the error message that
4507 is given when verify = headers_syntax fails (if there are lots of them,
4508 the message gets confusing).
4510 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4511 names, as there are now some PTR records that contain slashes. This check
4512 is only to protect against broken name servers that fall over on strange
4513 characters, so the fact that it applies to all lookups doesn't matter.
4515 PH/36 Now that the new test suite is complete, we can remove some of the
4516 special code in Exim that was needed for the old test suite. For example,
4517 sorting DNS records because real resolvers return them in an arbitrary
4518 order. The new test suite's fake resolver always returns records in the
4521 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4522 bounces) except when queue_only is set, to avoid logging races between
4523 the different processes.
4525 PH/38 Panic-die if .include specifies a non-absolute path.
4527 PH/39 A tweak to the "H" retry rule from its user.
4529 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4530 a label. They prevented compilation on older perls.
4532 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4533 a warning to be raised on newish perls.
4535 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4536 on queue. Changes to match documented behaviour of showing count of
4537 messages matching specified criteria.
4539 PH/40 Changed the default ident timeout from 30s to 5s.
4541 PH/41 Added support for the use of login_cap features, on those BSD systems
4542 that have them, for controlling the resources used by pipe deliveries.
4544 PH/42 The content-scanning code uses fopen() to create files in which to put
4545 message data. Previously it was not paying any attention to the mode of
4546 the files. Exim runs with umask(0) because the rest of the code creates
4547 files with open(), and sets the required mode explicitly. Thus, these
4548 files were ending up world-writeable. This was not a big issue, because,
4549 being within the spool directory, they were not world-accessible. I have
4550 created a function called modefopen, which takes an additional mode
4551 argument. It sets umask(777), creates the file, chmods it to the required
4552 mode, then resets the umask. All the relevant calls to fopen() in the
4553 content scanning code have been changed to use this function.
4555 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4556 to 24 hours. This avoids potential overflow problems when processing G
4557 and H retry rules. I suspect nobody ever tinkers with this value.
4559 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4561 PH/45 When the plaintext authenticator is running as a client, the server's
4562 challenges are checked to ensure they are valid base64 strings. By
4563 default, the authentication attempt is cancelled if an invalid string is
4564 received. Setting client_ignore_invalid_base64 true ignores these errors.
4565 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4566 they are received. Thus, the responses can be made to depend on the
4567 challenges. If an invalid string is ignored, an empty string is placed in
4570 PH/46 Messages that are created by the autoreply transport now contains a
4571 References: header, in accordance with RFCs 2822 and 3834.
4573 PH/47 Added authenticated_sender_force to the smtp transport.
4575 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4577 PH/49 Installed latest patch from the Sieve maintainer.
4579 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4580 was also set, the appendfile transport was unnecessarily scanning a
4581 directory of message files (e.g. for maildir delivery) to find the count
4582 of files (along with the size), even though it did not need this
4583 information. It now does the scan only if it needs to find either the
4584 size of the count of files.
4586 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4588 PH/52 Two bugs concerned with error handling when the smtp transport is
4591 (i) Exim was not creating retry information for temporary errors given
4592 for individual recipients after the DATA command when the smtp transport
4593 was used in LMTP mode. This meant that they could be retried too
4594 frequently, and not timed out correctly.
4596 (ii) Exim was setting the flag that allows error details to be returned
4597 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4598 recipients that were returned after the DATA command.
4600 PH/53 This is related to PH/52, but is more general: for any failing address,
4601 when detailed error information was permitted to be returned to the
4602 sender, but the error was temporary, then after the final timeout, only
4603 "retry timeout exceeded" was returned. Now it returns the full error as
4604 well as "retry timeout exceeded".
4606 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4607 do this, and (what is worse) MTAs that accept it.
4609 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4610 will now be deprecated.
4612 PH/56 New os.c-cygwin from the Cygwin maintainer.
4614 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4615 formats (previously only available in exim formats via -bpr, -bpru,
4616 and -bpra. Now also available in native and exiqgrep formats)
4618 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4619 with very large, slow to parse queues
4621 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4623 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4625 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4626 responses to authentication challenges, though it was showing the
4627 challenges; (ii) I've removed the CR characters from the debug output for
4630 PH/58 Allow for the insertion of a newline as well as a space when a string
4631 is turned into more than one encoded-word during RFC 2047 encoding. The
4632 Sieve code now uses this.
4634 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4635 data_4xx, lost_connection, tls_required.
4637 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4638 message was being sent as an SMTP response.
4640 PH/61 Add -l and -k options to exicyclog.
4642 PH/62 When verifying, if an address was redirected to one new address, so that
4643 verification continued, and the new address failed or deferred after
4644 having set something in $address_data, the value of $address_data was not
4645 passed back to the ACL. This was different to the case when no
4646 redirection occurred. The value is now passed back in both cases.
4648 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4649 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4650 use login_cap.h, so on its own it isn't the distinguishing feature. The
4651 new name refers directly to the setclassresources() function.
4653 PH/65 Added configuration files for NetBSD3.
4655 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4657 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4658 is preferred over IPv4.
4660 PH/68 The bounce_return_message and bounce_return_body options were not being
4661 honoured for bounces generated during the reception of non-SMTP messages.
4662 In particular, this applied to messages rejected by the ACL. This bug has
4663 been fixed. However, if bounce_return_message is true and bounce_return_
4664 body is false, the headers that are returned for a non-SMTP message
4665 include only those that have been read before the error was detected.
4666 (In the case of an ACL rejection, they have all been read.)
4668 PH/69 The HTML version of the specification is now built in a directory called
4669 spec_html instead of spec.html, because the latter looks like a path with
4670 a MIME-type, and this confuses some software.
4672 PH/70 Catch two compiler warnings in sieve.c.
4674 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4675 function verify_get_ident() calls ip_connect() to connect a socket, but
4676 if the "connect()" function timed out, ip_connect() used to close the
4677 socket. However, verify_get_ident() also closes the socket later, and in
4678 between Exim writes to the log, which may get opened at this point. When
4679 the socket was closed in ip_connect(), the log could get the same file
4680 descriptor number as the socket. This naturally causes chaos. The fix is
4681 not to close the socket in ip_connect(); the socket should be closed by
4682 the function that creates it. There was only one place in the code where
4683 this was missing, in the iplookup router, which I don't think anybody now
4684 uses, but I've fixed it anyway.
4686 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4687 well as to direct DNS lookups. Otherwise the handling of names in host
4688 lists is inconsistent and therefore confusing.
4694 PH/01 Two changes to the default runtime configuration:
4696 (1) Move the checks for relay_from_hosts and authenticated clients from
4697 after to before the (commented out) DNS black list checks.
4699 (2) Add control=submission to the relay_from_hosts and authenticated
4700 clients checks, on the grounds that messages accepted by these
4701 statements are most likely to be submissions.
4703 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4705 (1) Generate an error if the third argument for the ${prvs expansion is
4708 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4711 (3) Reset the variables that are obtained from the first argument of
4712 ${prvscheck and used in the second argument before leaving the code,
4713 because their memory is reclaimed, so using them afterwards may do
4716 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4717 one (it's much easier than Tom thought :-).
4719 (5) Because of (4), we can now allow for the use of $prvscheck_result
4720 inside the third argument.
4722 PH/03 For some reason, the default setting of PATH when running a command from
4723 a pipe transport was just "/usr/bin". I have changed it to
4726 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4727 anything to be listed in the output from -bV.
4729 PH/05 When a filter generated an autoreply, the entire To: header line was
4730 quoted in the delivery log line, like this:
4732 => >A.N.Other <ano@some.domain> <original@ddress> ...
4734 This has been changed so that it extracts the operative address. There
4735 may be more than one such address. If so, they are comma-separated, like
4738 => >ano@some.domain,ona@other.domain <original@ddress> ...
4740 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4741 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4742 not being looked up in the rDNS to get a host name, Exim was showing the
4743 IP address twice in Received: lines, even though the IP addresses were
4744 identical. For example:
4746 Received: from [1.2.3.4] (helo=[1.2.3.4])
4748 However, if the real host name was known, it was omitting the HELO data
4749 if it matched the actual IP address. This has been tidied up so that it
4750 doesn't show the same IP address twice.
4752 PH/07 When both +timestamp and +memory debugging was on, the value given by
4753 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4754 called by the debug printing, thereby overwriting the timestamp buffer.
4755 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4757 PH/08 When the original message was included in an autoreply transport, it
4758 always said "this is a copy of the message, including all the headers",
4759 even if body_only or headers_only was set. It now gives an appropriate
4762 PH/09 Applied a patch from the Sieve maintainer which:
4764 o fixes some comments
4765 o adds the (disabled) notify extension core
4766 o adds some debug output for the result of if/elsif tests
4767 o points to the current vacation draft in the documentation
4768 and documents the missing references header update
4772 o fixes a bug in processing the envelope test (when testing
4773 multiple envelope elements, the last element determined the
4776 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4777 Electronic Mail") by including:
4779 Auto-submitted: auto-generated
4781 in the messages that it generates (bounce messages and others, such as
4782 warnings). In the case of bounce messages for non-SMTP messages, there was
4783 also a typo: it was using "Auto_submitted" (underscore instead of
4784 hyphen). Since every message generated by Exim is necessarily in response
4785 to another message, thes have all been changed to:
4787 Auto-Submitted: auto-replied
4789 in accordance with these statements in the RFC:
4791 The auto-replied keyword:
4793 - SHOULD be used on messages sent in direct response to another
4794 message by an automatic process,
4796 - MUST NOT be used on manually-generated messages,
4798 - MAY be used on Delivery Status Notifications (DSNs) and Message
4799 Disposition Notifications (MDNs),
4801 - MUST NOT be used on messages generated by automatic or periodic
4802 processes, except for messages which are automatic responses to
4805 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4806 to the default Received: header definition.
4808 PH/12 Added log selector acl_warn_skipped (default on).
4810 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4811 variables because (a) they are in the wrong storage pool and (b) even if
4812 they were copied, it wouldn't work properly because of the caching.
4814 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4815 checking when decoding. Apparently there are clients that generate
4816 overlong encoded strings. Why am I not surprised?
4818 PH/15 If the first argument of "${if match_address" was not empty, but did not
4819 contain an "@" character, Exim crashed. Now it writes a panic log message
4820 and treats the condition as false.
4822 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4824 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4825 extension "envelope-auth". The code is finished and in agreement with
4826 other implementations, but there is no documentation so far and in fact,
4827 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4828 not changing the active code.
4830 Print executed "if" and "elsif" statements when debugging is used. This
4831 helps a great deal to understand what a filter does.
4833 Document more things not specified clearly in RFC3028. I had all this
4834 sorted out, when out of a sudden new issues came to my mind. Oops."
4836 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4839 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4840 canonical form (as documented). However, after a host name lookup from
4841 the IP address, check_host() was doing a simple string comparison with
4842 addresses acquired from the DNS when checking that the found name did
4843 have the original IP as one of its addresses. Since any found IPv6
4844 addresses are likely to be in abbreviated form, the comparison could
4845 fail. Luckily, there already exists a function for doing the comparison
4846 by converting both addresses to binary, so now that is used instead of
4847 the text comparison.
4849 PH/20 There was another similar case to PH/19, when a complete host name was
4850 given in a host list; looking up its IP address could give an abbreviated
4851 form, whereas the current host's name might or might not be abbreviated.
4852 The same fix has been applied.
4858 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4859 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4862 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4863 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4865 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4867 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4868 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4869 Also, for "Auto-Submitted", treat anything other than "no" as
4870 non-personal, in accordance with RFC 3834. (Previously it treated
4871 anything starting "auto-" as non-personal.)
4873 TF/01 The control=submission/name=... option had a problem with syntax
4874 errors if the name included a slash character. The /name= option
4875 now slurps the rest of the string, so it can include any characters
4876 but it must come last in the list of options (after /sender_retain
4879 PH/05 Some modifications to the interface to the fake nameserver for the new
4887 TK/01 Added the "success_on_redirect" address verification option. See
4888 NewStuff for rationale and an example.
4890 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4892 PH/02 Patch to exigrep to allow it to work on syslog lines.
4894 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4895 fread() to skip over the body file's header line, because in Cygwin the
4896 header line is locked and is inaccessible.
4898 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4899 co-exist for some time) to make it clear that it is the Exim ID that is
4900 referenced, not the Message-ID: header line.
4902 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4903 string_format() function, because snprintf() does not exist on all
4906 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4907 ${stat: expansion item.
4909 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4910 protocol synchronization error", to keep the pedants happy.
4912 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4913 well as for IRIX systems, when gcc is being used. See the host.c source
4916 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4918 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4921 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4922 transport and to the smtp transport in LMTP mode.
4924 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4926 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4927 run a filter in a subprocess. This could lead to confusion in subsequent
4928 lookups in the parent process. There should also be a search_tidyup() at
4929 the end of the subprocess.
4931 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4932 only if the host matched helo_try_verify_hosts, which caused the
4933 verification to occur when the EHLO/HELO command was issued. The ACL just
4934 tested the remembered result. Now, if a previous verification attempt has
4935 not happened, "verify = helo" does it there and then.
4937 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4939 TK/03 Fix log output including CR from clamd.
4941 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4942 exist provoked a memory error which could cause a segfault.
4944 PH/15 Installed PCRE 6.2
4946 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4948 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4949 of the problem. Specifically, suggested +O2 rather than +O1 for the
4952 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4954 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4955 record, the port was not correctly logged when the outgoing_port log
4956 selector was set (it logged the transort's default port).
4958 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4959 fallback_hosts, and "hosts" in the smtp transport.
4961 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4962 host errors such as "Connection refused".
4964 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4965 authentication with radiusclient 0.4.9:
4967 - Error returned from rc_read_config was caught wrongly
4968 - Username/password not passed on to radius server due to wrong length.
4970 The presumption is that some radiusclient API changes for 4.51/PH/17
4971 were not taken care of correctly. The code is still untested by me (my
4972 Linux distribution still has 0.3.2 of radiusclient), but it was
4973 contributed by a Radius user.
4975 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4976 expanding the "port" option of the smtp transport.
4978 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4979 while reading a MIME header. Thanks to Tom Hughes for a patch.
4981 PH/24 Include config.h inside local_scan.h so that configuration settings are
4984 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4985 that in an ACL for RCPT (for example), you can examine exactly what was
4988 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4989 commands, but it was not correctly comparing the address with the actual
4990 client host address. Thus, it would show the EHLO address in Received:
4991 header lines when this was not necessary.
4993 PH/27 Added the % operator to ${eval:}.
4995 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4996 it should be ignoring failures (because with -C, for example, it has lost
4997 privilege). It wasn't ignoring creation failures other than "already
5000 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5003 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5004 because an input line was too long, either on its own, or by virtue of
5005 too many continuations, the temporary file was not being removed, and the
5006 return code was incorrect.
5008 PH/31 Missing "BOOL" in function definition in filtertest.c.
5010 PH/32 Applied Sieve patches from the maintainer.
5012 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5014 PH/33 Added "verify = not_blind".
5016 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5017 Local/Makefile (with some defaults set). These are used in built scripts
5018 such as exicyclog, but they have never been used in the exim_install
5019 script (though there are many overriding facilities there). I have
5020 arranged that the exim_install script now takes note of these two
5023 PH/35 Installed configuration files for Dragonfly.
5025 PH/36 When a locally submitted message by a trusted user did not contain a
5026 From: header, and the sender address was obtained from -f or from an SMTP
5027 MAIL command, and the trusted user did not use -F to supply a sender
5028 name, $originator_name was incorrectly used when constructing a From:
5029 header. Furthermore, $originator_name was used for submission mode
5030 messages from external hosts without From: headers in a similar way,
5031 which is clearly wrong.
