1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes. Previously debugging stopped any time Exim
74 re-execs, or for processing a queued message.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
86 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
87 supplied as the db file-pair basename. Previously, if a directory
88 path was given, for example via the autoreply "once" option, the DB
89 file.pag and file.dir files would be created in that directory's
92 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
93 "taint" log_selector. These were previously deprecated.
95 JH/22 Fix static address-list lookups to properly return the matched item.
96 Previously only the domain part was returned.
98 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
99 the call into OpenSSL to send a TLS Close was being repeated; this
100 resulted in the library waiting for the peer's Close. If that was never
101 sent we waited forever. Fix by tracking send calls.
103 JH/24 The ${run} expansion item now expands its command string elements after
104 splitting. Previously it was before; the new ordering makes handling
105 zero-length arguments simpler. The old ordering can be obtained by
106 appending a new option "preexpand", after a comma, to the "run".
108 JH/25 Taint-check exec arguments for transport-initiated external processes.
109 Previously, tainted values could be used. This affects "pipe", "lmtp" and
110 "queryprogram" transport, transport-filter, and ETRN commands.
111 The ${run} expansion is also affected: in "preexpand" mode no part of
112 the command line may be tainted, in default mode the executable name
115 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
116 the the facility was not passed across execs, and only the first message
117 passed over a connection could use BDAT; any further ones using DATA.
119 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
120 uses $sending_ip_address and an interface is specified.
121 Previously any use of the local address in the EHLO name disabled
122 PIPECONNECT, the common case being to use the rDNS of it.
124 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
125 resumption. Previously verify failed because no certificate status is
126 passed on the wire for the restarted session. Fix by using the recorded
127 ocsp status of the stored session for the new connection.
129 JH/29 TLS resumption: the key for session lookup in the client now includes
130 more info that a server could potentially use in configuring a TLS
131 session, avoiding oferring mismatching sessions to such a server.
132 Previously only the server IP was used.
134 JH/30 Fix string_copyn() for limit greater than actual string length.
135 Previously the copied amount was the limit, which could result in a
136 overlapping memcpy for newly allocated destination soon after a
137 source string shorter than the limit. Found/investigated by KM.
139 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
140 close; it may be needed for a subsequent connection. This caused a
141 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
143 JH/32 Fix CHUNKING for a second message on a connection when the first was
144 rejected. Previously we did not reset the chunking-offered state, and
145 erroneously rejected the BDAT command. Investigation help from
148 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
149 an empty address. Previously the expansion returned an error.
151 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
152 proxy. Previously these were misparsed, leading to paniclog entries.
158 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
159 related applications. Previously an "H" was used where available info
160 says that "M" should be, so change to match.
162 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
163 as arguments, so an implementation trying to copy these into a local
164 buffer was taking a taint-enforcement trap. Fix by using dynamically
165 created buffers. Similar fix for radius expansion condition.
167 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
168 reasonable, eg. to count headers. Fix by using dynamically created
169 buffers rather than a local. Do similar fixes for ACL actions "dcc",
170 "log_reject_target", "malware" and "spam"; the arguments are expanded
171 so could be handling tainted values.
173 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
174 broken the (no-op) support for this sendmail command. Restore it
175 to doing nothing, silently, and returning good status.
177 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
178 record path was given (or the default used) without a leading directory
179 path, an error occurred on trying to open it. Use the transport's working
182 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
183 transport. Previously it was the name on the DNS A-record; use instead
184 the head of the CNAME chain leading there (if there is one). This seems
185 to align better with RFC 6125.
187 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
188 smtp_accept_max_per_host allocated resources which were not released
189 when the limit was exceeded. This eventually crashed the daemon. Fix
190 by adding a release action in that path.
192 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
193 expanded; previously using tainted values was rejected. Fix by using
194 dynamically-created buffers.
196 JH/09 Relax restrictions on ACL verify condition needing access to message
197 headers. Previously they were only permitted in data and non-smtp ACLs;
198 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
199 not_blind, header_sender and header_names_ascii verification.
201 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
202 Previously a macro used one argument twice; when called with the
203 argument as an expression having side-effects, incorrect operation
204 resulted. Use an inlineable function.
206 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
207 held open for a verify callout. Previously this wan not accounted for
208 and a corrupt onward SMTP conversation resulted.
210 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
211 Previously they were embedded naively in the constructed address; when
212 needed, strip the quoting and quote the entire local_part.
213 Also make the inbound_srs expansion condition handle quoting.
215 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
216 excluded, not matching the documentation.
218 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
219 was given for the sqlite_dbfile a trap resulted.
221 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
222 "name" argument resulted in a trap. There is no reason to disallow such;
223 this was a coding error.
225 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
226 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
227 spent suspended, ignoring the POSIX definition. Previously we assumed
228 it did and a constant offset from real time could be used as a correction.
229 Change to using the same clock source for the start-of-message and the
230 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
231 exists, just to get a clock slightly more aligned to reality.
233 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
234 RFC says it is optional some validators care. The missing char was not
235 intended but triggered by a line-wrap alignment. Discovery and fix by
236 Guillaume Outters, hacked on by JH.
238 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
239 name being quoted was tainted a trap would be taken. Fix by using
240 dynamically created buffers. The routine could have been called by a
241 rewrite with the "h" flag, by using the "-F" command-line option, or
242 by using a "name=" option on a control=submission ACL modifier.
244 JH/19 SPF: change the Authentication-Results expansion component to give
245 smtp.helo when the sender domain is empty. Previously it gave
248 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
249 not in 127.0.0.0/8 to help in spotting list domains taken over by a
250 domain-parking registrar.
252 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
253 Previously when a whitespace character was specified it was not inserted
254 after removing the newline.
256 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
257 the domain part of the recipient address. This overrides any tls_sni
258 option set, which was previously used.
260 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
263 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
264 is_tainted() had an off-by-one error in the overenthusiastic direction.
265 Find and fix by Gavan. Although NetBSD is not a supported platform for
266 4.94 this bug could affect other platforms.
268 PP/01 Fix default prime selection to be consistent.
269 One path used ike23 still, instead of exim.dev.20160529.3; now both
270 execution flows will use the same DH primes (currently
271 exim.dev.20160529.3).
273 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
274 option in smtp transport, to match the documentation. Previously
275 verification was not being done in some cases where it should have been.
277 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
278 than one server was defined and depending on the platform memory layout
279 details, an internal consistency trap could be hit while walking the list
282 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
283 files. The value is used by the authresults expansion item. Previously
284 if this was used in a router or transport, a crash could result.
286 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
287 have a sufficient incidence of EINTR returns from select that an
288 interaction with other operations done by the main daemon loop exposed
289 a bug in the error-handling. This was benign apart from the log
292 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
293 for consistency with delivered "=>" and failed "**" lines. While we're
294 there, handle PRX and TFO.
296 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
297 applied. This resulted, if any header-line rewrite rules were configured,
298 in a panic-log triggerable by sending a message with a long address in
299 a header. Fix by increasing the arbitrary limit to larger than a single
300 (dewrapped) 5322 header line maximum size.
302 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
303 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
304 RFC 6648 which deprecates X- options in protocols as a general practice.
305 Changeover between the implementations is handled by the mechanisms
308 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
309 a remote which uses callout/hold. Previously the local was queued.
311 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
314 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
315 log line. The asterisk is now consistently appended to the remote IP
316 (and port, if given), and will also be provided on defer and fail log
317 lines. Previously it could be placed on the local IP if that was being
318 logged, and was only provided on delivery lines.
320 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
322 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
323 authenticator client_send option. Previously the next char, after a pair
324 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
325 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
326 documentation. There is still no way to get a leading ^ immediately
327 after a NUL (ie. for the password of a PLAIN method authenticator.
329 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
330 files. For bad sizes read, delete the record and whine to paniclog.
332 JH/38 When logging an AUTH failure, as server, do not include sensitive
333 information. Previously, the credentials would be included if given
334 as part of the AUTH command line and an ACL denied authentication.
336 JH/39 Bug 2691: fix $local_part_data. When the matching list element
337 referred to a file, bad data was returned. This likely also affected
340 JH/40 The gsasl authenticator now supports caching of the salted password
341 generated by the client-side implementation. This required the addition
342 of a new variable: $auth4.
344 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
345 left undeleted; the attempt to re-create it then failed - resulting in
346 the usual "SIGHUP tp have daemon reload configuration" to not work.
347 This affected any platform not supporting "abstract" Unix-domain
348 sockets (i.e. not Linux).
350 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
351 recipients" response to RCPT in a later response, with a 250. The
352 previous coding assumed this would not happen, and under PIPELINING
353 would result in both lost and duplicate recipients for a message.
355 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
356 Previously the weighting was incorrectly applied. Similar fix for socks
357 proxies. Found and fixed by Heiko Schlichting.
359 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
360 not handle sub-lists included using the +namedlist syntax. While
361 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
362 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
365 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
366 Previously the permanent pool was used, so the sore could not be freed.
367 This meant a connection with many messages would use continually-growing
370 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
371 per-pool so as not to waste too much space. Previously a constant size
372 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
373 DOS attacks harder. The cost is wasted memory use in the larger blocks.
375 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
376 same space repeatedly, and a smaller process footprint.
378 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
379 for GnuTLS. Previously the name started "%s" which, while not a bug,
380 looked as if if might be one.
382 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
383 "continued connection") make the $proxy_* variables available. Previously
384 the information was not passed across the exec() call for subsequent
385 transport executions. This also mean that the log lines for the
386 messages can show the proxy information.
388 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
389 receive time. With modern systems the difference is significant.
390 The historical behaviour can be restored by disabling (a new) log_selector
391 "queue_time_exclusive".
393 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
394 content) were specifically tested for. Now, also cover expansions
395 resulting in ACL names and inline ACL content.
397 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
398 making it unusable in complex expressions.
400 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
401 fail with a taint-check on the spoolfile name, and leave the message
404 HS/01 Enforce absolute PID file path name.
406 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
408 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
410 PP/02 Bug 2643: Correct TLS DH constants.
411 A missing NUL termination in our code-generation tool had led to some
412 incorrect Diffie-Hellman constants in the Exim source.
413 Reported by kylon94, code-gen tool fix by Simon Arlott.
415 PP/03 Impose security length checks on various command-line options.
416 Fixes CVE-2020-SPRSS reported by Qualys.
418 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
419 better. Reported by Qualys.
421 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
422 providing a particularly obnoxious sender full name.
425 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
427 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
430 PP/08 Change default for recipients_max from unlimited to 50,000.
432 PP/09 Fix security issue with too many recipients on a message (to remove a
433 known security problem if someone does set recipients_max to unlimited,
434 or if local additions add to the recipient list).
435 Fixes CVE-2020-RCPTL reported by Qualys.
437 PP/10 Fix security issue in SMTP verb option parsing
438 Fixes CVE-2020-EXOPT reported by Qualys.
440 PP/11 Fix security issue in BDAT state confusion.
441 Ensure we reset known-good where we know we need to not be reading BDAT
442 data, as a general case fix, and move the places where we switch to BDAT
443 mode until after various protocol state checks.
444 Fixes CVE-2020-BDATA reported by Qualys.
446 HS/03 Die on "/../" in msglog file names
448 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
449 the Exim runtime user are allowed to create files.
451 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
454 QS/03 When reading the output from interpreted forward files we do not
455 pass the pipe between the parent and the interpreting process to
456 executed child processes (if any).
458 QS/04 Always die if requested from internal logging, even is logging is
461 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
462 the API; compilation noo longer completes with DMARC support included.
463 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
464 been reported on other platforms.
466 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
468 JH/56 Make the majority of info read from config files readonly, for defence-in-
469 depth against exploits. Suggestion by Qualys.
470 Not supported on Solaris 10.
472 JH/57 Fix control=fakreject for a custom message containing tainted data.
473 Previously this resulted in a log complaint, due to a re-expansion present
474 since fakereject was originally introduced.
476 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
477 modifier was given, a loop resulted.
479 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
480 pipelined SMTP command followed the BDAT LAST then it would be
481 incorrectly treated as part of the message body, causing a verification
484 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
485 require looser validation rules than those for 821-level addresses,
486 which only permit IP addresses.
492 JH/01 Avoid costly startup code when not strictly needed. This reduces time
493 for some exim process initialisations. It does mean that the logging
494 of TLS configuration problems is only done for the daemon startup.
496 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
498 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
499 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
502 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
503 DKIM signing is being done. Previously a transport_filter would always
504 disable CHUNKING, falling back to traditional DATA.
506 JH/05 Regard command-line recipients as tainted.
508 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
510 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
511 PAM library frees one of the arguments given to it, despite the
512 documentation. Therefore a plain malloc must be used.
514 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
515 on-stack buffers were used, resulting in a taint trap when DSN information
516 copied from a received message was written into the buffer.
518 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
519 the ordering of its ARC headers. This caused a crash.
521 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
522 a new record was being constructed with information from the peer, a trap
525 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
526 installation would get error messages from DMARC verify, when it hit the
527 nonexistent file indicated by the default. Distros wanting DMARC enabled
528 should both provide the file and set the option.
529 Also enforce no DMARC verification for command-line sourced messages.
531 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
532 could, depending on the platform, hang at the STARTTLS response.
534 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
535 message on a connection. Previously if one message had ARC headers and
536 the following one did not, a crash could result when adding an
537 Authentication-Results: header.
539 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
540 Exim main code and Exim-related utities. The introduction of taint
541 tracking also did many adjustments to string handling. Since then, eximon
542 frequently terminated with an assert failure.
544 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
545 check for 452 responses. This slightly helps the inefficieny of doing
546 a large alias-expansion into a recipient-limited target. The max_rcpt
547 transport option still applies (and at the current default, will override
548 the new feature). The check is done for either cause of synch, and forces
549 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
550 connection. The new facility is not tunable at this time.
552 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
553 library live data was being used, so the results became garbage. Make
554 copies while it is still usable.
556 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
557 on delivery deferred (==) and failed (**) lines (if a delivery was
558 attemtped). Previously it was only on completion (=>) lines.
560 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
561 for the expansion of the server_scram_iter and server_scram_salt options.
563 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
564 are now specifically given a NO_DATA response without hitting the system
565 resolver. The library goes on to do the now-standard TXT lookup.
566 Use of dnsdb lookups is not affected.
568 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
569 only retrieve the errormessage once. Previously two calls to dlerror()
570 were used, and the second one (for mainlog/paniclog) retrieved null
573 JH/20 Taint checking: disallow use of tainted data for
574 - the appendfile transport file and directory options
575 - the pipe transport command
576 - the autoreply transport file, log and once options
577 - file names used by the redirect router (including filter files)
579 - paths used by single-key lookups
580 Previously this was permitted.
582 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
583 adjusted the size of a major service buffer; this failed because the
584 buffer was in use at the time. Change to a compile-time increase in the
585 buffer size, when this authenticator is compiled into exim.
587 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
588 previous fast-mode was untenable in the face of glibs using mmap to
589 support larger malloc requests.
591 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
592 New values supported, if defined on system where compiled:
593 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
594 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
596 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
597 running a limited number of proceses in parallel, a benefit is gained. The
598 amount varies with the platform hardware and load. The use of the option
599 queue_run_in_order means we cannot do this, as ordering becomes
602 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
603 had introduced a string-copy (for ensuring NUL-termination) which was not
604 appropriate for that case, which can include embedded NUL bytes in the
605 block of data. Investigation showed the copy to actually be needless, the
606 data being length-specified.
608 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
609 done during a receiving connection, and both used TLS, global info was
610 used rather than per-connection info for tracking the state of data
611 queued for transmission. This could result in a connection hang.
613 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
614 Previously, when delivering serveral messages down a single connection
615 only the first would provide a SIZE. This was due to the size information
616 not being properly tracked.