5033 PH/37 Added control=suppress_local_fixups.
5035 PH/38 When log_selector = +received_sender was set, and the addition of the
5036 sender made the log line's construction buffer exactly full, or one byte
5037 less than full, an overflow happened when the terminating "\n" was
5040 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5041 when the result of a list match is failure because a DNS lookup failed.
5043 PH/40 RM_COMMAND is now used in the building process.
5045 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5046 the "build-* directories that it finds.
5048 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5049 address in a domain literal was a prefix of an interface address.
5051 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5052 when verifying a sender address, unless rewrite_headers is false.
5054 PH/44 Wrote a long comment about why errors_to addresses are verified as
5055 recipients, not senders.
5057 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5058 the ratelimit ACL was added.
5060 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5062 PH/46 Added extra information about PostgreSQL errors to the error string.
5064 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5065 avoiding the need to install special zones in a real server. This is
5066 backwards compatible; if it can't find the fake resolver, it drops back.
5067 Thus, both old and new test suites can be run.
5069 TF/02 Added util/ratelimit.pl
5071 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5072 clock is set back in time.
5074 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5075 Candler <B.Candler@pobox.com>.
5077 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5078 OK for addresses that are the result of redirections.
5080 PH/48 A number of further additions for the benefit of the new test suite,
5081 including a fake gethostbyname() that interfaces to the fake DNS resolver
5084 TF/06 The fix for widen_domains has also been applied to qualify_single and
5085 search_parents which are the other dnslookup options that can cause
5088 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5091 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5093 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5094 service" if it encountered an MX record with an empty target hostname.
5095 The message is now "an MX or SRV record indicated no SMTP service".
5097 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5098 if the DNS of the sending site is misconfigured. This is quite a
5099 common situation. This change restores the behaviour of treating a
5100 helo verification defer as a failure.
5102 PH/51 If self=fail was set on a router, the bounce message did not include the
5103 actual error message.
5109 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5111 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5112 command itself ended in error, the underlying message about the transport
5113 filter timeout was being overwritten with the pipe command error. Now the
5114 underlying error message should be appended to the second error message.
5116 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5118 PH/02 Reduce the amount of output that "make" produces by default. Full output
5119 can still be requested.
5121 PH/03 The warning log line about a condition test deferring for a "warn" verb
5122 was being output only once per connection, rather than after each
5123 occurrence (because it was using the same function as for successful
5124 "warn" verbs). This seems wrong, so I have changed it.
5126 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5127 it should not have, which might have caused a crash in the right
5128 circumstances, but probably never did.
5130 PH/04 Installed a modified version of Tony Finch's patch to make submission
5131 mode fix the return path as well as the Sender: header line, and to
5132 add a /name= option so that you can make the user's friendly name appear
5135 TF/03 Added the control = fakedefer ACL modifier.
5137 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5138 Mark Lowes for thorough testing.
5140 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5142 TK/03 Merged latest SRS patch from Miles Wilton.
5144 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5145 with the definition in sysexits.h (which is #included earlier).
5146 Fortunately, Exim does not actually use EX_OK. The code used to try to
5147 preserve the sysexits.h value, by assuming that macro definitions were
5148 scanned for macro replacements. I have been disabused of this notion,
5149 so now the code just undefines EX_OK before #including unistd.h.
5151 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5152 in the smtp transport. When a block could not be written in a single
5153 write() function, the timeout was being re-applied to each part-write.
5154 This seems wrong - if the receiver was accepting one byte at a time it
5155 would take for ever. The timeout is now adjusted when this happens. It
5156 doesn't have to be particularly precise.
5158 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5159 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5161 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5162 without a domain if the check to <postmaster@domain> fails.
5164 SC/01 Eximstats: added -xls and the ability to specify output files
5165 (patch written by Frank Heydlauf).
5167 SC/02 Eximstats: use FileHandles for outputting results.
5169 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5171 SC/04 Eximstats: fixed display of large numbers with -nvr option
5173 SC/05 Eximstats: fixed merging of reports with empty tables.
5175 SC/06 Eximstats: added the -include_original_destination flag
5177 SC/07 Eximstats: removed tabs and trailing whitespace.
5179 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5181 TK/06 MBOX spool code: Add real "From " MBOX separator line
5182 so the .eml file is really in mbox format (even though
5183 most programs do not really care). Patch from Alex Miller.
5185 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5186 The latter is generated from $received_to and is only set if the
5187 message has one envelope recipient. SA can use these headers,
5188 obviously out-of-the-box. Patch from Alex Miller.
5190 PH/08 The ${def test on a variable was returning false if the variable's
5191 value was "0", contrary to what the specification has always said!
5192 The result should be true unless the variable is empty.
5194 PH/09 The syntax error of a character other than { following "${if
5195 def:variable_name" (after optional whitespace) was not being diagnosed.
5196 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5197 accidental colon was present, for example, could give incorrect results.
5199 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5200 result is used (not including appendfile, where other changes are about
5203 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5204 This involved changing a lot of size variables from int to off_t. It
5205 should work with maildirs and everything.
5207 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5208 spamd dying while we are connected to it.
5210 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5213 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5214 function for BDB 4.3.
5216 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5218 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5219 the sources for PCRE have been re-arranged and more files are now
5222 PH/15 The code I had for printing potentially long long variables in PH/11
5223 above was not the best (it lost precision). The length of off_t variables
5224 is now inspected at build time, and an appropriate printing format (%ld
5225 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5226 to be "long long int" or "long int". This is needed for the internal
5227 formatting function string_vformat().
5229 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5230 the configuration file to be ":syslog", then the script "guesses" where
5231 the logs files are, rather than using the compiled in default. In our
5232 case the guess is not the same as the compiled default, so the script
5233 suddenly stopped working when I started to use syslog. The patch checks
5234 to see if log_file_path is "". If so, it attempts to read it from exim
5235 with no configuration file to get the compiled in version, before it
5236 falls back to the previous guessing code."
5238 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5239 implementing BATV in an Exim configuration. See NewStuff for the gory
5242 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5243 Makefile that are specific to HP-UX.
5245 PH/18 If the "use_postmaster" option was set for a recipient callout together
5246 with the "random" option, the postmaster address was used as the MAIL
5247 FROM address for the random test, but not for the subsequent recipient
5248 test. It is now used for both.
5250 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5251 patch removes a few documentation additions to RFC 3028, because the
5252 latest draft now contains them. It adds the new en;ascii-case comparator
5253 and a new error check for 8bit text in MIME parts. Comparator and
5254 require names are now matched exactly. I enabled the subaddress
5255 extension, but it is not well tested yet (read: it works for me)."
5257 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5258 rework some of the code of TK/09 above to avoid the hardwired use of
5259 "%lld" and "long long". Replaced the call to snprintf() with a call to
5262 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5263 records point to non-existent hosts", "retry timeout exceeded", and
5264 "retry time not reached for any host after a long failure period".
5266 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5267 experimental DomainKeys support:
5269 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5270 (2) On an error such as an illegally used "control", the wrong name for
5271 the control was given.
5273 These problems did NOT occur unless DomainKeys support was compiled.
5275 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5277 PH/24 Added ${if match_ip condition.
5279 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5280 databases so that it will be absolutely obvious if a crash occurs in the
5281 DB library. This is a regular occurrence (often caused by mis-matched
5284 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5285 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5286 picked up on a user's system that detects such things. There doesn't seem
5287 to be a gcc warning option for this - only an attribute that has to be
5288 put on the function's prototype. It seems that in Fedora Core 4 they have
5289 set this on a number of new functions. No doubt there will be more in due
5292 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5293 specify a transport. However, if an address that was verified by such a
5294 router was the subject of a callout, Exim crashed because it tried to
5295 read the rcpt_include_affixes from the non-existent transport. Now it
5296 just assumes that the setting of that option is false. This bug was
5297 introduced by 4.51/PH/31.
5299 PH/28 Changed -d+all to exclude +memory, because that information is very
5300 rarely of interest, but it makes the output a lot bigger. People tend to
5301 do -d+all out of habit.
5303 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5304 code in os-type was giving problems when libc.so lives in lib64, like on
5307 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5308 aren't the modern standard, and it seems that some systems' include files
5309 don't always have them. Exim was already checking for some of the newer
5310 ones like T_AAAA, and defining it itself. I've added checks for all the
5311 record types that Exim uses.
5313 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5314 not automatically generating a new one, as it is supposed to. This
5315 prevented TLS from working. If the file did exist, but contained invalid
5316 data, a new version was generated, as expected. It was only the case of a
5317 non-existent file that was broken.
5319 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5320 with a change in libdomainkeys > 0.64.
5322 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5323 from DNS. If the selector record carries the flag, it now has
5324 precedence over the domain-wide flag.
5326 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5328 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5329 the use of a port name that isn't defined in /etc/services) occurred, the
5330 message was deferred as in a normal delivery, and thus remained on the
5331 spool, instead of being failed because of the mua_wrapper setting. This
5332 is now fixed, and I tidied up some of the mua_wrapper messages at the
5335 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5336 the messages in a hash of arrays rather than using individual hashes.
5337 This is a bit cleaner and results in dramatic memory savings, albeit
5338 at a slight CPU cost.
5340 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5341 as requested by Marc Sherman.
5343 SC/10 Eximstats: added histograms for user specified patterns as requested
5346 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5348 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5349 fopen() in the content-scanning modules that did not already have it.
5355 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5356 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5358 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5360 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5362 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5363 to test Sieve filters that use "vacation".
5365 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5366 that changes the way the GnuTLS parameters are stored in the cache file.
5367 The new format can be generated externally. For backward compatibility,
5368 if the data in the cache doesn't make sense, Exim assumes it has read an
5369 old-format file, and it generates new data and writes a new file. This
5370 means that you can't go back to an older release without removing the
5373 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5374 work if there are any delivery delays because "one_time" forces the
5375 parent to be marked "delivered", so its unseen clone is never tried
5376 again. For this reason, Exim now forbids the simultaneous setting of
5379 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5380 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5381 entry if you want to know the details. The fix, however, neglected to
5382 consider the case when local delivery batching is involved. The test for
5383 "previously delivered" was not happening when checking to see if an
5384 address could be batched with a previous (undelivered) one; under
5385 certain circumstances this could lead to multiple deliveries to the same
5388 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5389 in its include files, and this causes problems building Exim.
5391 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5392 header_syntax) but Exim was just ignoring anything given after a slash.
5393 In particular, this caused confusion with an attempt to use "verify =
5394 reverse_host_lookup/defer_ok". An error is now given when options are
5395 supplied for verify items that do not have them. (Maybe reverse_host_
5396 lookup should have a defer_ok option, but that's a different point.)
5398 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5399 defined by RFC 821) to 2048, because there were problems with some AUTH
5400 commands, and RFC 1869 says the size should be increased for extended
5401 SMTP commands that take arguments.
5403 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5406 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5407 "unknown" error; now it says that the functionality isn't in the binary.
5409 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5410 an address' error message when a string expansion fails (syntax or
5411 whatever). Otherwise the password may appear in the log. Following change
5412 PH/42 below, there is no longer a chance of it appearing in a bounce
5415 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5417 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5418 of its addresses was ever tried. (Bugzilla bug #2.)
5420 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5421 the result incorrectly in the debug output. (It correctly added a newline
5422 to what was transported.)
5424 TF/01 Added $received_time.
5426 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5427 commented out examples of how to interface to a virus scanner and to
5428 SpamAssassin. Also added commented examples of av_scanner and
5429 spamd_address settings.
5431 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5432 and controls are allowed in which ACLs. There were a couple of minor
5433 errors. Some of the entries in the conditions table (which is a table of
5434 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5435 negation of where the condition IS allowed.
5437 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5439 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5440 header file does not have a version number, so I've had to invent a new
5441 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5442 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5443 radiusclient), but it was contributed by a Radius user.
5445 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5446 files or directories, for OpenSSL.
5448 PH/19 When an Exim process that is running as root has to create an Exim log
5449 file, it does so in a subprocess that runs as exim:exim so as to get the
5450 ownership right at creation (otherwise, other Exim processes might see
5451 the file with the wrong ownership). There was no test for failure of this
5452 fork() call, which would lead to the process getting stuck as it waited
5453 for a non-existent subprocess. Forks do occasionally fail when resources
5454 run out. I reviewed all the other calls to fork(); they all seem to check
5457 PH/20 When checking for unexpected SMTP input at connect time (before writing
5458 the banner), Exim was not dealing correctly with a non-positive return
5459 from the read() function. If the client had disconnected by this time,
5460 the result was a log entry for a synchronization error with an empty
5461 string after "input=" when read() returned zero. If read() returned -1
5462 (an event I could not check), uninitialized data bytes were printed.
5463 There were reports of junk text (parts of files, etc) appearing after
5466 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5468 PH/22 Added support for macro redefinition, and (re)definition in between
5469 driver and ACL definitions.
5471 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5472 forgetting to use the resulting value; it was using the unexpanded value.
5474 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5475 hadn't been configured. The fix is from Juergen Kreileder, who
5476 understands it better than I do:
5478 "Here's what I see happening with three configured cyrus_sasl
5479 authenticators configured (plain, login, cram-md5):
5481 On startup auth_cyrus_sasl_init() gets called for each of these.
5482 This means three calls to sasl_listmech() without a specified mech_list.
5483 => SASL tests which mechs of all available mechs actually work
5484 => three warnings about OTP not working
5485 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5487 With the patch, sasl_listmech() also gets called three times. But now
5488 SASL's mech_list option is set to the server_mech specified in the the
5489 authenticator. Or in other words, the answer from sasl_listmech()
5490 gets limited to just the mech you're testing for (which is different
5492 => the return list contains just 'plain' or 'login', 'cram-md5' or
5493 nothing depending on the value of ob->server_mech.
5495 I've just tested the patch: Authentication still works fine,
5496 unavailable mechs specified in the exim configuration are still
5497 caught, and the auth.log warnings about OTP are gone."
5499 PH/25 When debugging is enabled, the contents of the command line are added
5500 to the debugging output, even when log_selector=+arguments is not
5503 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5504 answer is "GNU", and only if the return is "GNU/something" is the answer
5507 PH/27 $acl_verify_message is now set immediately after the failure of a
5508 verification in an ACL, and so is available in subsequent modifiers. In
5509 particular, the message can be preserved by coding like this:
5511 warn !verify = sender
5512 set acl_m0 = $acl_verify_message
5514 Previously, $acl_verify_message was set only while expanding "message"
5515 and "log_message" when a very denied access.
5517 PH/28 Modified OS/os.c-Linux with
5519 -#ifndef OS_LOAD_AVERAGE
5520 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5522 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5523 nomenclature these days.)
5525 PH/29 Installed patch from the Sieve maintainer that adds the options
5526 sieve_useraddress and sieve_subaddress to the redirect router.
5528 PH/30 In these circumstances:
5529 . Two addresses routed to the same list of hosts;
5530 . First host does not offer TLS;
5531 . First host accepts first address;
5532 . First host gives temporary error to second address;
5533 . Second host offers TLS and a TLS session is established;
5534 . Second host accepts second address.
5535 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5536 and peerdn, if requested) that were in fact used only for the second
5539 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5540 attention to any local part prefix or suffix that was matched by the
5541 router that accepted the address. It now behaves in the same way as it
5542 does for delivery: the affixes are removed from the local part unless
5543 rcpt_include_affixes is set on the transport.
5545 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5546 timeout during the DATA phase of an incoming message.
5548 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5549 applied a patch sanctioned by the Sieve maintainer.