618 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
619 TAI (at 37 seconds currently), pretend to be in UTC for time-related
620 expansion and logging. Previously, spurious values such as a future
621 minute could be seen.
623 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
624 it could crash from a null-deref. This could also affect the
625 ${addresses: } operator and ${readsock } item.
627 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
628 message following a mime one, the variable was not reset.
630 JH/30 When an pipelined-connect fails at the first response, assume incorrect
631 cached capability (perhaps the peer reneged?) and immediately retry in
634 JH/31 Fix spurious detection of timeout while writing to transport filter.
636 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
637 an attempt to copy the string was made before checking it.
639 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
640 taint of the lookup key was maintained; we now regard the presence in the
641 filesystem as sufficient validation.
643 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
644 argument is supplied.
646 JH/35 The dsearch lookup now requires that the directory is an absolute path.
647 Previously this was not checked, and nonempty relative paths made an
648 access under Exim's current working directory.
650 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
651 Previously no event was raised.
653 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
654 parameter supplied by the sender MAIL FROM command. Previously it was
655 ignored, and only the check_spool_space option value for the required
658 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
659 the size of the signing public-key. Previously it was instead giving
660 the size of the signature hash.
662 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
663 the default. See the (new) dkim_verify_min_keysizes option.
665 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
666 and an ACL use a lookup for checking either the local_part or domain,
667 stale data could be accessed. Ensure that variable references are
668 dropped between messages.
670 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
671 by the client was not checked as pointing within response data before
672 being used. A malicious client could thus cause an out-of-bounds read and
673 possibly gain authentication. Fix by adding the check.
675 JH/42 Internationalisation: change the default for downconversion in the smtp
676 transport to be "if needed". Previously it was "as previously set" for
677 the message, which usually meant "if needed" for message-submission but
678 "no" for everything else. However, MTAs have been seen using SMTPUTF8
679 even when the envelope addresses did not need it, resulting in forwarding
680 failures to non-supporting MTAs. A downconvert in such cases will be
681 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
682 transport. The change does mean that addresses needing conversion will
683 be converted when previously a delivery failure would occur.
685 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
686 response was received it would be used unchecked in a fail-DSN, violating
687 standards on line-length limits. Truncate if needed.
689 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
690 doesn't work. (Additionally add a new main config option to configure the
697 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
698 side, to decode a TLS 1.3 packet capture.
700 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
701 Previously the default library behaviour applied, sending two, each in
704 JH/03 Debug output for ACL now gives the config file name and line number for
707 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
709 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
711 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
712 buffer overrun for (non-chunking) other transports.
714 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
715 TLS1.3, means that a server rejecting a client certificate is not visible
716 to the client until the first read of encrypted data (typically the
717 response to EHLO). Add detection for that case and treat it as a failed
718 TLS connection attempt, so that the normal retry-in-clear can work (if
719 suitably configured).
721 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
722 and/or domain. Found and fixed by Jason Betts.
724 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
725 configuration). If a CNAME target was not a wellformed name pattern, a
728 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
729 the OS reports them interleaved with other addresses.
731 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
732 used both for input and for a verify callout, both encrypted, SMTP
733 responses being sent by the server could be lost. This resulted in
734 dropped connections and sometimes bounces generated by a peer sending
737 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
738 string. Previously it was possible to cause undefined behaviour in a
739 library routine (usually a crash). Found by "zerons".
741 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
744 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
745 API was removed, so update to use the newer ones.
747 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
748 any timeout set, is taking a long time. Previously we would hang on to a
749 rotated logfile "forever" if the input was arriving with long gaps
750 (a previous attempt to fix addressed lack, for a long time, of initial
753 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
754 shared (NFS) environment. The length of the tempfile name is now
755 4 + 16 ("hdr.$message_exim_id") which might break on file
756 systems which restrict the file name length to lower values.
759 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
760 shared (NFS) environment.
762 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
763 did for all versions <4.90). Notably -M, -m, --invert, -I may be
766 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
767 on some platforms for bit 31.
769 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
770 to changes apparently associated with TLS1.3 handling some of the APIs
771 previously used were either nonfunctional or inappropriate. Strings
772 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
773 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
774 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
775 This affects log line X= elements, the $tls_{in,out}_cipher variables,
776 and the use of specific cipher names in the encrypted= ACL condition.
778 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
780 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
781 verification result was not updated unless hosts_require_ocsp applied.
783 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
784 queue_list_requires_admin set to false, non-admin users were denied the
787 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
788 directory-of-certs mode. Previously they were advertised despite the
791 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
792 A single TCP connection by a client will now hold a TLS connection open
793 for multiple message deliveries, by default. Previously the default was to
796 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
797 default. If built with the facility, DANE will be used. The facility
798 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
800 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
801 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
802 must be defined and you must still, unless you define DISABLE_TLS, manage
803 the the include-dir and library-file requirements that go with that
804 choice. Non-TLS builds are still supported.
806 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
809 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
810 default. If the platform supports and has the facility enabled, it will
811 be requested on all coneections.
813 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
814 controlled by the build-time option SUPPORT_PIPE_CONNECT.
816 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
818 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
819 success-DSN messages. Previously the From: header was always the default
820 one for these; the option was ignored.
822 JH/28 Fix the timeout on smtp response to apply to the whole response.
823 Previously it was reset for every read, so a teergrubing peer sending
824 single bytes within the time limit could extend the connection for a
825 long time. Credit to Qualsys Security Advisory Team for the discovery.
827 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
828 delivery address, which leaked information of the results of local
829 forwarding. Change to the original envelope recipient address, per
832 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
833 requested. Previously not bounce was generated and a log entry of
834 error ignored was made.
836 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
838 JH/32 Introduce a general tainting mechanism for values read from the input
839 channel, and values derived from them. Refuse to expand any tainted
840 values, to catch one form of exploit.
842 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
843 was unused and the unexpanded text used for the test. Found and
844 fixed by Ruben Jenster.
846 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
847 an attempt to use a TLS library read routine dereffed a nul pointer,
850 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
851 them in our smtp response.
853 JH/36 Have the general router option retry_use_local_part default to true when
854 any of the restrictive preconditions are set (to anything). Previously it
855 was only for check_local user. The change removes one item of manual
856 configuration which is required for proper retries when a remote router
857 handles a subset of addresses for a domain.
859 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
860 link count into consideration.
862 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
863 caused the extension of big_buffer, the following lines were ignored.
865 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
866 accordance with RFC 2308. Previously there was no expiry, so a longlived
867 receive process (eg. due to ACL delays) versus a short SOA value could
870 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
872 JH/39 Promote DMARC support to mainline.
874 JH/40 Bug 2452: Add a References: header to DSNs.
876 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
877 parameters. The relevant library call is documented as "Deprecated: This
878 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
879 3.6.0, DH parameters are negotiated following RFC7919."
881 HS/06 Change the default of dnssec_request_domains to "*"
883 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
884 carried on and emitted a BDAT command, even when PIPELINING was not
887 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
888 buffer was used for the filename, resulting in a trap when tainted
889 arguments (eg. $domain) were used.
891 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
892 recommended to avoid a possible server-load attack. The feature can be
893 re-enabled via the openssl_options main cofiguration option.
895 JH/45 local_scan API: documented the current smtp_printf() call. This changed
896 for version 4.90 - adding a "more data" boolean to the arguments.
897 Bumped the ABI version number also, this having been missed previously;
898 release versions 4.90 to 4.92.3 inclusive were effectively broken in
899 respect of usage of smtp_printf() by either local_scan code or libraries
900 accessed via the ${dlfunc } expansion item. Both will need coding
901 adjustment for any calls to smtp_printf() to match the new function
902 signature; a FALSE value for the new argument is always safe.
904 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
905 the file-offset (which the Linux syscall does, and exim expects); this
906 resulted in an indefinite loop.
908 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
909 to do ARC verification. The Authentication-Results: header line added
910 by the configuration then had no ARC item.
912 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
913 an incorrect mode was used for reading the result, resulting in it being
920 JH/01 Remove code calling the customisable local_scan function, unless a new
921 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
923 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
924 non-signal-safe functions being used.