5551 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5552 the uid or gid is negative. A case of a negative gid caused this to be
5553 noticed. The fix allows for either to be negative.
5555 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5556 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5559 PH/36 The change PH/12 above was broken. Fixed it.
5561 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5562 the grounds that routing the same address twice would always produce the
5563 same answer. This might have been true once, but it is certainly no
5564 longer true now. Routing a child address may depend on the previous
5565 routing that produced that child. Some complicated redirection strategies
5566 went wrong when messages had multiple recipients, and made Exim's
5567 behaviour dependent on the order in which the addresses were given.
5569 I have moved the duplicate checking until after the routing is complete.
5570 Exim scans the addresses that are assigned to local and remote
5571 transports, and removes any duplicates. This means that more work will be
5572 done, as duplicates will always all be routed, but duplicates are
5573 presumably rare, so I don't expect this is of any significance.
5575 For deliveries to pipes, files, and autoreplies, the duplicate checking
5576 still happens during the routing process, since they are not going to be
5579 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5580 It corrects a timeout issue with spamd. This is Ian's comment: "The
5581 background is that sometimes spamd either never reads data from a
5582 connection it has accepted, or it never writes response data. The exiscan
5583 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5584 blindly assumes that writes won't block so it may never time out."
5586 PH/39 Allow G after quota size as well as K and M.
5588 PH/40 The value set for $authenticated_id in an authenticator may not contain
5589 binary zeroes or newlines because the value is written to log lines and
5590 to spool files. There was no check on this. Now the value is run through
5591 the string_printing() function so that such characters are converted to
5592 printable escape sequences.
5594 PH/41 $message_linecount is a new variable that contains the total number of
5595 lines in the message. Compare $body_linecount, which is the count for the
5598 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5599 bounce and delay warning messages, except in certain special cases, which
5602 (a) An SMTP error message from a remote host;
5603 (b) A message specified in a :fail: redirection;
5604 (c) A message specified in a "fail" command in a system filter;
5605 (d) A message specified in a FAIL return from the queryprogram router;
5606 (e) A message specified by the cannot_route_message router option.
5608 In these cases only, Exim does include the error details in bounce and
5609 warning messages. There are also a few cases where bland messages such
5610 as "unrouteable address" or "local delivery error" are given.
5612 PH/43 $value is now also set for the "else" part of a ${run expansion.
5614 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5615 being worked on, but at least Exim now implements the latest version to
5618 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5619 process to complete was treated as a delivery failure, a timeout while
5620 writing the message to the pipe was logged, but erroneously treated as a
5621 successful delivery. Such timeouts include transport filter timeouts. For
5622 consistency with the overall process timeout, these timeouts are now
5623 treated as errors, giving rise to delivery failures by default. However,
5624 there is now a new Boolean option for the pipe transport called
5625 timeout_defer, which, if set TRUE, converts the failures into defers for
5626 both kinds of timeout. A transport filter timeout is now identified in
5629 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5630 systems where "make" and "gmake" are different, calling "gmake" at top
5631 level broke things. I've arranged for the value of $(MAKE) to be passed
5632 from the Makefile to this script so that it can call the same version of
5636 A note about Exim versions 4.44 and 4.50
5637 ----------------------------------------
5639 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5640 changes of various kinds. As a consequence, a big documentation update was
5641 needed. This delayed the release for rather longer than seemed good, especially
5642 in the light of a couple of (minor) security issues. Therefore, the changes
5643 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5644 release. So 4.44 and 4.50 are in effect two different branches that both start
5647 I have left the 4.50 change log unchanged; it contains all the changes since
5648 4.43. The change log for 4.44 is below; many of its items are identical to
5649 those for 4.50. This seems to be the most sensible way to preserve the
5650 historical information.
5656 1. Minor wording change to the doc/README.SIEVE file.
5658 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5659 computation of the current number of files was incorrect.
5661 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5662 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5665 4. Give more explanation in the error message when the command for a transport
5666 filter fails to execute.
5668 5. There are several places where Exim runs a non-Exim command in a
5669 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5670 was being done only for the command run by the queryprogram router. It is
5671 now done for all such subprocesses. The other cases are: ${run, transport
5672 filters, and the commands run by the lmtp and pipe transports.
5674 6. Added CONFIGURE_GROUP build-time option.
5676 7. Some older OS have a limit of 256 on the maximum number of file
5677 descriptors. Exim was using setrlimit() to set 1000 as a large value
5678 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5679 systems. I've change it so that if it can't get 1000, it tries for 256.
5681 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5682 was an oversight, and furthermore, ever since the addition of extra
5683 controls (e.g. 4.43/32), the checks on when to allow different forms of
5684 "control" were broken. There should now be diagnostics for all cases when a
5685 control that does not make sense is encountered.
5687 9. Added the /retain_sender option to "control=submission".
5689 10. $recipients is now available in the predata ACL (oversight).
5691 11. Tidy the search cache before the fork to do a delivery from a message
5692 received from the command line. Otherwise the child will trigger a lookup
5693 failure and thereby defer the delivery if it tries to use (for example) a
5694 cached ldap connection that the parent has called unbind on.
5696 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5697 of $address_data from the recipient verification was clobbered by the
5698 sender verification.
5700 13. The value of address_data from a sender verification is now available in
5701 $sender_address_data in subsequent conditions in the ACL statement.
5703 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5705 15. Added a new option "connect=<time>" to callout options, to set a different
5708 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5709 was its contents. (It was OK if the option was not defined at all.)
5711 17. A "Completed" log line is now written for messages that are removed from
5712 the spool by the -Mrm option.
5714 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5715 information about exactly what failed.
5717 19. Added -dd to debug only the daemon process.
5719 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5720 handles timeouts, both on the server side and network timeouts. Renamed the
5721 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5723 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5724 It is now set to "smtps".
5726 22. $host_address is now set to the target address during the checking of
5727 ignore_target_hosts.
5729 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5730 being passed; this would have caused $sender_host_name to have been used if
5731 matching the list had actually called for a host name (not very likely,
5732 since this list is usually IP addresses). A host name is now passed as
5735 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5736 code that specifies a non-restarting handler (typically sigaction() in
5737 modern systems) in an attempt to fix a rare and obscure crash bug.
5739 25. Narrowed the window for a race in the daemon that could cause it to ignore
5740 SIGCHLD signals. This is not a major problem, because they are used only to
5741 wake it up if nothing else does.
5743 26. A malformed maildirsize file could cause Exim to calculate negative values
5744 for the mailbox size or file count. Odd effects could occur as a result.
5745 The maildirsize information is now recalculated if the size or filecount
5748 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5749 support for a long time. Removed HAVE_SYS_VFS_H.
5751 28. Installed the latest version of exipick from John Jetmore.
5753 29. In an address list, if the pattern was not a regular expression, an empty
5754 subject address (from a bounce message) matched only if the pattern was an
5755 empty string. Non-empty patterns were not even tested. This was the wrong
5756 because it is perfectly reasonable to use an empty address as part of a
5757 database query. An empty address is now tested by patterns that are
5758 lookups. However, all the other forms of pattern expect the subject to
5759 contain a local part and a domain, and therefore, for them, an empty
5760 address still always fails if the pattern is not itself empty.
5762 30. Exim went into a mad DNS loop when attempting to do a callout where the
5763 host was specified on an smtp transport, and looking it up yielded more
5764 than one IP address.
5766 31. Re-factored the code for checking spool and log partition space into a
5767 function that finds that data and another that does the check. The former
5768 is then used to implement four new variables: $spool_space, $log_space,
5769 $spool_inodes, and $log_inodes.
5771 32. The RFC2047 encoding function was originally intended for short strings
5772 such as real names; it was not keeping to the 75-character limit for
5773 encoded words that the RFC imposes. It now respects the limit, and
5774 generates multiple encoded words if necessary. To be on the safe side, I
5775 have increased the buffer size for the ${rfc2047: expansion operator from
5778 33. It is now permitted to omit both strings after an "if" condition; if the
5779 condition is true, the result is "true". As before, when the second string
5780 is omitted, a false condition yields an empty string. This makes it less
5781 cumbersome to write custom ACL and router conditions.
5783 34. Failure to deliver a bounce message always caused it to be frozen, even if
5784 there was an errors_to setting on the router. The errors_to setting is now
5787 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5788 canonical form (fully expanded) before being placed in
5789 $sender_host_address.
5791 36. The table in the code that translates DNS record types into text (T_A to
5792 "A" for instance) was missing entries for NS and CNAME. It is just possible
5793 that this could have caused confusion if both these types were looked up
5794 for the same domain, because the text type is used as part of Exim's
5795 per-process caching. But the chance of anyone hitting this buglet seems
5798 37. The dnsdb lookup has been extended in a number of ways.
5800 (1) There is a new type, "zns", which walks up the domain tree until it
5801 finds some nameserver records. It should be used with care.
5803 (2) There is a new type, "mxh", which is like "mx" except that it returns
5804 just the host names, not the priorities.
5806 (3) It is now possible to give a list of domains (or IP addresses) to be
5807 looked up. The behaviour when one of the lookups defers can be
5808 controlled by a keyword.
5810 (4) It is now possible to specify the separator character for use when
5811 multiple records are returned.
5813 38. The dnslists ACL condition has been extended: it is now possible to supply
5814 a list of IP addresses and/or domains to be looked up in a particular DNS
5817 39. Added log_selector=+queue_time_overall.
5819 40. When running the queue in the test harness, wait just a tad after forking a
5820 delivery process, to get repeatability of debugging output.
5822 41. Include certificate and key file names in error message when GnuTLS fails
5823 to set them up, because the GnuTLS error message doesn't include the name
5824 of the failing file when there is a problem reading it.
5826 42. Allow both -bf and -bF in the same test run.
5828 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5830 44. The "Exiscan patch" is now merged into the mainline Exim source.
5832 45. Sometimes the final signoff response after QUIT could fail to get
5833 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5834 < 0 before doing a fflush(). This bug looks as though it goes back to the
5835 introduction of TLS in release 3.20, but "sometimes" must have been rare
5836 because the tests only now provoked it.
5838 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5839 (this can affect the format of dates).
5841 47. exim_tidydb, when checking for the continued existence of a message for
5842 which it has found a message-specific retry record, was not finding
5843 messages that were in split spool directories. Consequently, it was
5844 deleting retry records that should have stayed in existence.
5846 48. Steve fixed some bugs in eximstats.
5848 49. The SPA authentication driver was not abandoning authentication and moving
5849 on to the next authenticator when an expansion was forced to fail,
5850 contradicting the general specification for all authenticators. Instead it
5851 was generating a temporary error. It now behaves as specified.
5853 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5854 (the order specifies the preference for clients). The order is now AES256,
5855 AES128, 3DES, ARCFOUR128.
5857 51. Small patch to Sieve code - explicitly set From: when generating an
5860 52. Exim crashed if a remote delivery caused a very long error message to be
5861 recorded - for instance if somebody sent an entire SpamAssassin report back
5862 as a large number of 550 error lines. This bug was coincidentally fixed by
5863 increasing the size of one of Exim's internal buffers (big_buffer) that
5864 happened as part of the Exiscan merge. However, to be on the safe side, I
5865 have made the code more robust (and fixed the comments that describe what
5868 53. Now that there can be additional text after "Completed" in log lines (if
5869 the queue_time_overall log selector is set), a one-byte patch to exigrep
5870 was needed to allow it to recognize "Completed" as not the last thing in
5873 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5874 patch that reportedly fixes this has been added. I am not expert enough to
5875 create a test for it. This is what the patch creator wrote:
5877 "I found a little strange behaviour of ldap code when working with
5878 Windows 2003 AD Domain, where users was placed in more than one
5879 Organization Units. When I tried to give exim partial DN, the exit code
5880 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5881 But simultaneously result of request was absolutely normal ldap result,
5882 so I produce this patch..."
5884 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5885 REFERENCE, so I have modified the code to exclude the patch when that macro
5888 55. Some experimental protocols are using DNS PTR records for new purposes. The
5889 keys for these records are domain names, not reversed IP addresses. The
5890 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5891 leaves it alone. Component reversal etc. now happens only for IP addresses.
5894 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5896 57. Double the size of the debug message buffer (to 2048) so that more of very
5897 long debug lines gets shown.
5899 58. The exicyclog utility now does better if the number of log files to keep
5900 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5902 59. Two changes related to the smtp_active_hostname option:
5904 (1) $smtp_active_hostname is now available as a variable.
5905 (2) The default for smtp_banner uses $smtp_active_hostname instead
5906 of $primary_hostname.
5908 60. The host_aton() function is supposed to be passed a string that is known
5909 to be a valid IP address. However, in the case of IPv6 addresses, it was
5910 not checking this. This is a hostage to fortune. Exim now panics and dies
5911 if the condition is not met. A case was found where this could be provoked
5912 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5913 components; fortuitously, this particular loophole had already been fixed
5914 by change 4.50/55 above.
5916 If there are any other similar loopholes, the new check in host_aton()
5917 itself should stop them being exploited. The report I received stated that
5918 data on the command line could provoke the exploit when Exim was running as
5919 exim, but did not say which command line option was involved. All I could
5920 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5921 running as the user.
5924 61. There was a buffer overflow vulnerability in the SPA authentication code
5925 (which came originally from the Samba project). I have added a test to the
5926 spa_base64_to_bits() function which I hope fixes it.
5929 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5930 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5932 63. The daemon start-up calls getloadavg() while still root for those OS that
5933 need the first call to be done as root, but it missed one case: when
5934 deliver_queue_load_max is set with deliver_drop_privilege. This is
5935 necessary for the benefit of the queue runner, because there is no re-exec
5936 when deliver_drop_privilege is set.
5938 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5939 This has been fixed.
5941 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5942 was in use, was not putting the data itself into the right store pool;
5943 consequently, it could be overwritten for a subsequent message in the same
5944 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5947 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5949 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5950 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5951 the function were treating the return as a boolean value, which happened to
5952 work because 0=false and not-0=true, but is not correct code.
5954 68. The host_aton() function was not handling scoped IPv6 addresses (those
5955 with, for example, "%eth0" on the end) correctly.
5957 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5958 negated items (that is, ~something) in unsigned ints. Some compilers
5959 apparently mutter when there is no cast.
5961 70. If an address verification called from an ACL failed, and did not produce a
5962 user-specific message (i.e. there was only a "system" message), nothing was
5963 put in $acl_verify_message. In this situation, it now puts the system
5966 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5967 session; change 4.31/43 added the unwanted input to the log line - except
5968 that it did not do this in the start of session case. It now does.
5970 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5971 This is wrong and can cause the other end to generate a synchronization
5972 error if it is another Exim or anything else that does the synchronization
5973 check. A QUIT command is no longer sent after a timeout.
5975 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5976 during host lookups.
5978 74. The defer_ok option of callout verification was not working if it was used
5979 when verifying addresses in header lines, that is, for this case:
5981 verify = header_sender/callout=defer_ok
5983 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5984 those file descriptors could be used for SMTP connections. If anything
5985 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5986 could be sent to the SMTP client, causing chaos. The daemon now opens
5987 stdin, stdout, and stderr to /dev/null when it puts itself into the
5990 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5991 log by default. The user can override this with suitable Perl magic.
5993 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5994 the log message when discard triggers, was not working for the DATA ACL or
5995 for the non-SMTP ACL.
5997 78. Error message wording change in sieve.c.