926 JH/03 Bug 2269: When presented with a received message having a stupidly large
927 number of DKIM-Signature headers, disable DKIM verification to avoid
928 a resource-consumption attack. The limit is set at twenty.
930 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
931 report of oldest_pass in ${authres } in consequence, and separate out
932 some descriptions of reasons for verification fail.
934 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
935 files in the spool were present and unlocked. A queue-runner could spot
936 them, resulting in a duplicate delivery. Fix that by doing the unlock
937 after the unlink. Investigation by Tim Stewart. Take the opportunity to
938 add more error-checking on spoolfile handling while that code is being
941 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
942 No known attacks, no CVE, this is defensive hardening.
944 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
945 a queue-runner could start a delivery while other operations were ongoing.
946 Cutthrough delivery was a common victim, resulting in duplicate delivery.
947 Found and investigated by Tim Stewart. Fix by using the open message data
948 file handle rather than opening another, and not locally closing it (which
949 releases a lock) for that case, while creating the temporary .eml format
950 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
952 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
953 $sender_verify_failure/$recipient_verify_failure to "random".
955 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
958 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
959 Previously this would segfault.
961 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
964 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
965 like zero, since the resolver should be doing this for us, But we need one
966 as a CNAME but no MX presence gets the CNAME returned; we need to check
967 that doesn't point to an MX to declare it "no MX returned" rather than
968 "error, loop". A new main option is added so the older capability of
969 following some limited number of chain links is maintained.
971 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
973 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
974 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
975 now-deprecated earlier definitions used only specified the range up to TLS
976 1.2 (in the older-version library docs).
978 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
980 JH/15 Rework TLS client-side context management. Stop using a global, and
981 explicitly pass a context around. This enables future use of TLS for
982 connections to service-daemons (eg. malware scanning) while a client smtp
983 connection is using TLS; with cutthrough connections this is quite likely.
985 JH/16 Fix ARC verification to do AS checks in reverse order.
987 JH/17 Support a "tls" option on the ${readsocket } expansion item.
989 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
990 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
991 Previously the "utf8" would be re-prepended for every additional message.
993 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
994 Previously thery were accepted, resulting in issues when attempting to
995 forward messages to a non-supporting MTA.
997 PP/02 Let -n work with printing macros too, not just options.
999 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1000 one parent address was copied, and bogus data was used at delivery-logging
1001 time. Either a crash (after delivery) or bogus log data could result.
1002 Discovery and analysis by Tim Stewart.
1004 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1005 Previously if the string ended mid-character, we did not insert the
1006 promised '?' replacement.
1008 PP/04 Documentation: current string operators work on bytes, not codepoints.
1010 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1011 should pack well giving a smaller memory footprint so better caching and
1012 therefore performance. Group the declarations where this can't be done so
1013 that the byte-sized flag variables are not interspersed among pointer
1014 variables, giving a better chance of good packing by the compiler.
1016 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1017 non-null, to avoid issues with sites running BATV. Previously reports were
1018 sent with an empty envelope sender so looked like bounces.
1020 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1021 The ignore_error flag wasn't being returned from the filter subprocess so
1022 was not set for later routers. Investigation and fix by Matthias Kurz.
1024 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1025 and a msg:complete for the whole, when a message is manually removed using
1026 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1028 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1029 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1030 hope that is portable enough.
1032 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1033 requiring a known-CA anchor certificate; make it now rely entirely on the
1034 TLSA as an anchor. Checking the name on the leaf cert against the name
1035 on the A-record for the host is still done for TA (but not for EE mode).
1037 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1038 included in delivery lines for non-proxied connections, when compiled with
1039 SUPPORT_SOCKS and running with proxy logging enabled.
1041 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1042 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1043 move the existing event to fire before the normal logging of message
1044 failure so that custom logging is bracketed by normal logging.
1046 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1047 msg:fail:internal event. Developement by Matthias Kurz.
1049 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1050 far too small for todays use of crypto signatures stored there. Go all
1051 the way to the max DNS message size of 64kB, even though this might be
1052 overmuch for IOT constrained device use.
1054 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1055 copy a string over itself. The library routine is documented as not
1056 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1058 JH/32 For main options check_spool_space and check_inode_space, where the
1059 platform supports 64b integers, support more than the previous 2^31 kB
1060 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1061 the previous G, M, k.
1063 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1064 $authenticated_fail_id variable on authentication failure. Previously
1067 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1068 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1069 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1070 GNUTLS_SEC_PARAM_MEDIUM.
1072 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1073 side. Previously we would continue as if no SNI had been received.
1075 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1076 "<" character, which should be a list-separator specification, we walked
1077 off past the nul-terimation.
1079 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1080 causes) even when the retry time is not yet met. Previously they were
1081 not, meaning that when (say) an account was over-quota and temp-rejecting,
1082 and multiple senders' messages were queued, only one sender would get
1083 notified on each configured delay_warning cycle.
1085 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1087 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1088 cipher-suites, an error can be left on the stack even for a succeeding
1089 accept; this results in impossible error messages when a later operation
1092 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1093 return error codes indicating retry. Under TLS1.3 this becomes required.
1095 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1096 it only wrote the new authenticators, resulting in a lack of tracking of
1097 peer changes of ESMTP extensions until the next cache flush.
1099 JH/41 Fix the loop reading a message header line to check for integer overflow,
1100 and more-often against header_maxsize. Previously a crafted message could
1101 induce a crash of the recive process; now the message is cleanly rejected.
1103 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1104 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1110 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1111 When redis_servers is set to a list of > 1 element, and the Redis servers
1112 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1113 case of MOVED into a DEFER case instead, thus moving the query onto the
1114 next server in the list. For a cluster of N elements, all N servers must
1115 be defined in redis_servers.
1117 GF/02 Catch and remove uninitialized value warning in exiqsumm
1118 Check for existence of @ARGV before looking at $ARGV[0]
1120 JH/01 Replace the store_release() internal interface with store_newblock(),
1121 which internalises the check required to safely use the old one, plus
1122 the allocate and data copy operations duplicated in both (!) of the
1123 extant use locations.
1125 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1126 modifier. This matches the restriction on the commandline.
1128 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1129 Previously only the last row was returned.
1131 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1132 we assumed that tags in the header were well-formed, and parsed the
1133 element content after inspecting only the first char of the tag.
1134 Assumptions at that stage could crash the receive process on malformed
1137 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1138 While running the DKIM ACL we operate on the Permanent memory pool so that
1139 variables created with "set" persist to the DATA ACL. Also (at any time)
1140 DNS lookups that fail create cache records using the Permanent pool. But
1141 expansions release any allocations made on the current pool - so a dnsdb
1142 lookup expansion done in the DKIM ACL releases the memory used for the
1143 DNS negative-cache, and bad things result. Solution is to switch to the
1144 Main pool for expansions.
1145 While we're in that code, add checks on the DNS cache during store_reset,
1146 active in the testsuite.
1147 Problem spotted, and debugging aided, by Wolfgang Breyha.
1149 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1150 When none of the hosts presented to a transport match an already-open
1151 connection, close it and proceed with the list. Previously we would
1152 queue the message. Spotted by Lena with Yahoo, probably involving
1155 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1156 Previously a spurious "250 OK id=" response was appended to the proper
1159 JH/08 The "support for" informational output now, which built with Content
1160 Scanning support, has a line for the malware scanner interfaces compiled
1161 in. Interface can be individually included or not at build time.
1163 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1164 by the template makefile "src/EDITME". The "STREAM" support for an older
1165 ClamAV interface method is removed.
1167 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1168 rows affected is given instead).
1170 JH/11 The runtime Berkeley DB library version is now additionally output by
1171 "exim -d -bV". Previously only the compile-time version was shown.
1173 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1174 SMTP connection. Previously, when one had more recipients than the
1175 first, an abortive onward connection was made. Move to full support for
1176 multiple onward connections in sequence, handling cutthrough connection
1177 for all multi-message initiating connections.