5999 79. If smtp_accept_max_per_host was set, the number of connections could be
6000 restricted to fewer than expected, because the daemon was trying to set up
6001 a new connection before checking whether the processes handling previous
6002 connections had finished. The check for completed processes is now done
6003 earlier. On busy systems, this bug wouldn't be noticed because something
6004 else would have woken the daemon, and it would have reaped the completed
6007 80. If a message was submitted locally by a user whose login name contained one
6008 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6009 It caused a spool format error. I have fixed the spool reading code. A
6010 related problem was that the "from" clause in the Received: line became
6011 illegal because of the space(s). It is now covered by ${quote_local_part.
6013 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6016 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6018 83. Patch from Sieve maintainer to fix unterminated string problem in
6019 "vacation" handling.
6021 84. Some minor changes to the Linux configuration files to help with other
6022 OS variants using glibc.
6024 85. One more patch for Sieve to update vacation handling to latest spec.
6027 ----------------------------------------------------
6028 See the note above about the 4.44 and 4.50 releases.
6029 ----------------------------------------------------
6035 1. Change 4.43/35 introduced a bug that caused file counts to be
6036 incorrectly computed when quota_filecount was set in an appendfile
6039 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6040 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6043 3. Give more explanation in the error message when the command for a transport
6044 filter fails to execute.
6046 4. There are several places where Exim runs a non-Exim command in a
6047 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6048 was being done only for the command run by the queryprogram router. It is
6049 now done for all such subprocesses. The other cases are: ${run, transport
6050 filters, and the commands run by the lmtp and pipe transports.
6052 5. Some older OS have a limit of 256 on the maximum number of file
6053 descriptors. Exim was using setrlimit() to set 1000 as a large value
6054 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6055 systems. I've change it so that if it can't get 1000, it tries for 256.
6057 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6058 was an oversight, and furthermore, ever since the addition of extra
6059 controls (e.g. 4.43/32), the checks on when to allow different forms of
6060 "control" were broken. There should now be diagnostics for all cases when a
6061 control that does not make sense is encountered.
6063 7. $recipients is now available in the predata ACL (oversight).
6065 8. Tidy the search cache before the fork to do a delivery from a message
6066 received from the command line. Otherwise the child will trigger a lookup
6067 failure and thereby defer the delivery if it tries to use (for example) a
6068 cached ldap connection that the parent has called unbind on.
6070 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6071 of $address_data from the recipient verification was clobbered by the
6072 sender verification.
6074 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6075 was its contents. (It was OK if the option was not defined at all.)
6077 11. A "Completed" log line is now written for messages that are removed from
6078 the spool by the -Mrm option.
6080 12. $host_address is now set to the target address during the checking of
6081 ignore_target_hosts.
6083 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6084 being passed; this would have caused $sender_host_name to have been used if
6085 matching the list had actually called for a host name (not very likely,
6086 since this list is usually IP addresses). A host name is now passed as
6089 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6090 code that specifies a non-restarting handler (typically sigaction() in
6091 modern systems) in an attempt to fix a rare and obscure crash bug.
6093 15. Narrowed the window for a race in the daemon that could cause it to ignore
6094 SIGCHLD signals. This is not a major problem, because they are used only to
6095 wake it up if nothing else does.
6097 16. A malformed maildirsize file could cause Exim to calculate negative values
6098 for the mailbox size or file count. Odd effects could occur as a result.
6099 The maildirsize information is now recalculated if the size or filecount
6102 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6103 support for a long time. Removed HAVE_SYS_VFS_H.
6105 18. Updated exipick to current release from John Jetmore.
6107 19. Allow an empty sender to be matched against a lookup in an address list.
6108 Previously the only cases considered were a regular expression, or an
6111 20. Exim went into a mad DNS lookup loop when doing a callout where the
6112 host was specified on the transport, if the DNS lookup yielded more than
6115 21. The RFC2047 encoding function was originally intended for short strings
6116 such as real names; it was not keeping to the 75-character limit for
6117 encoded words that the RFC imposes. It now respects the limit, and
6118 generates multiple encoded words if necessary. To be on the safe side, I
6119 have increased the buffer size for the ${rfc2047: expansion operator from
6122 22. Failure to deliver a bounce message always caused it to be frozen, even if
6123 there was an errors_to setting on the router. The errors_to setting is now
6126 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6127 canonical form (fully expanded) before being placed in
6128 $sender_host_address.
6130 24. Updated eximstats to version 1.33
6132 25. Include certificate and key file names in error message when GnuTLS fails
6133 to set them up, because the GnuTLS error message doesn't include the name
6134 of the failing file when there is a problem reading it.
6136 26. Expand error message when OpenSSL has problems setting up cert/key files.
6139 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6140 (this can affect the format of dates).
6142 28. exim_tidydb, when checking for the continued existence of a message for
6143 which it has found a message-specific retry record, was not finding
6144 messages that were in split spool directories. Consequently, it was
6145 deleting retry records that should have stayed in existence.
6147 29. eximstats updated to version 1.35
6148 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6149 1.35 - bugfix such that pie charts by volume are generated correctly
6151 30. The SPA authentication driver was not abandoning authentication and moving
6152 on to the next authenticator when an expansion was forced to fail,
6153 contradicting the general specification for all authenticators. Instead it
6154 was generating a temporary error. It now behaves as specified.
6156 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6157 (the order specifies the preference for clients). The order is now AES256,
6158 AES128, 3DES, ARCFOUR128.
6160 31. Small patch to Sieve code - explicitly set From: when generating an
6163 32. Exim crashed if a remote delivery caused a very long error message to be
6164 recorded - for instance if somebody sent an entire SpamAssassin report back
6165 as a large number of 550 error lines. This bug was coincidentally fixed by
6166 increasing the size of one of Exim's internal buffers (big_buffer) that
6167 happened as part of the Exiscan merge. However, to be on the safe side, I
6168 have made the code more robust (and fixed the comments that describe what
6171 33. Some experimental protocols are using DNS PTR records for new purposes. The
6172 keys for these records are domain names, not reversed IP addresses. The
6173 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6174 leaves it alone. Component reversal etc. now happens only for IP addresses.
6177 34. The host_aton() function is supposed to be passed a string that is known
6178 to be a valid IP address. However, in the case of IPv6 addresses, it was
6179 not checking this. This is a hostage to fortune. Exim now panics and dies
6180 if the condition is not met. A case was found where this could be provoked
6181 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6182 components; fortuitously, this particular loophole had already been fixed
6183 by change 4.50/55 or 4.44/33 above.
6185 If there are any other similar loopholes, the new check in host_aton()
6186 itself should stop them being exploited. The report I received stated that
6187 data on the command line could provoke the exploit when Exim was running as
6188 exim, but did not say which command line option was involved. All I could
6189 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6190 running as the user.
6193 35. There was a buffer overflow vulnerability in the SPA authentication code
6194 (which came originally from the Samba project). I have added a test to the
6195 spa_base64_to_bits() function which I hope fixes it.
6198 36. The daemon start-up calls getloadavg() while still root for those OS that
6199 need the first call to be done as root, but it missed one case: when
6200 deliver_queue_load_max is set with deliver_drop_privilege. This is
6201 necessary for the benefit of the queue runner, because there is no re-exec
6202 when deliver_drop_privilege is set.
6204 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6205 was in use, was not putting the data itself into the right store pool;
6206 consequently, it could be overwritten for a subsequent message in the same
6207 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6210 38. Sometimes the final signoff response after QUIT could fail to get
6211 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6212 < 0 before doing a fflush(). This bug looks as though it goes back to the
6213 introduction of TLS in release 3.20, but "sometimes" must have been rare
6214 because the tests only now provoked it.
6220 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6221 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6222 Now it returns an integer. A number of calls were still expecting a T/F
6223 return. Fortuitously, in all cases, the tests worked in OK situations,
6224 which is the norm. However, things would have gone wrong on any write
6225 failures on the smtp file descriptor. This function is used when sending
6226 messages over SMTP and also when doing verify callouts.
6228 2. When Exim is called to do synchronous delivery of a locally submitted
6229 message (the -odf or -odi options), it no longer closes stderr before doing
6232 3. Implemented the mua_wrapper option.
6234 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6236 5. Implemented the functions header_remove(), header_testname(),
6237 header_add_at_position(), and receive_remove_recipient(), and exported them
6240 6. If an ACL "warn" statement specified the addition of headers, Exim already
6241 inserted X-ACL-Warn: at the start if there was no header name. However, it
6242 was not making this test for the second and subsequent header lines if
6243 there were newlines in the string. This meant that an invalid header could
6244 be inserted if Exim was badly configured.
6246 7. Allow an ACL "warn" statement to add header lines at the start or after all
6247 the Received: headers, as well as at the end.
6249 8. Added the rcpt_4xx retry error code.
6251 9. Added postmaster_mailfrom=xxx to callout verification option.
6253 10. Added mailfrom=xxxx to the callout verification option, for verify=
6256 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6257 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6259 12. Inserted some casts to stop certain compilers warning when using pointer
6260 differences as field lengths or precisions in printf-type calls (mostly
6261 affecting debugging statements).
6263 13. Added optional readline() support for -be (dynamically loaded).
6265 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6266 same clock tick as a message's arrival, so that its received time was the
6267 same as the "first fail" time on the retry record, and that message
6268 remained on the queue past the ultimate address timeout, every queue runner
6269 would try a delivery (because it was past the ultimate address timeout) but
6270 after another failure, the ultimate address timeout, which should have then
6271 bounced the address, did not kick in. This was a "< instead of <=" error;
6272 in most cases the first failure would have been in the next clock tick
6273 after the received time, and all would be well.
6275 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6276 being recognized when the domain list was tested by the match_domain
6277 condition in an expansion string.
6279 16. Added the ${str2b64: operator.
6281 17. Exim was always calling setrlimit() to set a large limit for the number of
6282 processes, without checking whether the existing limit was already
6283 adequate. (It did check for the limit on file descriptors.) Furthermore,
6284 errors from getrlimit() and setrlimit() were being ignored. Now they are
6285 logged to the main and panic logs, but Exim does carry on, to try to do its
6286 job under whatever limits there are.
6288 18. Imported PCRE 5.0.
6290 19. Trivial typo in log message " temporarily refused connection" (the leading
6293 20. If the log selector return_path_on_delivery was set and an address was
6294 redirected to /dev/null, the delivery process crashed because it assumed
6295 that a return path would always be set for a "successful" delivery. In this
6296 case, the whole delivery is bypassed as an optimization, and therefore no
6299 21. Internal re-arrangement: the function for sending a challenge and reading
6300 a response while authentication was assuming a zero-terminated challenge
6301 string. It's now changed to take a pointer and a length, to allow for
6302 binary data in such strings.
6304 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6306 23. Exim was not respecting finduser_retries when seeking the login of the
6307 uid under which it was called; it was always trying 10 times. (The default
6308 setting of finduser_retries is zero.) Also, it was sleeping after the final
6309 failure, which is pointless.
6311 24. Implemented tls_on_connect_ports.
6313 25. Implemented acl_smtp_predata.
6315 26. If the domain in control=submission is set empty, Exim assumes that the
6316 authenticated id is a complete email address when it generates From: or
6317 Sender: header lines.
6319 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6320 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6321 chown and chgrp in /bin and hostname in /usr/bin.
6323 28. Exim was keeping the "process log" file open after each use, just as it
6324 does for the main log. This opens the possibility of it remaining open for
6325 long periods when the USR1 signal hits a daemon. Occasional processlog
6326 errors were reported, that could have been caused by this. Anyway, it seems
6327 much more sensible not to leave this file open at all, so that is what now
6330 29. The long-running daemon process does not normally write to the log once it
6331 has entered its main loop, and it closes the log before doing so. This is
6332 so that log files can straightforwardly be renamed and moved. However,
6333 there are a couple of unusual error situations where the daemon does write
6334 log entries, and I had neglected to close the log afterwards.
6336 30. The text of an SMTP error response that was received during a remote
6337 delivery was being truncated at 512 bytes. This is too short for some of
6338 the long messages that one sometimes sees. I've increased the limit to
6341 31. It is now possible to make retry rules that apply only when a message has a
6342 specific sender, in particular, an empty sender.
6344 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6345 it possible to be selective about when SMTP synchronization is enforced.
6347 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6349 32. Implemented hosts_connection_nolog.
6351 33. Added an ACL for QUIT.
6353 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6356 35. Added mailbox_size and mailbox_filecount to appendfile.
6358 36. Added control = no_multiline_responses to ACLs.
6360 37. There was a bug in the logic of the code that waits for the clock to tick
6361 in the case where the clock went backwards by a substantial amount such
6362 that the microsecond fraction of "now" was more than the microsecond
6363 fraction of "then" (but the whole seconds number was less).
6365 38. Added support for the libradius Radius client library this is found on
6366 FreeBSD (previously only the radiusclient library was supported).
6372 1. When certain lookups returned multiple values in the form name=value, the
6373 quoting of the values was not always being done properly. Specifically:
6374 (a) If the value started with a double quote, but contained no whitespace,
6376 (b) If the value contained whitespace other than a space character (i.e.
6377 tabs or newlines or carriage returns) it was not quoted.
6378 This fix has been applied to the mysql and pgsql lookups by writing a
6379 separate quoting function and calling it from the lookup code. The fix
6380 should probably also be applied to nisplus, ibase and oracle lookups, but
6381 since I cannot test any of those, I have not disturbed their existing code.
6383 2. A hit in the callout cache for a specific address caused a log line with no
6384 reason for rejecting RCPT. Now it says "Previous (cached) callout
6385 verification failure".
6387 3. There was an off-by-one bug in the queryprogram router. An over-long
6388 return line was truncated at 256 instead of 255 characters, thereby
6389 overflowing its buffer with the terminating zero. As well as fixing this, I
6390 have increased the buffer size to 1024 (and made a note to document this).
6392 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6393 when Exim is waiting for an SMTP response from a remote server, Exim
6394 restarts its select() call on the socket, thereby resetting its timeout.
6395 This is not a problem when such interrupts are rare. Somebody set up a cron
6396 job to run exiwhat every 2 minutes, which is less than the normal select()
6397 timeout (5 or 10 minutes). This meant that the select() timeout never
6398 kicked in because it was always reset. I have fixed this by comparing the
6399 time when an interrupt arrives with the time at the start of the first call
6400 to select(). If more time than the timeout has elapsed, the interrupt is
6401 treated as a timeout.
6403 5. Some internal re-factoring in preparation for the addition of Sieve
6404 extensions (by MH). In particular, the "personal" test is moved to a
6405 separate function, and given an option for scanning Cc: and Bcc: (which is
6406 not set for Exim filters).
6408 6. When Exim created an email address using the login of the caller as the
6409 local part (e.g. when creating a From: or Sender: header line), it was not
6410 quoting the local part when it contained special characters such as @.
6412 7. Installed new OpenBSD configuration files.
6414 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6415 try to make them clearer.
6417 9. Callout options, other than the timeout value, were being ignored when
6418 verifying sender addresses in header lines. For example, when using
6420 verify = header_sender/callout=no_cache
6422 the cache was (incorrectly) being used.
6424 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6425 only the Cygwin environment.
6427 11. When return_path_on_delivery was set as a log selector, if different remote
6428 addresses in the same message used different return paths and parallel
6429 remote delivery occurred, the wrong values would sometimes be logged.
6430 (Whenever a remote delivery process finished, the return path value from
6431 the most recently started remote delivery process was logged.)
6433 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6434 lines when AUTH and/or TLS are in use. This is the "received protocol"
6435 field. Exim used to use "asmtp" for authenticated SMTP, without any
6436 indication (in the protocol name) for TLS use. Now it follows the RFC and
6437 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6438 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6439 names appear in log lines as well as in Received: header lines.
6441 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6442 extensions, and comparison tests, and to fix some bugs.