1179 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1180 routers. Previously, a multi-recipient message would fail to match the
1181 onward-connection opened for the first recipient, and cause its closure.
1183 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1184 a timeout on read on a GnuTLS initiating connection, resulting in the
1185 initiating connection being dropped. This mattered most when the callout
1186 was marked defer_ok. Fix to keep the two timeout-detection methods
1189 JH/15 Relax results from ACL control request to enable cutthrough, in
1190 unsupported situations, from error to silently (except under debug)
1191 ignoring. This covers use with PRDR, frozen messages, queue-only and
1194 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1196 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1197 metadata, resulting in a crash in free().
1199 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1200 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1201 Broken also in d185889f4, with init system revamp.
1203 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1204 Previously we abruptly closed the connection after reading a malware-
1205 found indication; now we go on to read the "scan ok" response line,
1208 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1209 ACL. Previously, a crash would result.
1211 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1212 macro text after a replacement (previously it was only once per line) and
1213 by skipping builtin macros when searching for an uppercase lead character.
1215 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1216 for the build is renamed.
1218 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1219 was allocated for every new TLS startup, meaning one per message. Fix
1220 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1222 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1223 reported the original. Fix to report (as far as possible) the ACL
1224 result replacing the original.
1226 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1227 OpenSSL. Certificate information is loaded for every new TLS startup,
1228 and the resources needed to be freed.
1230 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1232 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1235 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1236 DATA response info to the (existing) per-recipient response info for
1237 the "C=" log element. It can have useful tracking info from the
1238 destination system. Patch from Simon Arlott.
1240 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1241 length value. Previously this would segfault.
1243 HS/02 Support Avast multiline protoocol, this allows passing flags to
1244 newer versions of the scanner.
1246 JH/28 Ensure that variables possibly set during message acceptance are marked
1247 dead before release of memory in the daemon loop. This stops complaints
1248 about them when the debug_store option is enabled. Discovered specifically
1249 for sender_rate_period, but applies to a whole set of variables.
1250 Do the same for the queue-runner and queue-list loops, for variables set
1251 from spool message files. Do the same for the SMTP per-message loop, for
1252 certain variables indirectly set in ACL operations.
1254 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1255 as a multi-recipient message from a mailinglist manager). The coding had
1256 an arbitrary cutoff number of characters while checking for more input;
1257 enforced by writing a NUL into the buffer. This corrupted long / fast
1258 input. The problem was exposed more widely when more pipelineing of SMTP
1259 responses was introduced, and one Exim system was feeding another.
1260 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1261 receiving system, and refused recipients seen by the sending system
1262 (propating to people being dropped from mailing lists).
1263 Discovered and pinpointed by David Carter.
1265 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1266 replaced by the ${authresults } expansion.
1268 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1270 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1271 allows proper process termination in container environments.
1273 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1274 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1276 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1277 and "err_perm", deprecated since 4.83 when the RFC-defined words
1278 "temperror" and "permerror" were introduced.
1280 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1281 transport-filters or DKIM-signing. The restriction was lost in the
1282 consolidation of verify-callout and delivery SMTP handling.
1283 Extend the restriction to also cover ARC-signing.
1285 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1286 in defer=pass mode supply a 450 to the initiator. Previously the message
1289 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1290 tls_require_ciphers is used as before.
1292 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1293 "pass_unscanned". Only tmpfails from the scanner are written to
1294 the paniclog, as they may require admin intervention (permission
1295 denied, license issues). Other scanner errors (like decompression
1296 bombs) do not cause a paniclog entry.
1298 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1299 Previously it was possible to log spurious information in receive log
1302 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1303 triggered odd behaviour from Outlook Express clients.
1305 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1308 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1309 since the IETF WG has not yet settled on that versus the original
1310 "bare" representation.
1312 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1313 Previously the millisecond value corrupted the output.
1314 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1315 corrupted the output.
1321 JH/01 Rework error string handling in TLS interface so that the caller in
1322 more cases is responsible for logging. This permits library-sourced
1323 string to be attached to addresses during delivery, and collapses
1324 pairs of long lines into single ones.
1326 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1327 during configuration. Wildcards are allowed and expanded.
1329 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1330 This permits better logging.
1332 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1333 do not close it down and have the child transport start it up again on
1334 the passed-on TCP connection. Instead, proxy the child (and any
1335 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1336 affected: the continued delivery log lines do not have any DNSSEC, TLS
1337 Certificate or OCSP information. TLS cipher information is still logged.
1339 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1340 identical IP addresses on different listening ports. Will also affect
1343 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1344 add noisy ifdef guards to special-case this sillyness.
1345 Patch from Bernd Kuhls.
1347 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1348 than 255 are no longer allowed.
1350 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1351 Disable the session-cache too, which might reduce our load. Since we
1352 currrectly use a new context for every connection, both as server and
1353 client, there is no benefit for these.
1354 GnuTLS appears to not support tickets server-side by default (we don't
1355 call gnutls_session_ticket_enable_server()) but client side is enabled
1356 by default on recent versions (3.1.3 +) unless the PFS priority string
1359 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1360 <https://reproducible-builds.org/specs/source-date-epoch/>.
1362 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1363 the check for any unsuccessful recipients did not notice the limit, and
1364 erroneously found still-pending ones.
1366 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1367 MSG_MORE. Only in-clear (not on TLS connections).
1369 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1370 transport-filter is involved we can buffer the headers in memory for
1371 creating the signature, and read the spool data file once for the
1372 signature and again for transmission.
1374 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1375 4.77 as the kernel support then wasn't solid, having issues in 64bit
1376 mode. Now, it's been long enough. Add support for FreeBSD also.
1378 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1379 case where the routing stage had gathered several addresses to send to
1380 a host before calling the transport for the first, we previously failed
1381 to close down TLS in the old transport process before passing the TCP
1382 connection to the new process. The new one sent a STARTTLS command
1383 which naturally failed, giving a failed delivery and bloating the retry
1384 database. Investigation and fix prototype from Wolfgang Breyha.
1386 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1387 false-negatives in the check that the sender had not preempted a response
1388 or prompt from Exim (running as a server), due to that code's lack of
1389 awareness of the SMTP input buffering.
1391 PP/04 Add commandline_checks_require_admin option.
1392 Exim drops privileges sanely, various checks such as -be aren't a
1393 security problem, as long as you trust local users with access to their
1394 own account. When invoked by services which pass untrusted data to
1395 Exim, this might be an issue. Set this option in main configuration
1396 AND make fixes to the calling application, such as using `--` to stop
1399 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1400 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1401 responses to those, into a single TLS record each way (this usually means
1402 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1405 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1406 affects you only if you're dancing at the edge of the param size limits.
1407 If you are, and this message makes sense to you, then: raise the
1408 configured limit or use OpenSSL 1.1. Nothing we can do for older
1411 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1412 cmdline element to get the documented default one. Previously it was
1415 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1416 are unsuable for non-message-specific reasons - eg. connection timeout,
1417 banner-time rejection.
1419 JH/16 Fix logging of delivery remote port, when specified by router, under
1422 PP/06 Repair manualroute's ability to take options in any order, even if one
1423 is the name of a transport.
1426 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1428 JH/17 Change the list-building routines interface to use the expanding-string
1429 triplet model, for better allocation and copying behaviour.
1431 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1432 Previously it was constructed the first time a possibly-matching string
1433 was met in the configuration file input during startup; now it is done
1436 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1437 compatible one, to avoid the (poorly documented) possibility of a config
1438 file in the working directory redirecting the DB files, possibly correpting
1439 some existing file. CVE-2017-10140 assigned for BDB.
1441 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1442 cache-hot. Previously, although the result was properly cached, the
1443 initial verify call returned a defer.
1445 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1446 the main verify for receipient in uncached-mode.
1448 JH/22 Retire historical build files to an "unsupported" subdir. These are
1449 defined as "ones for which we have no current evidence of testing".
1451 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1452 if present. Previously it was ignored.
1454 JH/24 Start using specified-initialisers in C structure init coding. This is
1455 a C99 feature (it's 2017, so now considered safe).