6444 14. Changes to the "personal" filter test:
6446 (1) The test was buggy in that it was just doing the equivalent of
6447 "contains" tests on header lines. For example, if a user's address was
6448 anne@some.where, the "personal" test would incorrectly be true for
6450 To: susanne@some.where
6452 This test is now done by extracting each address from the header in turn,
6453 and checking the entire address. Other tests that are part of "personal"
6454 are now done using regular expressions (for example, to check local parts
6455 of addresses in From: header lines).
6457 (2) The list of non-personal local parts in From: addresses has been
6458 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6459 taken from the Sieve specification recommendations.
6461 (3) If the message contains any header line starting with "List-" it is
6462 treated as non-personal.
6464 (4) The test for "circular" in the Subject: header line has been removed
6465 because it now seems ill-conceived.
6467 15. Minor typos in src/EDITME comments corrected.
6469 16. Installed latest exipick from John Jetmore.
6471 17. If headers_add on a router specified a text string that was too long for
6472 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6473 of string_sprintf() is now avoided.
6475 18. $message_body_size was not set (it was always zero) when running the DATA
6476 ACL and the local_scan() function.
6478 19. For the "mail" command in an Exim filter, no default was being set for
6479 the once_repeat time, causing a random time value to be used if "once" was
6480 specified. (If the value happened to be <= 0, no repeat happened.) The
6481 default is now 0s, meaning "never repeat". The "vacation" command was OK
6482 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6483 (I found it when inspecting the code).
6485 20. There is now an overall timeout for performing a callout verification. It
6486 defaults to 4 times the callout timeout, which applies to individual SMTP
6487 commands during the callout. The overall timeout applies when there is more
6488 than one host that can be tried. The timeout is checked before trying the
6489 next host. This prevents very long delays if there are a large number of
6490 hosts and all are timing out (e.g. when the network connections are timing
6491 out). The value of the overall timeout can be changed by specifying an
6492 additional sub-option for "callout", called "maxwait". For example:
6494 verify = sender/callout=5s,maxwait=20s
6496 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6497 to the end before writing, but this should make it even safer).
6499 22. Exim was forgetting that it had advertised PIPELINING for the second and
6500 subsequent messages on an SMTP connection. It was also not resetting its
6501 memory on STARTTLS and an internal HELO.
6503 23. When Exim logs an SMTP synchronization error within a session, it now
6504 records whether PIPELINING has been advertised or not.
6506 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6507 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6508 rather than long int.
6510 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6512 26. Added the never_mail option to autoreply.
6518 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6519 crash if the getsockname() call failed; this can happen if a connection is
6520 closed very soon after it is established. The problem was simply in the
6521 order in which certain operations were done, causing Exim to try to write
6522 to the SMTP stream before it had set up the file descriptor. The bug has
6523 been fixed by making things happen in the correct order.
6529 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6530 before the connection was closed, thus losing the rejection response.
6532 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6533 some early Solaris releases, but causes trouble in current releases where
6534 socklen_t is defined.
6536 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6539 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6542 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6543 that releasing the top of it at the end releases what was used for sub-
6544 expansions (unless the block got too big). However, discard this block if
6545 the first thing is a variable or header, so that we can use its block when
6546 it is dynamic (useful for very large $message_headers, for example).
6548 6. Lookups now cache *every* query, not just the most recent. A new, separate
6549 store pool is used for this. It can be recovered when all lookup caches are
6550 flushed. Lookups now release memory at the end of their result strings.
6551 This has involved some general refactoring of the lookup sources.
6553 7. Some code has been added to the store_xxx() functions to reduce the amount
6554 of flapping under certain conditions.
6556 8. log_incoming_interface used to affect only the <= reception log lines. Now
6557 it causes the local interface and port to be added to several more SMTP log
6558 lines, for example "SMTP connection from", and rejection lines.
6560 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6562 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6564 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6565 could be overwritten at the end of the current message (or the start of a
6566 new message if it was set in a HELO ACL). The value is now preserved for
6567 the duration of the SMTP connection.
6569 12. If a transport had a headers_rewrite setting, and a matching header line
6570 contained an unqualified address, that address was qualified, even if it
6571 did not match any rewriting rules. The underlying bug was that the values
6572 of the flags that permit the existence of unqualified sender and recipient
6573 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6574 non-local messages, and by -bnq for local messages) were not being
6575 preserved with the message after it was received.
6577 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6578 "next input=" as part of the text comprising the host identity instead of
6579 the correct text. The code was using the same buffer for two different
6580 strings. However, depending on which order the printing function evaluated
6581 its arguments, the bug did not always show up. Under Linux, for example, my
6582 test suite worked just fine.
6584 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6585 doesn't work with some older versions of Perl. It has been changed to "my",
6586 which in any case is probably the better facility to use.
6588 15. A really picky compiler found some instances of statements for creating
6589 error messages that either had too many or two few arguments for the format
6592 16. The size of the buffer for calls to the DNS resolver has been increased
6593 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6594 for addresses that have a lot of PTR records. This alleviates a problem; it
6595 does not fully solve it.
6597 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6598 buffer now truncates the list and logs the incident, which is the same
6599 action as happens when Exim is looking up a host name and its aliases.
6600 Previously in this situation something unpredictable would happen;
6601 sometimes it was "internal error: store_reset failed".
6603 18. If a server dropped the connection unexpectedly when an Exim client was
6604 using GnuTLS and trying to read a response, the client delivery process
6605 crashed while trying to generate an error log message.
6607 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6608 string, for example:
6610 warn message = H1: something\nH2: something
6612 the text was added as a single header line from Exim's point of view
6613 though it ended up OK in the delivered message. However, searching for the
6614 second and subsequent header lines using $h_h2: did not work. This has been
6615 fixed. Similarly, if a system filter added multiple headers in this way,
6616 the routers could not see them.
6618 20. Expanded the error message when iplsearch is called with an invalid key to
6619 suggest using net-iplsearch in a host list.
6621 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6622 ACLs are no longer actually imposed (and a message to that effect is
6625 22. If a "gecos" field in a passwd entry contained escaped characters, in
6626 particular, if it contained a \" sequence, Exim got it wrong when building
6627 a From: or a Sender: header from that name. A second bug also caused
6628 incorrect handling when an unquoted " was present following a character
6629 that needed quoting.
6631 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6632 was not being matched caselessly.
6634 24. Arranged for all hyphens in the exim.8 source to be escaped with
6637 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6638 properly. Recipient callout cache records were still being keyed to include
6639 the sender, even when use_sender was set false. This led to far more
6640 callouts that were necessary. The sender is no longer included in the key
6641 when use_sender is false.
6643 26. Added "control = submission" modifier to ACLs.
6645 27. Added the ${base62d: operator to decode base 62 numbers.
6647 28. dnsdb lookups can now access SRV records.
6649 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6650 the configuration file.
6652 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6653 (-v) mode. This makes the output for a verbose queue run more intelligible.
6655 31. Added a use_postmaster feature to recipient callouts.
6657 32. Added the $body_zerocount variable, containing the number of binary zero
6658 bytes in the message body.
6660 33. The time of last modification of the "new" subdirectory is now used as the
6661 "mailbox time last read" when there is a quota error for a maildir
6664 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6666 35. Added +ignore_unknown as a special item in host lists.
6668 36. Code for decoding IPv6 addresses in host lists is now included, even if
6669 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6670 address was recognized as an IP address, but was then not correctly decoded
6671 into binary, causing unexpected and incorrect effects when compared with
6678 1. Very minor rewording of debugging text in manualroute to say "list of
6679 hosts" instead of "hostlist".
6681 2. If verify=header_syntax was set, and a header line with an unqualified
6682 address (no domain) and a large number of spaces between the end of the
6683 name and the colon was received, the reception process suffered a buffer
6684 overflow, and (when I tested it) crashed. This was caused by some obsolete
6685 code that should have been removed. The fix is to remove it!
6687 3. When running in the test harness, delay a bit after writing a bounce
6688 message to get a bit more predictability in the log output.
6690 4. Added a call to search_tidyup() just before forking a reception process. In
6691 theory, someone could use a lookup in the expansion of smtp_accept_max_
6692 per_host which, without the tidyup, could leave open a database connection.
6694 5. Added the variables $recipient_data and $sender_data which get set from a
6695 lookup success in an ACL "recipients" or "senders" condition, or a router
6696 "senders" option, similar to $domain_data and $local_part_data.
6698 6. Moved the writing of debug_print from before to after the "senders" test
6701 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6702 problems for message scanning, either using a data ACL, or using
6703 local_scan() because the Received: header was not generated till after they
6704 were called (in order to set the time as the time of reception completion).
6705 I have revised the way this works. The header is now generated after the
6706 body is received, but before the ACL or local_scan() are called. After they
6707 are run, the timestamp in the header is updated.
6713 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6714 before starting a queue runner without re-exec. This happened only when
6715 deliver_drop_privilege was set or when the Exim user was set to root. The
6716 effect of the bug was that timeouts during subsequent deliveries caused
6717 crashes instead of being properly handled. The handler is now left at its
6718 default (and expected) setting.
6720 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6721 message, again when deliver_drop_privilege is set or Exim is run as root.
6722 The bug described in (1) was not present in this case, but the tidying up
6723 of the other signals was missing. I have made the two cases consistent.
6725 3. The ignore_target_hosts setting on a manualroute router was being ignored
6726 for hosts that were looked up using the /MX notation.
6728 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6731 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6732 operated on the sender address. After changing the $sender_address to <>
6733 for the sender address verify, Exim was re-instated it as the original
6734 (before rewriting) address, but remembering that it had rewritten it, so it
6735 wasn't rewriting it again. This bug also had the effect of breaking the
6736 sender address verification caching when the sender address was rewritten.
6738 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6739 This has been changed so that if the ip literal address matches
6740 ignore_target_hosts, the router declines.
6742 7. Added expansion conditions match_domain, match_address, and match_local_
6743 part (NOT match_host).
6745 8. The placeholder for the Received: header didn't have a length field set.
6747 9. Added code to Exim itself and to exim_lock to test for a specific race
6748 condition that could lead to file corruption when using MBX delivery. The
6749 issue is with the lockfile that is created in /tmp. If this file is removed
6750 after a process has opened it but before that process has acquired a lock,
6751 there is the potential for a second process to recreate the file and also
6752 acquire a lock. This could lead to two Exim processes writing to the file
6753 at the same time. The added code performs the same test as UW imapd; it
6754 checks after acquiring the lock that its file descriptor still refers to
6755 the same named file.
6757 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6758 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6761 11. Added the smtp_active_hostname option. If used, this will typically be made
6762 to depend on the incoming interface address. Because $interface_address is
6763 not set up until the daemon has forked a reception process, error responses
6764 that can happen earlier (such as "too many connections") no longer contain
6767 12. If an expansion in a condition on a "warn" statement fails because a lookup
6768 defers, the "warn" statement is abandoned, and the next ACL statement is
6769 processed. Previously this caused the whole ACL to be aborted.
6771 13. Added the iplsearch lookup type.
6773 14. Added ident_timeout as a log selector.
6775 15. Added tls_certificate_verified as a log selector.
6777 16. Added a global option tls_require_ciphers (compare the smtp transport
6778 option of the same name). This controls incoming TLS connections.
6780 17. I finally figured out how to make tls_require_ciphers do a similar thing
6781 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6782 before starting the TLS session.
6784 18. Tabs are now shown as \t in -bP output.
6786 19. If the log selector return_path_on_delivery was set, Exim crashed when
6787 bouncing a message because it had too many Received: header lines.
6789 20. If two routers both had headers_remove settings, and the first one included
6790 a superfluous trailing colon, the final name in the first list and the
6791 first name in the second list were incorrectly joined into one item (with a
6792 colon in the middle).
6798 1. Added -C and -D options to the exinext utility, mainly to make it easier
6799 to include in the automated testing, but these could be helpful when
6800 multiple configurations are in use.
6802 2. The exinext utility was not formatting the output nicely when there was
6803 an alternate port involved in the retry record key, nor when there was a
6804 message id as well (for retries that were specific to a specific message
6805 and a specific host). It was also confused by IPv6 addresses, because of
6806 the additional colons they contain. I have fixed the IPv4 problem, and
6807 patched it up to do a reasonable job for IPv6.
6809 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6810 delivery, the log line now contains "pipelined" if PIPELINING was used.
6812 4. An SMTP transport process used to panic and die if the bind() call to set
6813 an explicit outgoing interface failed. This has been changed; it is now
6814 treated in the same way as a connect() failure.
6816 5. A reference to $sender_host_name in the part of a conditional expansion
6817 that was being skipped was still causing a DNS lookup. This no longer
6820 6. The def: expansion condition was not recognizing references to header lines
6821 that used bh_ and bheader_.
6823 7. Added the _cache feature to named lists.
6825 8. The code for checking quota_filecount in the appendfile transport was
6826 allowing one more file than it should have been.
6828 9. For compatibility with Sendmail, the command line option
6836 and sets the incoming protocol and host name (for trusted callers). The
6837 host name and its colon can be omitted when only the protocol is to be set.
6838 Note the Exim already has two private options, -pd and -ps, that refer to
6839 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6840 "s", but I don't think that's a major issue.
6842 10. A number of refactoring changes to the code, none of which should affect
6845 (a) The number of logging options was getting close to filling up the
6846 32-bit word that was used as a bit map. I have split them into two classes:
6847 those that are passed in the argument to log_write(), and those that are
6848 only ever tested independently outside of that function. These are now in
6849 separate 32-bit words, so there is plenty of room for expansion again.
6850 There is no change in the user interface or the logging behaviour.
6852 (b) When building, for example, log lines, the code previously used a
6853 macro that called string_cat() twice, in order to add two strings. This is
6854 not really sufficiently general. Furthermore, there was one instance where
6855 it was actually wrong because one of the argument was used twice, and in
6856 one call a function was used. (As it happened, calling the function twice
6857 did not affect the overall behaviour.) The macro has been replaced by a
6858 function that can join an arbitrary number of extra strings onto a growing
6861 (c) The code for expansion conditions now uses a table and a binary chop
6862 instead of a serial search (which was left over from when there were very
6863 few conditions). Also, it now recognizes conditions like "pam" even when
6864 the relevant support is not compiled in: a suitably worded error message is
6865 given if an attempt is made to use such a condition.
6867 11. Added ${time_interval:xxxxx}.
6869 12. A bug was causing one of the ddress fields not to be passed back correctly
6870 from remote delivery subprocesses. The field in question was not being
6871 subsequently used, so this caused to problems in practice.
6873 13. Added new log selectors queue_time and deliver_time.
6875 14. Might have fixed a bug in maildirsizefile handling that threw up
6876 "unexpected character" debug warnings, and recalculated the data
6877 unnecessarily. In any case, I expanded the warning message to give more
6880 15. Added the message "Restricted characters in address" to the statements in
6881 the default ACL that block characters like @ and % in local parts.
6883 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6884 Three changes have been made:
6886 (a) There was a serious bug; a negative response to MAIL caused the whole
6887 recipient domain to be cached as invalid, thereby blocking all messages
6888 to all local parts at the same domain, from all senders. This bug has
6889 been fixed. The domain is no longer cached after a negative response to
6890 MAIL if the sender used is not empty.
6892 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6895 (c) A new callout option, "use_sender" has been added for people who want
6896 the modified behaviour.
6902 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6905 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6906 indeed breaks things for older releases.
6908 3. Added additional logging to the case where there is a problem reading data
6909 from a filter that is running in a subprocess using a pipe, in order to
6910 try to track down a specific problem.
6912 4. Testing facility fudge: when running in the test harness and attempting
6913 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6914 getting "No route to host". Convert this to a timeout.