1457 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1458 if was a fixed-sized field and bitmask ops via macros; it is now more
1461 PP/07 GitHub PR 56: Apply MariaDB build fix.
1462 Patch provided by Jaroslav Škarvada.
1464 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1465 during Coverity cleanups [4.87 JH/47]
1466 Diagnosis and fix provided by Michael Fischer v. Mollard.
1468 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1469 the right size to place the terminating semicolon on its own folded
1470 line, the header hash was calculated to an incorrect value thanks to
1471 the (relaxed) space the fold became.
1473 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1474 and confused the parent.
1476 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1477 which could crash as a result. This could lead to undeliverable messages.
1479 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1482 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1483 matters on fast-turnover and PID-randomising systems, which were getting
1484 out-of-order delivery.
1486 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1487 a possibly-overlapping copy. The symptom was that "Remote host closed
1488 connection in response to HELO" was logged instead of the actual 4xx
1491 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1492 Previously only that bufferd was discarded, resulting in SYMTP command
1495 JH/32 DKIM: when a message has multiple signatures matching an identity given
1496 in dkim_verify_signers, run the dkim acl once for each. Previously only
1497 one run was done. Bug 2189.
1499 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1500 "panic the current process" to "deliberately defer". The panic log is
1501 still written with the problem list name; the mail and reject logs now
1502 get a temp-reject line for the message that was being handled, saying
1503 something like "domains check lookup or other defer". The SMTP 451
1504 message is still "Temporary local problem".
1506 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1507 A crafted sequence of BDAT commands could result in in-use memory beeing
1508 freed. CVE-2017-16943.
1510 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1511 from SMTP input. Previously it was always done; now only done for DATA
1512 and not BDAT commands. CVE-2017-16944.
1514 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1515 to the message (such as an overlong header line). Previously this was
1516 not done and we did not exit BDAT mode. Followon from the previous item
1517 though a different problem.
1523 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1524 than -2003 did; needs libidn2 in addition to libidn.
1526 JH/02 The path option on a pipe transport is now expanded before use.
1528 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1529 Patch provided by "Björn", documentation fix added too.
1531 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1532 missing a wire-to-host endian conversion.
1534 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1535 close after a BDAT command line could be taken as a following command,
1536 giving a synch failure. Fix by only checking for synch immediately
1537 before acknowledging the chunk.
1539 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1540 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1541 macro. Patches provided by Josh Soref.
1543 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1544 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1545 by IANA as a service extension. However, John Klensin suggests that we
1548 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1549 the dkim code may be unix-mode line endings rather than smtp wire-format
1550 CRLF, so prepend a CR to any bare LF.
1552 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1553 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1555 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1556 the first were themselves being wrongly included in the feed into dkim
1557 processing; with most chunk sizes in use this resulted in an incorrect
1558 body hash calculated value.
1560 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1561 DKIM signature block, for verification. Although advised against by
1562 standards it is specifically not ruled illegal.
1564 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1566 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1567 missing a body hash (the bh= tag).
1569 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1570 It seems that HAProxy sends the Proxy Protocol information in clear and
1571 only then does a TLS startup, so do the same.
1573 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1574 TCP connections (such as for Spamd) unless the daemon successfully set
1575 Fast Open mode on its listening sockets. This fixes breakage seen on
1576 too-old kernels or those not configured for Fast Open, at the cost of
1577 requiring both directions being enabled for TFO, and TFO never being used
1578 by non-daemon-related Exim processes.
1580 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1581 endings, at least on the first header line. Try to canonify any that get
1582 past that check, despite the cost.
1584 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1585 now limited to an arbitrary five deep, while parsing addresses with the
1586 strip_excess_angle_brackets option enabled.
1588 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1589 instead leave the unprompted TLS handshake in socket buffer for the
1590 TLS library to consume.
1592 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1594 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1596 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1597 region in one operation, for speed, and this leaves assigned pointers
1598 dangling. Add checks run only under the testsuite which checks all
1599 variables at a store-reset and panics on a dangling pointer; add code
1600 explicitly nulling out all the variables discovered. Fixes one known
1601 bug: a transport crash, where a dangling pointer for $sending_ip_address
1602 originally assigned in a verify callout, is re-used.
1604 PP/06 Drop '.' from @INC in various Perl scripts.
1606 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1608 PP/08 Reduce a number of compilation warnings under clang; building with
1609 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1610 should be warning-free.
1612 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1614 HS/01 Fix portability problems introduced by PP/08 for platforms where
1615 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1617 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1618 chunk. This allows us to accept broken chunked messages. We need a more
1619 general solution here.
1621 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1622 already-broken messages in the queue.
1624 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1626 JH/19 Fix reference counting bug in routing-generated-address tracking.
1632 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1633 supports it and a size is available (ie. the sending peer gave us one).
1635 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1636 years of being deprecated). The replacements are the ACLs
1637 acl_smtp_mime and acl_not_smtp_mime.
1639 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1640 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1641 or even an in-clear connection were permitted. Now, if the host lookup
1642 was dnssec and dane was requested then the host is only used if the
1643 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1644 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1645 if one fails this test.
1646 This means that a poorly-configured remote DNS will make it incommunicado;
1647 but it protects against a DNS-interception attack on it.
1649 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1650 non-noisy when a race steals the message being considered.
1652 JH/05 If main configuration option tls_certificate is unset, generate a
1653 self-signed certificate for inbound TLS connections.
1655 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1656 in rewrites and routers.
1658 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1659 and logged a warning sing 4.83; now they are a configuration file error.
1661 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1662 (lacking @domain). Apply the same qualification processing as RCPT.
1664 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1666 JH/10 Support ${sha256:} applied to a string (as well as the previous
1669 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1670 a cutthrough deliver is pending, as we always want to make a connection.
1671 This also avoids re-routing the message when later placing the cutthrough
1672 connection after a verify cache hit.
1673 Do not update it with the verify result either.
1675 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1676 when routing results in more than one destination address.
1678 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1679 signing (which inhibits the cutthrough capability). Previously only
1680 the presence of an option was tested; now an expansion evaluating as
1681 empty is permissible (obviously it should depend only on data available
1682 when the cutthrough connection is made).
1684 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1685 the relevant preceding SMTP command did not note the pipelining mode.
1687 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1688 Previously they were not counted.
1690 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1691 as one having no matching records. Previously we deferred the message
1692 that needed the lookup.
1694 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1695 distinguished as "(=".
1697 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1698 for missing MX records. Previously it only worked for missing A records.
1700 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1702 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1703 after the data-go-ahead and data-ack. Patch from Jason Betts.
1705 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1706 even for a "none" policy. Patch from Tony Meyer.
1708 JH/22 Fix continued use of a connection for further deliveries. If a port was
1709 specified by a router, it must also match for the delivery to be
1712 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1713 When one of the recipients of a message was unsuitable for the connection
1714 (has no matching addresses), we lost track of needing to mark it
1715 deferred. As a result mail would be lost.
1717 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1719 JH/25 Decoding ACL controls is now done using a binary search; the source code
1720 takes up less space and should be simpler to maintain. Merge the ACL
1721 condition decode tables also, with similar effect.
1723 JH/26 Fix problem with one_time used on a redirect router which returned the
1724 parent address unchanged. A retry would see the parent address marked as
1725 delivered, so not attempt the (identical) child. As a result mail would
1728 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1729 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1730 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1733 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1734 Main config options check_{log,spool}_{inodes,space} are now
1735 100 inodes, 10MB unless set otherwise in the configuration.
1737 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1738 Previously it only applied to the main-section connection policy
1741 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1743 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1744 by me. Added RFC7919 DH primes as an alternative.
1746 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1747 are not in the system include path.
1749 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1750 GnuTLS, when a session startup failed (eg because the client disconnected)
1751 Exim did stdio operations after fclose. This was exposed by a recent
1752 change which nulled out the file handle after the fclose.
1754 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1755 signed directly by the cert-signing cert, rather than an intermediate
1756 OCSP-signing cert. This is the model used by LetsEncrypt.
1758 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1760 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1761 an incoming connection.