6916 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6919 6. Some OS don't have socklen_t but use size_t instead. This affects the
6920 fifth argument of getsockopt() amongst other things. This is now
6921 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6922 can be set for individual OS. I have set it for SunOS5, OSF1, and
6923 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6924 some earlier ones do not.
6926 7. Change 4.30/15 was not doing the test caselessly.
6928 8. The standard form for an IPv6 address literal was being rejected by address
6929 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6930 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6931 this, as well as the form without the "IPv6" on the front (but only when
6932 address literals are enabled, of course).
6934 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6936 10. Exim crashed if a message with an empty sender address specified by -f
6937 encountered a router with an errors_to setting. This could be provoked only
6938 by a command such as
6942 where an empty string was supplied; "<>" did not hit this bug.
6944 11. Installed PCRE release 4.5.
6946 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6947 remained set. It is now erased.
6949 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6950 times from message ids (which are base 36 rather than the normal 62).
6952 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6953 were being counted as actual protocol errors, and logged if the log
6954 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6955 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6956 and DATA following a set of rejected RCPTs do not count as protocol errors.
6957 In other words, Exim assumes they were pipelined, though this may not
6958 actually be the case. Of course, in all cases the client gets an
6959 appropriate error code.
6961 15. If a lookup fails in an ACL condition, a message about the failure may
6962 be available; it is used if testing the ACL cannot continue, because most
6963 such messages specify what the cause of the deferral is. However, some
6964 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6965 that caused an old message to be retained and used if a later statement
6966 caused a defer, replacing the real cause of the deferral.
6968 16. If an IP address had so many PTR records that the DNS lookup buffer
6969 was not large enough to hold them, Exim could crash while trying to process
6970 the truncated data. It now detects and logs this case.
6972 17. Further to 4.21/58, another change has been made: if (and only if) the
6973 first line of a message (the first header line) ends with CRLF, a bare LF
6974 in a subsequent header line has a space inserted after it, so as not to
6975 terminate the header.
6977 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6978 unnecessarily, used atoi() instead of strtol(), and didn't check the
6979 termination when getting file sizes from file names by regex.
6981 19. Completely re-implemented the support for maildirsize files, in the light
6982 of a number of problems with the previous contributed implementation
6983 (4.30/29). In particular:
6985 . If the quota is zero, the maildirsize file is maintained, but no quota is
6988 . If the maildir directory does not exist, it is created before any attempt
6989 to write a maildirsize file.
6991 . The quota value in the file is just a cache; if the quota is changed in
6992 the transport, the new value overrides.
6994 . A regular expression is available for excluding directories from the
6997 20. The autoreply transport checks the characters in options that define the
6998 message's headers; it allows continued headers, but it was checking with
6999 isspace() after an embedded newline instead of explicitly looking for a
7002 21. If all the "regular" hosts to which an address was routed had passed their
7003 expiry times, and had not reached their retry times, the address was
7004 bounced, even if fallback hosts were defined. Now Exim should go on to try
7007 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7008 equivalent code in the SMTP transport. Some hosts send humungous responses
7009 to HELO/EHLO, more than 1024 it seems.
7011 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7012 gives compiler warnings in some environments. I've now done it "properly",
7015 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7016 (because of problems with the built-in one) was declared to return uschar *
7017 instead of char *, causing compiler failure.
7019 25. Fixed a file descriptor leak when processing alias/forward files.
7021 26. Fixed a minor format string issue in dbfn.c.
7023 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7025 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7026 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7029 29. When a sender address is verified, it is cached, to save repeating the test
7030 when there is more than one recipient in a message. However, when the
7031 verification involves a callout, it is possible for different callout
7032 options to be set for different recipients. It is too complicated to keep
7033 track of this in the cache, so now Exim always runs a verification when a
7034 callout is required, relying on the callout cache for the optimization.
7035 The overhead is duplication of the address routing, but this should not be
7038 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7039 to be verified with callout=postmaster, and the main callout worked but the
7040 postmaster check failed, the verification correctly failed. However, if a
7041 subsequent RCPT command asked for sender verification *without* the
7042 postmaster check, incorrect caching caused this verification also to fail,
7045 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7046 it was not caching the DNS options (qualify_single, search_parents) that
7047 were used when the lookup failed. A subsequent lookup with different
7048 options therefore always gave the same answer, though there were cases
7049 where it should not have. (Example: a "domains = !$mx_any" option on a
7050 dnslookup router: the "domains" option is always processed without any
7051 widening, but the router might have qualify_single set.) Now Exim uses the
7052 cached value only when the same options are set.
7054 32. Added John Jetmore's "exipick" utility to the distribution.
7056 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7057 than a timeout (e.g. a certificate is required, and is not provided), an
7058 Exim server now closes the connection immediately. Previously it waited for
7059 the client to close - but if the client is SSL, it seems that they each
7060 wait for each other, leading to a delay before one of them times out.
7062 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7063 maintained 0.8.x compatibility because I don't think many are using it, and
7064 it is clearly obsolete.
7066 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7069 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7070 client certificate was expired. A simple patch fixes this, though I don't
7071 understand the full logic of why the verify callback is called multiple
7074 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7075 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7076 which causes problems with some clients (such as the Certicom SSL Plus
7077 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7078 disables the coutermeasure allowing Eudora to connect."
7080 38. Exim was not checking that a write() to a log file succeeded. This could
7081 lead to Bad Things if a log got too big, in particular if it hit a file
7082 size limit. Exim now panics and dies if it cannot write to a log file, just
7083 as it does if it cannot open a log file.
7085 39. Modified OS/Makefile-Linux so that it now contains
7087 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7089 The two -D definitions ensure that Exim is compiled with large file
7090 support, which makes it possible to handle log files that are bigger than
7093 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7094 instance) a domain was checked against a named list that involved a lookup,
7095 causing $domain_data to be set, then another domain was checked against the
7096 same list, then the first domain was re-checked, the value of $domain_data
7097 after the final check could be wrong. In particular, if the second check
7098 failed, it could be set empty. This bug probably also applied to
7101 41. The strip_trailing_dot option was not being applied to the address given
7102 with the -f command-line option.
7104 42. The code for reading a message's header from the spool was incrementing
7105 $received_count, but never initializing it. This meant that the value was
7106 incorrect (doubled) while delivering a message in the same process in which
7107 it was received. In the most common configuration of Exim, this never
7108 happens - a fresh exec is done - but it can happen when
7109 deliver_drop_privilege is set.
7111 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7112 it now includes up to 150 characters of the unexpected data in the log
7115 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7116 and building data strings. The size of both of these buffers was 10 000
7117 bytes - far larger than anybody would *ever* want, thought I. Needless to
7118 say, somebody hit the limit. I have increased the maximum line length to
7119 20 000 and the maximum data length of concatenated lines to 100 000. I have
7120 also fixed two bugs, because there was no checking on these buffers. Tsk,
7121 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7122 buffer is too small.
7124 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7125 lsearch lookups. Now it does.
7127 46. When parsing a route_list item in a manualroute router, a fixed-length
7128 buffer was used for the list of hosts. I made this 1024 bytes long,
7129 thinking that nobody would ever have a list of hosts that long. Wrong.
7130 Somebody had a whole pile of complicated expansion conditions, and the
7131 string was silently truncated, leading to an expansion error. It turns out
7132 that it is easier to change to an unlimited length (owing to other changes
7133 that have happened since this code was originally written) than to build
7134 structure for giving a limitation error. The length of the item that
7135 expands into the list of hosts is now unlimited.
7137 47. The lsearch lookup could not handle data where the length of text line was
7138 more than 4095 characters. Such lines were truncated, leading to shortened
7139 data being returned. It should now handle lines of any length.
7141 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7142 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7145 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7146 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7147 no longer happens in comment lines. A list of replacements is now placed
7148 at the head of all of the source files, except those whose only change is
7149 to replace PERL_COMMAND in the very first #! line.
7151 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7152 messages on the queue, with a bottom-up merge sort, using code contributed
7153 by Michael Haardt. This should make operations like -bp somewhat faster on
7154 large queues. It won't affect queue runners, except when queue_run_in_order
7157 51. Installed eximstats 1.31 in the distribution.
7159 52. Added support for SRV lookups to the dnslookup router.
7161 53. If an ACL referred to $message_body or $message_body_end, the value was not
7162 reset for any messages that followed in the same SMTP session.
7164 54. The store-handling optimization for building very long strings was not
7165 differentiating between the different store pools. I don't think this
7166 actually made any difference in practice, but I've tidied it.
7168 55. While running the routers to verify a sender address, $sender_address
7169 was still set to the sender address. This is wrong, because when routing to
7170 send a bounce to the sender, it would be empty. Therefore, I have changed
7171 it so that, while verifying a sender address, $sender_address is set to <>.
7172 (There is no change to what happens when verifying a recipient address.)
7174 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7175 target A or AAAA records (if not already returned) without resetting the
7176 qualify_single or search_parents options of the DNS resolver. These are
7177 inappropriate in this case because the targets of MX and SRV records must
7178 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7179 target that, when qualified, matched something in the local domain. These
7180 two options are now turned off when doing these lookups.
7182 57. It seems that at least some releases of Reiserfs (which does not have the
7183 concept of a fixed number of inodes) returns zero and not -1 for the
7184 number of available inodes. This interacted badly with check_spool_inodes,
7185 which assumed that -1 was the "no such thing" setting. What I have done is
7186 to check that the total number of inodes is greater than zero before doing
7187 the test of how many are available.
7189 58. When a "warn" ACL statement has a log_message modifier, the message is
7190 remembered, and not repeated. This is to avoid a lot of repetition when a
7191 message has many recipients that cause the same warning to be written.
7192 However, Exim was preserving the list of already written lines for an
7193 entire SMTP session, which doesn't seem right. The memory is now reset if a
7194 new message is started.
7196 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7197 debugging output unless log_rewrite was also set.
7199 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7200 of (int)(handle) when we know that handle contains (void *)(-1).
7202 61. The Exim daemon panic-logs an error return when it closes the incoming
7203 connection. However "connection reset by peer" seems to be common, and
7204 isn't really an error worthy of noting specially, so that particular error
7207 62. When Exim is trying to find all the local interfaces, it used to panic and
7208 die if the ioctl to get the interface flags failed. However, it seems that
7209 on at least one OS (Solaris 9) it is possible to have an interface that is
7210 included in the list of interfaces, but for which you get a failure error
7211 for this call. This happens when the interface is not "plumbed" into a
7212 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7213 failure of the "get flags" call assumes that the interface is down.
7215 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7216 makes life easier for people who are doing arithmetic on fields extracted
7217 from dates, where you often get leading zeros that should not be
7218 interpreted as octal.
7220 64. Added qualify_domain to the redirect router, to override the global
7223 65. If a pathologically long header line contained very many addresses (the
7224 report of this problem mentioned 10 000) and each of them was rewritten,
7225 Exim could use up a very large amount of memory. (It kept on making new
7226 copies of the header line as it rewrote, and never released the old ones.)
7227 At the expense of a bit more processing, the header rewriting function has
7228 been changed so that it no longer eats memory in this way.
7230 66. The generation of the Received: header has been moved from the time that a
7231 message starts to be received, to the time that it finishes. The timestamp
7232 in the Received: header should now be very close to that of the <= log
7233 line. There are two side-effects of this change:
7235 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7236 logged header lines no longer include the local Received: line, because
7237 it has not yet been created. The same applies to a copy of the message
7238 that is returned to a non-SMTP sender when a message is rejected.
7240 (b) When a filter file is tested using -bf, no additional Received: header
7241 is added to the test message. After some thought, I decided that this
7244 This change does not affect the value of $received_for. It is still set
7245 after address rewriting, but before local_scan() is called.
7247 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7249 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7250 gave an unhelpful panic error message, and a defer error. I have managed to
7251 change this behaviour so that it now rejects any supplied certificate,
7252 which seems right, as the list of acceptable certificates is empty.
7254 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7255 gave an unhelpful defer error. I have not managed to make this reject any
7256 supplied certificates, but the error message it gives is "no certificate
7257 supplied", which is not helpful.
7259 70. exigrep's output now also includes lines that are not associated with any
7260 message, but which match the given pattern. Implemented by a patch from
7261 Martin Sluka, which also tidied up the Perl a bit.
7263 71. Recipient callout verification, like sender verification, was using <> in
7264 the MAIL FROM command. This isn't really the right thing, since the actual
7265 sender may affect whether the remote host accepts the recipient or not. I
7266 have changed it to use the actual sender in the callout; this means that
7267 the cache record is now keyed on a recipient/sender pair, not just the
7268 recipient address. There doesn't seem to be a real danger of callout loops,
7269 since a callout by the remote host to check the sender would use <>.
7270 [SEE ABOVE: changed after hitting problems.]
7272 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7273 temporary errors. However, in the case of such a code being given after
7274 the end of a data transmission (i.e. after ".") Exim was failing to write
7275 a retry record for the message. (Yes, there was some broken host that was
7276 actually sending 8xx at this point.)
7278 73. An unknown lookup type in a host list could cause Exim to panic-die when
7279 the list was checked. (An example that provoked this was putting <; in the
7280 middle of a list instead of at the start.) If this happened during a DATA
7281 ACL check, a -D file could be left lying around. This kind of configuration
7282 error no longer causes Exim to die; instead it causes a defer error. The
7283 incident is still logged to the main and panic logs.
7285 74. Buglet left over from Exim 3 conversion. The message "too many messages
7286 in one connection" was written to the rejectlog but not the mainlog, except
7287 when address rewriting (yes!) was being logged.
7289 75. Added write_rejectlog option.
7291 76. When a system filter was run not as root (that is, when system_filter_user
7292 was set), the values of the $n variables were not being returned to the
7293 main process; thus, they were not subsequently available in the $sn
7296 77. Added +return_path_on_delivery log selector.
7298 78. A connection timeout was being treated differently from recipients deferred
7299 when testing hosts_max_try with a message that was older than the host's
7300 retry timeout. (The host should not be counted, thus allowing all hosts to
7301 be tried at least once before bouncing.) This may have been the cause of an
7302 occasionally reported bug whereby a message would remain on the queue
7303 longer than the retry timeout, but would be bounced if a delivery was
7304 forced. I say "may" because I never totally pinned down the problem;
7305 setting up timeout/retry tests is difficult. See also the next item.
7307 79. The ultimate address timeout was not being applied to errors that involved
7308 a combination of host plus message (for example, a timeout on a MAIL
7309 command). When an address resolved to a number of possible hosts, and they
7310 were not all tried for each delivery (e.g. because of hosts_max_try), a
7311 message could remain on the queue longer than the retry timeout.
7313 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7316 81. Fixed an obscure SMTP outgoing bug which required at least the following
7317 conditions: (a) there was another message waiting for the same server;
7318 (b) the server returned 5xx to all RCPT commands in the first message so
7319 that the message was not completed; (c) the server dropped the connection
7320 or gave a negative response to the RSET that Exim sends to abort the
7321 transaction. The observed case was a dropped connection after DATA that had
7322 been sent in pipelining mode. That is, the server had advertised PIPELINING
7323 but was not implementing it correctly. The effect of the bug was incorrect
7324 behaviour, such as trying another host, and this could lead to a crash.
7330 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7331 and daemon.c were passed as pointers to ints; they should have been
7332 pointers to socklen_t variables (which are typically unsigned ints).
7334 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7337 3. Fixed a really odd bug that affected only the testing scheme; patching a
7338 certain fixed string in the binary changed the value of another string that
7339 happened to be identical to the end of the original first string.