1763 HS/02 Bug 1802: Do not half-close the connection after sending a request
1766 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1767 fallback to "prime256v1".
1769 JH/34 SECURITY: Use proper copy of DATA command in error message.
1770 Could leak key material. Remotely exploitable. CVE-2016-9963.
1776 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1777 and 3.4.4 - once the server is enabled to respond to an OCSP request
1778 it does even when not requested, resulting in a stapling non-aware
1779 client dropping the TLS connection.
1781 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1782 support variable-length bit vectors. No functional change.
1784 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1785 The I= interface field on outgoing lines is now after the H= remote
1786 host field, same as incoming lines. There is a separate
1787 outgoing_interface log selector which allows you to disable the
1790 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1791 If not running log_selector +smtp_connection the mainlog would be held
1792 open indefinitely after a "too many connections" event, including to a
1793 deleted file after a log rotate. Leave the per net connection logging
1794 leaving it open for efficiency as that will be quickly detected by the
1795 check on the next write.
1797 HS/01 Bug 1671: Fix post transport crash.
1798 Processing the wait-<transport> messages could crash the delivery
1799 process if the message IDs didn't exist for some reason. When
1800 using 'split_spool_directory=yes' the construction of the spool
1801 file name failed already, exposing the same netto behaviour.
1803 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1804 mime_regex ACL conditions.
1806 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1807 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1808 to HELO, local diagnostic string.
1810 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1811 log line to debug. Even when configured with a tls authenticator many
1812 client connections are expected to not authenticate in this way, so
1813 an authenticate fail is not an error.
1815 HS/02 Add the Exim version string to the process info. This way exiwhat
1816 gives some more detail about the running daemon.
1818 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1819 matter for fast-change records such as DNSBLs.
1821 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1822 retry record, even if constant. There may be multiple transports with
1823 different interface settings and the retry behaviour needs to be kept
1826 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1828 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1830 JH/10 Bug 840: fix log_defer_output option of pipe transport
1832 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1833 hosts_randomize. This matters a lot when combined with mua_wrapper.
1835 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1836 ${quote_pgsql:<string>} operator.
1838 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1840 JH/14 Change header file name for freeradius-client. Relevant if compiling
1841 with Radius support; from the Gentoo tree and checked under Fedora.
1843 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1845 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1846 option transport_filter, meaning no filtering.
1848 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1850 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1851 defaults to "*" (all hosts). The variable is now available when not built
1852 with TLS, default unset, mainly to enable keeping the testsuite sane.
1853 If a server certificate is not supplied (via tls_certificate) an error is
1854 logged, and clients will find TLS connections fail on startup. Presumably
1855 they will retry in-clear.
1856 Packagers of Exim are strongly encouraged to create a server certificate
1857 at installation time.
1859 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1860 with the $config_file variable.
1862 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1863 in transport context, after the attempt, and per-recipient. The latter type
1864 is per host attempted. The event data is the error message, and the errno
1865 information encodes the lookup type (A vs. MX) used for the (first) host,
1866 and the trailing two digits of the smtp 4xx response.
1868 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1869 to write to mainlog (or rejectlog, paniclog) in the window between file
1870 creation and permissions/ownership being changed. Particularly affects
1871 installations where exicyclog is run as root, rather than exim user;
1872 result is that the running daemon panics and dies.
1874 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1876 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1877 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1878 "pri" and "weight". Note that the previous implicit priority given by the
1879 list order is no longer honoured.
1881 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1882 for DKIM processing.
1884 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1885 by defining SUPPORT_SOCKS.
1887 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1888 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1889 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1890 variables are renamed to proxy_{local,external}_{address,port}.
1892 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1893 for a build by defining SUPPORT_I18N
1895 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1896 of the query string, and make ${quote_redis:} do that quoting.
1898 JH/29 Move Events support from Experimental to mainline, enabled by default
1899 and removable for a build by defining DISABLE_EVENT.
1901 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1903 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1904 cached by the daemon.
1906 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1907 by defining LOOKUP_REDIS. The libhiredis library is required.
1909 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1910 keys are given for lookup.
1912 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1913 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1914 only supported when built with TLS support. The PolarSSL SHA routines
1915 are still used when the TLS library is too old for convenient support.
1917 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1918 openssl_options), for security. OpenSSL forces this from version 1.1.0
1919 server-side so match that on older versions.
1921 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1922 allocation for $value could be released as the expansion processing
1923 concluded, but leaving the global pointer active for it.
1925 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1926 and to use the domains and local_parts ACL conditions.
1928 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1929 incorrectly not doubled on cutthrough transmission, hence seen as a
1930 body-termination at the receiving system - resulting in truncated mails.
1931 Commonly the sender saw a TCP-level error, and retransmitted the message
1932 via the normal store-and-forward channel. This could result in duplicates
1933 received - but deduplicating mailstores were liable to retain only the
1934 initial truncated version.
1936 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1938 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1940 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1941 we're in there, support oversigning also; bug 1309.
1943 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1945 HS/04 Add support for keep_environment and add_environment options.
1947 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1948 either intentional arithmetic overflow during PRNG, or testing config-
1951 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1952 delivery resulted in actual delivery. Cancel cutthrough before DATA
1955 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1956 rejects a recipient: pass the reject to the originator.
1958 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1959 Many were false-positives and ignorable, but it's worth fixing the
1962 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1963 for the new environment-manipulation done at startup. Move the routines
1964 from being local to tls.c to being global via the os.c file.
1966 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1967 an extract embedded as result-arg for a map, the first arg for extract
1968 is unavailable so we cannot tell if this is a numbered or keyed
1969 extraction. Accept either.
1975 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1978 JH/02 The smtp transport option "multi_domain" is now expanded.
1980 JH/03 The smtp transport now requests PRDR by default, if the server offers
1983 JH/04 Certificate name checking on server certificates, when exim is a client,
1984 is now done by default. The transport option tls_verify_cert_hostnames
1985 can be used to disable this per-host. The build option
1986 EXPERIMENTAL_CERTNAMES is withdrawn.
1988 JH/05 The value of the tls_verify_certificates smtp transport and main options
1989 default to the word "system" to access the system default CA bundle.
1990 For GnuTLS, only version 3.0.20 or later.
1992 JH/06 Verification of the server certificate for a TLS connection is now tried
1993 (but not required) by default. The verification status is now logged by
1994 default, for both outbound TLS and client-certificate supplying inbound
1997 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2000 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2001 Status Notification (bounce) messages are now MIME format per RFC 3464.
2002 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2003 under the control of the dsn_advertise_hosts option, and routers may
2004 have a dsn_lasthop option.
2006 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2007 default, modifiable by a malware= option. The list separator for
2008 the options can now be changed in the usual way. Bug 68.
2010 JH/10 The smtp_receive_timeout main option is now expanded before use.
2012 JH/11 The incoming_interface log option now also enables logging of the
2013 local interface on delivery outgoing connections.
2015 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2016 if the interface and destination host and port all match.
2018 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2021 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2022 Patch from Andrew Lewis.
2024 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2025 now supports optional time-restrictions, weighting, and priority
2026 modifiers per server. Patch originally by <rommer@active.by>.
2028 JH/16 The spamd_address main option now supports a mixed list of local
2029 and remote servers. Remote servers can be IPv6 addresses, and
2030 specify a port-range.
2032 JH/17 Bug 68: The spamd_address main option now supports an optional
2033 timeout value per server.
2035 JH/18 Bug 1581: Router and transport options headers_add/remove can
2036 now have the list separator specified.
2038 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2041 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2044 JH/21 Support for the A6 type of dns record is withdrawn.
2046 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2047 rather than the verbs used.
2049 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2050 from 255 to 1024 chars.
2052 JH/24 Verification callouts now attempt to use TLS by default.
2054 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2055 are generic router options now. The defaults didn't change.
2057 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2058 Original patch from Alexander Shikoff, worked over by JH.
2060 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2061 files if ZCAT_COMMAND is not executable.
2063 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2065 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2067 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2068 Normally benign, it bites when the pair was led to by a CNAME;
2069 modern usage is to not canonicalize the domain to a CNAME target
2070 (and we were inconsistent anyway for A-only vs AAAA+A).
2072 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2074 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2075 when evaluating $sender_host_dnssec.
2077 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2078 $sender_helo_dnssec variable.
2080 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2082 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2084 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2086 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2087 documented as working, but never had. Support all but $spam_report.
2089 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2090 added for tls authenticator.
2092 HS/03 Add perl_taintmode main config option
2098 TL/01 When running the test suite, the README says that variables such as
2099 no_msglog_check are global and can be placed anywhere in a specific
2100 test's script, however it was observed that placement needed to be near
2101 the beginning for it to behave that way. Changed the runtest perl
2102 script to read through the entire script once to detect and set these
2103 variables, reset to the beginning of the script, and then run through
2104 the script parsing/test process like normal.
2106 TL/02 The BSD's have an arc4random API. One of the functions to induce
2107 adding randomness was arc4random_stir(), but it has been removed in
2108 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2109 function when detected.
2111 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2112 cause callback expansion.
2114 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2115 syntax errors in an expansion can be treated as a string instead of
2116 logging or causing an error, due to the internal use of bool_lax
2117 instead of bool when processing it.
2119 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2120 server certificates when making smtp deliveries.
2122 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2124 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2126 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2128 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2129 Merged patch from Sebastian Wiedenroth.
2131 JH/05 Fix results-pipe from transport process. Several recipients, combined
2132 with certificate use, exposed issues where response data items split
2133 over buffer boundaries were not parsed properly. This eventually
2134 resulted in duplicates being sent. This issue only became common enough
2135 to notice due to the introduction of connection certificate information,
2136 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2138 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2139 size buffer was used, resulting in syntax errors when an expansion
2142 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2143 version 3.3.6 or later.
2145 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2146 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2147 both become "event_action", the variables become $event_name, $event_data
2148 and $event_defer_errno. There is a new variable $verify_mode, usable in
2149 routers, transports and related events. The tls:cert event is now also
2150 raised for inbound connections, if the main configuration event_action
2153 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2154 early OCSP support, but no stapling (appears to be less than 1.0.0).
2156 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2157 server certificate names available under the smtp transport option
2158 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2161 JH/10 Time-related extraction expansions from certificates now use the main
2162 option "timezone" setting for output formatting, and are consistent
2163 between OpenSSL and GnuTLS compilations. Bug 1541.
2165 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2166 encoded parameter in the incoming message. Bug 1558.
2168 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2169 include certificate info, eximon was claiming there were spoolfile
2172 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2174 JH/14 Log delivery-related information more consistently, using the sequence
2175 "H=<name> [<ip>]" wherever possible.
2177 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2178 are problematic for Debian distribution, omit them from the release
2181 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2183 JH/16 Fix string representation of time values on 64bit time_t architectures.
2186 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2187 output list separator was used.
2192 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2193 checkers that were complaining about end of non-void function with no
2196 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2197 This was a regression introduced in 4.83 by another bugfix.
2199 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2201 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2202 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2208 TF/01 Correctly close the server side of TLS when forking for delivery.
2210 When a message was received over SMTP with TLS, Exim failed to clear up
2211 the incoming connection properly after forking off the child process to
2212 deliver the message. In some situations the subsequent outgoing
2213 delivery connection happened to have the same fd number as the incoming
2214 connection previously had. Exim would try to use TLS and fail, logging
2215 a "Bad file descriptor" error.
2217 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2218 utilities have not been installed.
2220 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2221 temporary space as the ACL may create new global variables.
2223 TL/01 LDAP support uses per connection or global context settings, depending
2224 upon the detected version of the libraries at build time.
2226 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2227 to extract and use the src ip:port in logging and expansions as if it
2228 were a direct connection from the outside internet. PPv2 support was
2229 updated based on HAProxy spec change in May 2014.
2231 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2233 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2234 Properly escape header and check for NULL return.
2236 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2239 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2241 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2242 characters in header names, implemented as a verify condition.
2243 Contributed by Michael Fischer v. Mollard.
2245 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2246 results permerror and temperror. Previous values are deprecated but
2247 still accepted. In a future release, err_perm and err_temp will be
2248 completely removed, which will be a backward incompatibility if the
2249 ACL tests for either of these two old results. Patch contributed by
2250 user bes-internal on the mailing list.
2252 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2254 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2255 selectors, in both main and reject logs.
2257 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2260 JH/07 Add malware type "sock" for talking to simple daemon.
2262 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2264 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2265 routers/transports under cutthrough routing.
2267 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2268 numbers. Touch up "bool" conditional to keep the same definition.
2270 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2272 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2274 JH/12 Expand items in router/transport headers_add or headers_remove lists
2275 individually rather than the list as a whole. Bug 1452.
2277 Required for reasonable handling of multiple headers_ options when
2278 they may be empty; requires that headers_remove items with embedded
2279 colons must have them doubled (or the list-separator changed).
2281 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2282 view the policy declared in the DMARC record. Currently, $dmarc_status
2283 is a combined value of both the record presence and the result of the
2286 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2288 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2289 dnslookup router and the smtp transport (applying to the forward
2292 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2293 of ldap servers used for a specific lookup. Patch provided by Heiko
2296 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2297 New variable $lookup_dnssec_authenticated for observability.
2299 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2300 Patch submitted by Lars Timman.
2302 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2304 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2305 Requires trusted mode and valid format message id, aborts otherwise.
2306 Patch contributed by Heiko Schlichting.
2308 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2309 certextract with support for various fields. Bug 1358.
2311 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2312 is requested by default, modifiable by smtp transport option
2315 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2316 operate on certificate variables to give certificate fingerprints
2317 Also new ${sha256:cert_variable}.
2319 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2321 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2322 Christian Aistleitner.
2324 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2326 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2327 file. Patch from Wolfgang Breyha.
2329 JH/25 Expand the coverage of the delivery $host and $host_address to
2330 client authenticators run in verify callout. Bug 1476.
2332 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2333 align with daemon_smtp_ports. Bug 72.
2335 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2336 support and error reporting did not work properly.
2338 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2339 and is readable. Patch from Andrew Colin Kissa.
2341 TL/14 Enhance documentation of ${run expansion and how it parses the
2342 commandline after expansion, particularly in the case when an
2343 unquoted variable expansion results in an empty value.
2345 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2347 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2348 by renaming away from it. Addresses GH issue 15, reported by
2351 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2353 TL/15 SECURITY: prevent double expansion in math comparison functions
2354 (can expand unsanitized data). Not remotely exploitable.
2361 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2363 PP/02 Make -n do something, by making it not do something.
2364 When combined with -bP, the name of an option is not output.
2366 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2369 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2370 $sender_host_name and config options to manage this, and basic check
2373 PP/05 DSCP support for outbound connections and control modifier for inbound.
2375 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2376 (Only plugin which currently uses this is kerberos4, which nobody should
2377 be using, but we should make it available and other future plugins might
2378 conceivably use it, even though it would break NAT; stuff *should* be
2379 using channel bindings instead).
2381 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2382 name; added for Sendmail compatibility; requires admin caller.
2383 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2384 just ignore it); requires trusted caller.
2385 Also parse but ignore: -Ac -Am -X<logfile>
2388 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2390 TL/02 Add +smtp_confirmation as a default logging option.
2392 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2393 Patch by Magnus Holmgren from 2007-02-20.
2395 TL/04 Bugzilla 1281 - Spec typo.
2396 Bugzilla 1283 - Spec typo.
2397 Bugzilla 1290 - Spec grammar fixes.
2399 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2401 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2403 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2404 mailing list by by Dmitry Isaikin.
2406 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2408 JH/02 Support "G" suffix to numbers in ${if comparisons.
2410 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2412 NM/01 Bugzilla 1197 - Spec typo
2413 Bugzilla 1196 - Spec examples corrections
2415 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2417 