7341 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7342 name", it returns that address as the IP address. On some operating
7343 systems (e.g. Solaris), it also passes back the IP address string as the
7344 "host name". However, on others (e.g. Linux), it passes back an empty
7345 string. Exim wasn't checking for this, and was changing the host name to an
7346 empty string, assuming it had been canonicalized.
7348 5. Although rare, it is permitted to have more than one PTR record for a given
7349 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7350 all the names associated with an address, because they do in Solaris.
7351 However, it seems that they do not in Linux for data that comes from the
7352 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7353 I found this out when I moved to a new Linux workstation and tried to run
7354 the Exim test suite.
7356 To get round this problem I have changed the code so that it now does its
7357 own call to the DNS to look up PTR records when searching for a host name.
7358 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7359 addresses that are only in /etc/hosts are still found.
7361 This behaviour is, however, controlled by an option called host_lookup_
7362 order, which defaults to "bydns:byaddr". If people want to use the other
7363 order, or indeed, just use one or the other means of lookup, they can
7364 specify it in this variable.
7366 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7367 some operating systems, this comes back from gethostbyaddr() as an empty
7368 string, and this is what Exim used to test for. However, it seems that in
7369 other systems, "." is yielded. Exim now tests for this case too.
7371 7. The values of check_spool_space and check_log_space are now held internally
7372 as a number of kilobytes instead of an absolute number of bytes. If a
7373 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7374 kilobyte. This means that much larger values can be stored.
7376 8. Exim monitor: an attempt to get the action menu when not actually pointing
7377 at a message produces an empty menu entitled "No message selected". This
7378 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7379 no entries in it ("Shell widget menu has zero width and/or height"). So I
7380 have added a single, blank menu entry in this case.
7382 9. Added ${quote_local_part.
7384 10. MIME decoding is now applied to the contents of Subject: header lines when
7387 11. Now that a reference to $sender_host_address automatically causes a reverse
7388 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7389 host lookup before query-style lookups in lists that might use this
7390 variable. This has therefore been abolished, and the "net-" prefix is no
7391 longer necessary for query-style lookups.
7393 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7394 have been a typo for LFLAGS, so it has been changed.
7396 13. The install script calls Exim with "-C /dev/null" in order to find the
7397 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7398 to be output. However, since Exim outputs its version number before the
7399 error, it didn't break the script. It just looked ugly. I fixed this by
7400 always allowing "-C /dev/null" if the caller is root.
7402 14. Ignore overlarge ACL variable number when reading spool file - insurance
7403 against a later release with more variables having written the file.
7405 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7406 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7407 this, as well as the form without the "IPv6" on the front.
7409 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7410 OS/Makefile-Darwin file.
7412 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7413 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7415 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7416 variable. However, some OS use ints for inodes. I've added cast to long int
7417 to get rid of the compiler warning.
7419 19. I had forgotten to lock out "/../" in configuration file names when
7420 ALT_CONFIG_PREFIX was set.
7422 20. Routers used for verification do not need to specify transports. However,
7423 if such a router generated a host list, and callout was configured, Exim
7424 crashed, because it could not find a port number from the (non-existent)
7425 transport. It now assumes port 25 in this circumstance.
7427 21. Added the -t option to exigrep.
7429 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7430 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7431 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7432 them is set, LOOKUP_LSEARCH is forced.
7434 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7436 24. Added sender and host information to the "rejected by local_scan()" log
7437 line; previously there was no indication of these.
7439 25. Added .include_if_exists.
7441 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7442 Linux. It turns out that not all file systems support this. Apparently some
7443 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7444 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7445 not supported on the file descriptor, is now ignored when Exim is trying to
7446 sync a directory. This applies only to Linux.
7448 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7450 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7451 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7454 29. Added support for maildirsize files from supplied patch (modified a bit).
7456 30. The use of :fail: followed by an empty string could lead Exim to respond to
7457 sender verification failures with (e.g.):
7459 550 Verification failed for <xxx>
7460 550 Sender verify failed
7462 where the first response line was missing the '-' that indicates it is not
7463 the final line of the response.
7465 31. The loop for finding the name of the user that called Exim had a hardwired
7466 limit of 10; it now uses the value of finduser_retries, which is used for
7467 all other user lookups.
7469 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7472 33. Exim was neglecting to zero errno before one call of strtol() when
7473 expanding a string and expecting an integer value. On some systems this
7474 resulted in spurious "integer overflow" errors. Also, it was casting the
7475 result into an int without checking.
7477 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7478 did not work. The code looks as if it has *never* worked, though it appears
7479 to have been documented since at least release 1.62. I have made it work.
7481 35. The "timeout_DNS" error in retry rules, also documented since at least
7482 1.62, also never worked. As it isn't clear exactly what this means, and
7483 clearly it isn't a major issue, I have abolished the feature by treating it
7484 as "timeout", and writing a warning to the main and panic logs.
7486 36. The display of retry rules for -brt wasn't always showing the error code
7489 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7490 timeout_connect_A, timeout_connect_MX.
7492 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7493 to the empty sender.
7495 39. The daemon was not analysing the content of -oX till after it had closed
7496 stderr and disconnected from the controlling terminal. This meant that any
7497 syntax errors were only noted on the panic log, and the return code from
7498 the command was 0. By re-arranging the code a little, I've made the
7499 decoding happen first, so such errors now appear on stderr, and the return
7500 code is 1. However, the actual setting up of the sockets still happens in
7501 the disconnected process, so errors there are still only recorded on the
7504 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7505 connections (as happens on some IP stacks) was logged at start up time as
7506 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7507 it from "IPv6 and IPv4", which means that two separate sockets are being
7510 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7511 says whether AF_INET or AF_INET6 was passed as an argument.
7513 42. Exiwhat output was messed up when time zones were included in log
7516 43. Exiwhat now gives more information about the daemon's listening ports,
7517 and whether -tls-on-connect was used.
7519 44. The "port" option of the smtp transport is now expanded.
7521 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7522 silently ignored. Now an error message is written to the main and panic
7525 46. There's a new ACL modifier called "logwrite" which writes to a log file
7526 as soon as it is encountered.
7528 47. Added $local_user_uid and $local_user_gid at routing time.
7530 48. Exim crashed when trying to verify a sender address that was being
7533 49. Exim was recognizing only a space character after ".include". It now also
7534 recognizes a tab character.
7536 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7537 extracting the relevant information from the specification. The man page no
7538 longer contains scrambled data for the -d option, and I've added a section
7539 at the front about calling Exim under different names.
7541 51. Added "extra_headers" argument to the "mail" command in filter files.
7543 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7546 53. Installed eximstats 1.29.
7548 54. Added transport_filter_timeout as a generic transport option.
7550 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7551 Cc: header lines. This was required by RFC 822, but it not required by RFC
7554 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7555 incoming messages that did not have them. Now it does so only if the
7556 message originates locally, that is, if there is no associated remote host
7557 address. When Resent- header lines are present, this applies to the Resent-
7558 lines rather than the non-Resent- lines.
7560 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7561 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7563 58. Messages for configuration errors now include the name of the main
7564 configuration file - useful now that there may be more than one file in a
7565 list (.included file names were always shown).
7567 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7568 for those rare installations that do not start the daemon as root or run it
7569 setuid root. I've cut out the call to initgroups() if the daemon is not
7572 60. The Exim user and group can now be bound into the binary as text strings
7573 that are looked up at the start of Exim's processing.
7575 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7577 62. Added $mailstore_basename variable.
7579 63. Installed patch to sieve.c from Michael Haardt.
7581 64. When Exim failed to open the panic log after failing to open the main log,
7582 the original message it was trying to log was written to stderr and debug
7583 output, but if they were not available (the usual case in production), it
7584 was lost. Now it is written to syslog before the two lines that record the
7585 failures to open the logs.
7587 65. Users' Exim filters run in subprocesses under the user's uid. It is
7588 possible for a "deliver" command or an alias in a "personal" command to
7589 provoke an address rewrite. If logging of address rewriting is configured,
7590 this fails because the process is not running as root or exim. There may be
7591 a better way of dealing with this, but for the moment (because 4.30 needs
7592 to be released), I have disabled address rewrite logging when running a
7593 filter in a non-root, non-exim process.
7599 1. The buildconfig auxiliary program wasn't quoting the value set for
7600 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7601 not defined. This bug was masked in 4.22 by the effect that was fixed in
7604 2. Some messages that were rejected after a message id was allocated were
7605 shown as "incomplete" by exigrep. It no longer does this for messages that
7606 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7608 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7609 have allow_domain_literals set, the ID did not get logged in the <= line.
7610 Domain literals are now always recognized in Message-ID: header lines.
7612 4. The first argument for a ${extract expansion item is the key name or field
7613 number. Leading and trailing spaces in this item were not being ignored,
7614 causing some misleading effects.
7616 5. When deliver_drop_privilege was set, single queue runner processes started
7617 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7618 same command in the process it spins off) were not dropping privilege.
7620 6. When the daemon running as "exim" started a queue runner, it always
7621 re-executed Exim in the spun-off process. This is a waste of effort when
7622 deliver_drop_privilege is set. The new process now just calls the
7623 queue-runner function directly.
7629 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7632 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7633 ignored. Though the use of -oP was forcing the writing of a pid file, it
7634 was always written to the default place.
7636 3. If the message "no IP address found for host xxxx" is generated during
7637 incoming verification, it is now followed by identification of the incoming
7638 connection (so you can more easily find what provoked it).
7640 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7642 5. Added some features to "harden" Exim a bit more against certain attacks:
7644 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7645 be put in Local/Makefile. This is like the never_users runtime option,
7646 but it cannot be overridden. The default setting is "root".
7648 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7649 prefix string with which any file named in a -C command line option
7652 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7653 is retained for -C and -D only if the caller of Exim is root. Without
7654 it, the exim user may also use -C and -D and retain privilege.
7656 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7657 command line option is disabled.
7659 6. Macro names set by the -D option must start with an upper case letter, just
7660 like macro names defined in the configuration file.
7662 7. Added "dereference=" facility to LDAP.
7664 8. Two instances of the typo "uknown" in the source files are fixed.
7666 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7667 the Configure-Makefile script screwed up while processing it.
7669 10. Incorporated PCRE 4.4.
7671 11. The SMTP synchronization check was not operating right at the start of an
7672 SMTP session. For example, it could not catch a HELO sent before the client
7673 waited for the greeting. There is now a check for outstanding input at the
7674 point when the greeting is written. Because of the duplex, asynchronous
7675 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7676 way, but not yet received, when the check is performed.
7678 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7679 on TCP/IP sockets, because this apparently causes some broken clients to
7682 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7683 unchanged) from the Cygwin maintainer.
7685 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7686 was supported instead of testing for mbx. Effectively a typo.
7688 15. The spa authenticator server code was not checking that the input it
7689 received was valid base64.
7691 16. The debug output line for the "set" modifier in ACLs was not showing the
7692 name of the variable that was being set.
7694 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7696 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7697 lookup on its own. Something else was needed to trigger the lookup. For
7698 example, a match in host_lookup or the need for a host name in a host list.
7699 Now, if $sender_host_name is referenced and the host name has not yet been
7700 looked up, a lookup is performed. If the lookup fails, the variable remains
7701 empty, and $host_lookup_failed is set to "1".
7703 19. Added "eqi" as a case-independent comparison operator.
7705 20. The saslauthd authentication condition could segfault if neither service
7706 nor realm was specified.
7708 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7709 error message that was logged was misleading, and incoming SMTP
7710 connections were dropped. The message is now more accurate, and temporary
7711 errors are given to SMTP connections.
7713 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7714 (except RSET) with a 421 error, until QUIT is received. However, it was
7715 failing to send a response to QUIT.
7717 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7718 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7719 ACL is now run after the helo_try_verify_hosts code.
7721 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7722 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7723 case-independent, so other case variants are also recognized). Apparently
7724 some systems use these upper case variants.
7726 25. If more than two messages were waiting for the same host, and a transport
7727 filter was specified for the transport, Exim sent two messages over the
7728 same TCP/IP connection, and then failed with "socket operation on non-
7729 socket" when it tried to send the third.
7731 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7733 27. The extern definition of crypt16() in expand.c was not being excluded when
7734 the OS had its own crypt16() function.
7736 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7737 as a preferred synonym for return_size_limit, both as an option and as an
7740 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7742 30. Changed the default configuration ACL to relax the local part checking rule
7743 for addresses that are not in any local domains. For these addresses,
7744 slashes and pipe symbols are allowed within local parts, but the sequence
7745 /../ is explicitly forbidden.
7747 31. SPA server authentication was not clearing the challenge buffer before
7750 32. log_message in a "warn" ACL statement was writing to the reject log as
7751 well as to the main log, which contradicts the documentation and doesn't
7752 seem right (because no rejection is happening). So I have stopped it.
7754 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7755 However, I am unable to do any testing of this.
7757 34. Fixed an infelicity in the appendfile transport. When checking directories
7758 for a mailbox, to see if any needed to be created, it was accidentally
7759 using path names with one or more superfluous leading slashes; tracing
7760 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7762 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7763 contents are added to the log line that is written for every discarded
7764 recipient. (Previously a log_message setting was ignored.)
7766 36. The ${quote: operator now quotes the string if it is empty.
7768 37. The install script runs exim in order to find its version number. If for
7769 some reason other than non-existence or emptiness, which it checks, it
7770 could not run './exim', it was installing it with an empty version number,
7771 i.e. as "exim-". This error state is now caught, and the installation is
7774 38. An argument was missing from the function that creates an error message
7775 when Exim fails to connect to the socket for saslauthd authentication.
7776 This could cause Exim to crash, or give a corrupted message.
7778 39. Added isip, isip4, and isip6 to ${if conditions.
7780 40. The ACL variables $acl_xx are now saved with the message, and can be
7781 accessed later in routers, transports, and filters.
7783 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7784 strings in the file are not string-expanded.
7786 42. If a MAIL command specified a SIZE value that was too large to fit into an
7787 int variable, the check against message_size_limit failed. Such values are
7788 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7789 day this will have to be increased, but I don't think I want to be around
7790 when emails are that large.
7797 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7798 iconv() is not standard in FreeBSD.
7800 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7801 IPv6 enabled. The observed symptom was a segmentation fault on return from
7802 the function os_common_find_running_interfaces() in src/os.c.
7804 3. In the check_special_case() function in daemon.c I had used "errno" as an
7805 argument name, which causes warnings on some systems. This was basically a
7806 typo, since it was named "eno" in the comments!
7808 4. The code that waits for the clock to tick (at a resolution of some fraction
7809 of a second) so as to ensure message-id uniqueness was always waiting for
7810 at least one whole tick, when it could have waited for less. [This is
7811 almost certainly not relevant at current processor speeds, where it is
7812 unlikely to ever wait at all. But we try to future-proof.]
7814 5. The function that sleeps for a time interval that includes fractions of a
7815 second contained a race. It did not block SIGALRM between setting the
7816 timer, and suspending (a couple of lines later). If the interval was short
7817 and the sigsuspend() was delayed until after it had expired, the suspension
7818 never ended. On busy systems this could lead to processes getting stuck for
7821 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7822 process, before it forks any delivery processes. The open lookup caching
7823 mechanism meant that the open file or database connection was passed into
7824 the delivery process. The problem was that delivery processes always tidy
7825 up cached lookup data. This could cause a problem for the next delivery
7826 process started by the queue runner, because the external queue runner
7827 process does not know about the closure. So the next delivery process
7828 still has data in the lookup cache. In the case of a file lookup, there was
7829 no problem because closing a file descriptor in a subprocess doesn't affect
7830 the parent. However, if the lookup was caching a connection to a database,
7831 the connection was closed, and the second delivery process was likely to
7832 see errors such as "PGSQL: query failed: server closed the connection
7833 unexpectedly". The problem has been fixed by closing all cached lookups
7834 in a queue runner before running a delivery process.
7836 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7837 seem to have the "const" qualifier which it has on other OS. I've
7840 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7841 *of the same type* (client or server) with the same public name that an
7842 error should be diagnosed.
7844 9. When Exim looked up a host name for an IP address, but failed to find the
7845 original IP address when looking up the host name (a safety check), it
7846 output the message "<ip address> does not match any IP for NULL", which was
7847 confusing, to say the least. The bug was that the host name should have
7848 appeared instead of "NULL".
7850 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7851 user that is built into the binary, and the -C or -D options is used, root
7852 privilege is dropped before the configuration file is read. In addition,
7853 logging is switched to stderr instead of the normal log files. If the
7854 configuration then re-defines the Exim user, the unprivileged environment
7855 is probably not what is expected, so Exim logs a panic warning message (but
7858 However, if deliver_drop_privilege is set, the unprivileged state may well
7859 be exactly what is intended, so the warning has been cut out in that case,
7860 and Exim is allowed to try to write to its normal log files.
7866 1. smtp_return_error_details was not giving details for temporary sender
7867 or receiver verification errors.
7869 2. Diagnose a configuration error if two authenticators have the same public
7872 3. Exim used not to create the message log file for a message until the first
7873 delivery attempt. This could be confusing when incoming messages were held
7874 for policy or load reasons. The message log file is now created at the time
7875 the message is received, and an initial "Received" line is written to it.
7877 4. The automatically generated man page for command line options had a minor
7878 bug that caused no ill effects; however, a more serious problem was that
7879 the procedure for building the man page automatically didn't always
7880 operate. Consequently, release 4.20 contains an out-of-date version. This
7881 shouldn't happen again.
7883 5. When building Exim with embedded Perl support, the script that builds the
7884 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7885 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7887 6. The freeze_tell option was not being used for messages that were frozen on
7888 arrival, either by an ACL or by local_scan().
7890 7. Added the smtp_incomplete_transaction log selector.
7892 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7893 was accepting AUTH without a new EHLO.
7895 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7896 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7899 10. Logging of TCP/IP connections (when configured) now happens in the main
7900 daemon process instead of the child process, so that the TCP/IP connection
7901 count is more accurate (but it can never be perfect).
7903 11. The use of "drop" in a nested ACL was not being handled correctly in the
7904 outer ACL. Now, if condition failure induced by the nested "drop" causes
7905 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7906 or "require"), the connection is dropped.
7908 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7909 that yield "discard" can now be used with an "accept" or a "discard" verb,
7910 but an error is generated for any others (because I can't see a useful way
7911 to define what should happen).
7913 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7914 are now processed in the same way as lines in the Exim configuration file.
7915 In particular, continuation lines are supported.
7917 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7919 15. Added -ti meaning -t -i.
7921 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7922 domains, and warn by logging if others are found.
7924 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7925 returned from GIFCONF when Exim is trying to find the list of interfaces on
7926 a host. The code in os.c has been modified to copy each ifreq to an aligned
7927 structure in all cases.
7929 Also, in some cases, the returned ifreq's were being copied to a 'struct
7930 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7931 means the last couple of bytes of an IPv6 address could be chopped if the
7932 ifreq contained only a normal sockaddr (14 bytes storage).
7934 18. Named domain lists were not supported in the hosts_treat_as_local option.
7935 An entry such as +xxxx was not recognized, and was treated as a literal
7938 19. Ensure that header lines added by a DATA ACL are included in the reject log
7939 if the ACL subsequently rejects the message.
7941 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7942 MD5 (which is deprecated).
7944 21. When testing a filter file using -bf, Exim was writing a message when it
7945 took the sender from a "From " line in the message, but it was not doing so
7946 when it took $return_path from a Return-Path: header line. It now does.
7948 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7949 with a valid header line field name (a series of printing characters
7950 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7952 23. Changed "disc" in the source to "disk" to conform to the documentation and
7953 the book and for uniformity.
7955 24. Ignore Sendmail's -Ooption=value command line item.
7957 25. When execve() failed while trying to run a command in a pipe transport,
7958 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7959 could be confused with a return value of 69 from the command itself. This
7960 has been changed to 127, the value the shell returns if it is asked to run
7961 a non-existent command. The wording for the related log line suggests a
7962 non-existent command as the problem.
7964 26. If received_header_text expands to an empty string, do not add a Received:
7965 header line to the message. (Well, it adds a token one on the spool, but
7966 marks it "old" so that it doesn't get used or transmitted.)
7968 27. Installed eximstats 1.28 (addition of -nt option).
7970 28. There was no check for failure on the call to getsockname() in the daemon
7971 code. This can fail if there is a shortage of resources on the system, with
7972 ENOMEM, for example. A temporary error is now given on failure.
7974 29. Contrary to the C standard, it seems that in some environments, the
7975 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7976 program. Exim now does this explicitly; it affects the formatting of
7977 timestamps using strftime().
7979 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7980 complaints. I've now initialized all the variables, to avoid this.
7982 32. Header lines added by a system filter were not being "seen" during
7983 transport-time rewrites.
7985 33. The info_callback() function passed to OpenSSL is set up with type void
7986 (*)(SSL *, int, int), as described somewhere. However, when calling the
7987 function (actually a macro) that sets it up, the type void(*)() is
7988 expected. I've put in a cast to prevent warnings from picky compilers.
7990 34. If a DNS black list lookup found a CNAME record, but there were no A
7991 records associated with the domain it pointed at, Exim crashed.
7993 35. If a DNS black list lookup returned more than one A record, Exim ignored
7994 all but the first. It now scans all returned addresses if a particular IP
7995 value is being sought. In this situation, the contents of the
7996 $dnslist_value variable are a list of all the addresses, separated by a
7999 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8000 to accept a host name and all its aliases if the forward lookup for any of
8001 them yielded the IP address of the incoming connection. Now it accepts only
8002 those names whose forward lookup yields the correct IP address. Any other
8003 names are discarded. This closes a loophole whereby a rogue DNS
8004 administrator could create reverse DNS records to break through a
8005 wildcarded host restriction in an ACL.
8007 37. If a user filter or a system filter that ran in a subprocess used any of
8008 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8009 the wrong values were passed to the pipe command ($thisaddress had the
8010 value of $0, $0 had the value of $1, etc). This bug was introduced by
8011 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8013 38. Improved the line breaking for long SMTP error messages from ACLs.
8014 Previously, if there was no break point between 40 and 75 characters, Exim
8015 left the rest of the message alone. Two changes have been made: (a) I've
8016 reduced the minimum length to 35 characters; (b) if it can't find a break
8017 point between 35 and 75 characters, it looks ahead and uses the first one
8018 that it finds. This may give the occasional overlong line, but at least the
8019 remaining text gets split now.
8021 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8022 file descriptors might be low, and that setting 1000 would always raise it.
8023 It turns out that in some environments, the limit is already over 1000 and
8024 that lowering it causes trouble. So now Exim takes care not to decrease it.
8026 40. When delivering a message, the value of $return_path is set to $sender_
8027 address at the start of routing (routers may change the value). By an
8028 oversight, this default was not being set up when an address was tested by
8029 -bt or -bv, which affected the outcome if any router or filter referred to
8032 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8033 log only when "message" or "log_message" are set. However, if one of the
8034 conditions was an address verification, or a call to a nested ACL, the
8035 messages generated by the underlying test were being passed through. This
8036 no longer happens. The underlying message is available in $acl_verify_
8037 message for both "message" and "log_message" expansions, so it can be
8038 passed through if needed.
8040 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8041 new expansion $bh_ to give the encoded byte string without charset
8042 translation. Translation happens only if iconv() is available; HAVE_ICONV
8043 indicates this at build time. HEADERS_CHARSET gives the charset to
8044 translate to; headers_charset can change it in the configuration, and
8045 "headers charset" can change it in an individual filter file.
8047 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8048 that creates RFC 2047 encoded "words" labels them as that charset instead
8049 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8050 expansion operator; (ii) when Exim creates a From: line for a local
8051 message; (iii) when a header line is rewritten to include a "phrase" part.
8053 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8054 buggy, causing it to skip the first lines of messages whose message ID
8055 ended in 'D'. This would not have bitten before Exim release 4.14, because
8056 message IDs were unlikely to end in 'D' before then. The effect was to have
8057 incorrect size information for certain domains.
8059 45. #include "config.h" was missing at the start of the crypt16.c module. This
8060 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8063 46. If there was a timeout during a "random" callout check, Exim treated it as
8064 a failure of the random address, and carried on sending RSET and the real
8065 address. If the delay was just some slowness somewhere, the response to the
8066 original RCPT would be taken as a response to RSET and so on, causing
8067 mayhem of various kinds.
8069 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8070 when I implemented it. It didn't allow for the fact that some option values
8071 may legitimately be negative (e.g. size_addition), and it didn't even do
8072 the right test for positive values.
8074 48. Domain names in DNS records are case-independent. Exim always looks them up
8075 in lower case. Some resolvers return domain names in exactly the case they
8076 appear in the zone file, that is, they may contain uppercase letters. Not
8077 all resolvers do this - some return always lower case. Exim was treating a
8078 change of case by a resolver as a change of domain, similar to a widening
8079 of a domain abbreviation. This triggered its re-routing code and so it was
8080 trying to route what was effectively the same domain again. This normally
8081 caused routing to fail (because the router wouldn't handle the domain
8082 twice). Now Exim checks for this case specially, and just changes the
8083 casing of the domain that it ultimately uses when it transmits the message
8086 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8089 50. If a filter generated a file delivery with a non-absolute name (possible if
8090 no home directory exists for the router), the forbid_file option was not
8093 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8094 the existing equality matching.
8096 52. Exim was using ints instead of ino_t variables in some places where it was
8097 dealing with inode numbers.
8099 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8100 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8101 called TMPDIR, and if it finds it is different, it changes its value.
8103 54. The smtp_printf() function is now made available to local_scan() so
8104 additional output lines can be written before returning. There is also an
8105 smtp_fflush() function to enable the detection of a dropped connection.
8106 The variables smtp_input and smtp_batched_input are exported to
8109 55. Changed the default runtime configuration: the message "Unknown user"
8110 has been removed from the ACL, and instead placed on the localuser router,
8111 using the cannot_route_message feature. This means that any verification
8112 failures that generate their own messages won't get overridden. Similarly,
8113 the "Unrouteable address" message that was in the ACL for unverifiable
8114 relay addresses has also been removed.
8116 56. Added hosts_avoid_esmtp to the smtp transport.
8118 57. The exicyclog script was not checking for the esoteric option
8119 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8120 will work only if exicyclog is run under the appropriate euid.
8122 58. Following a discussion on the list, the rules by which Exim recognises line
8123 endings on incoming messages have been changed. The -dropcr and drop_cr
8124 options are now no-ops, retained only for backwards compatibility. The
8125 following line terminators are recognized: LF CRLF CR. However, special
8126 processing applies to CR:
8128 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8129 nor a local message in the state where . is a terminator.
8131 (ii) If a bare CR is encountered in a header line, an extra space is added
8132 after the line terminator so as not to end the header. The reasoning
8133 behind this is that bare CRs in header lines are most likely either
8134 to be mistakes, or people trying to play silly games.
8136 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8137 was being incorrectly given as 18 bytes larger than it should have been.
8138 This is a VOB (very old bug).
8140 60. This may never have affected anything current, but just in case it has:
8141 When the local host is found other than at the start of a list of hosts,
8142 the local host, those with the same MX, and any that follow, are discarded.
8143 When the list in question was part of a longer list of hosts, the following
8144 hosts (not currently being processed) were also being discarded. This no
8145 longer happens. I'm not sure if this situation could ever has previously
8148 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8151 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8152 header. This is something that is recommended in a new Internet Draft, and
8153 is something that is documented as being done by Sendmail. There are two
8154 possible values. For messages generated by the autoreply transport, Exim
8157 Auto-Submitted: auto-replied
8159 whereas for all other generated messages (e.g. bounces) it adds
8161 Auto-Submitted: auto-generated
8163 63. The "personal" condition in filters now includes a test for the
8164 Auto-Submitted: header. If it contains the string "auto-" the message it
8165 not considered personal.
8167 64. Added rcpt_include_affixes as a generic transport option.
8169 65. Added queue_only_override (default true).
8171 66. Added the syslog_duplication option.
8173 67. If what should have been the first header line of a message consisted of
8174 a space followed by a colon, Exim was mis-interpreting it as a header line.
8175 It isn't of course - it is syntactically invalid and should therefore be
8176 treated as the start of the message body. The misbehaviour could have
8177 caused a number of strange effects, including loss of data in subsequent
8178 header lines, and spool format errors.
8180 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8181 client host had authenticated. This control can now be exercised by an ACL
8182 for more flexibility.
8184 69. By default, callouts do not happen when testing with -bh. There is now a
8185 variant, -bhc, which does actually run the callout code, including
8186 consulting and updating the callout cache.
8188 70. Added support for saslauthd authentication, courtesy of Alexander
8191 71. If statvfs() failed on the spool or log directories while checking their
8192 size for availability, Exim confusingly gave the error "space shortage".
8193 Furthermore, in debugging mode it crashed with a floating point exception.
8194 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8195 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8196 serious problem, Exim now writes to the main and panic logs when this
8197 happens, with details of the failure. It then refuses to accept the
8198 incoming message, giving the message "spool directory problem" or "log
8199 directory problem" with a 421 code for SMTP messages.
8201 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8202 0, 1, and 2 exist, because some OS complain for execs without them (see
8203 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8204 descriptors. However, the code omitted to check that the open succeeded,
8205 causing mysterious errors if for some reason the permissions on /dev/null
8206 got screwed. Now Exim writes a message to the main and panic logs, and
8207 bombs out if it can't open /dev/null.
8209 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8210 interact so that it is all more flexible. It is supposed to remain
8211 backwards compatible. Also added extra_local_interfaces.
8213 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8214 to bomb out with an assertion failure - to the client this appears as a
8215 connection drop. This problem occurs in the part of the code that was taken
8216 from the Samba project. Fortunately, the assertion is in a very simple
8217 function, so I have fixed this by reproducing the function inline in the
8218 one place where it is called, and arranging for authentication to fail
8219 instead of killing the process with assert().
8221 75. The SPA client code was not working when the server requested OEM rather
8222 than Unicode encoding.
8224 76. Added code to make require_files with a specific uid setting more usable in
8225 the case where statting the file as root fails - usually a non-root-mounted
8226 NFS file system. When this happens and the failure is EACCES, Exim now
8227 forks a subprocess and does the per-uid checking as the relevant uid.
8229 77. Added process_log_path.
8231 78. If log_file_path was not explicitly set, a setting of check_log_space or
8232 check_log_inodes was ignored.
8234 79. If a space check for the spool or log partitions fails, the incident is now
8235 logged. Of course, in the latter case the data may get lost...
8237 80. Added the %p formatting code to string_format() so that it can be used to
8238 print addresses in debug_print(). Adjusted all the address printing in the
8239 debugging in store.c to use %p rather than %d.
8241 81. There was a concern that a line of code in smtp_in.c could overflow a
8242 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8243 initially expressed, the concern was not well-founded, because trailing
8244 spaces are removed early. However, if the trailing spaces were followed by
8245 a NULL, they did not get removed, so the overflow was possible. Two fixes
8248 (a) I re-wrote the offending code in a cleaner fashion.
8249 (b) If an incoming SMTP command contains a NULL character, it is rejected
8252 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8253 runs initgroups(), so that if the Exim user is in any additional groups,
8254 they will be used during message reception.
8260 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob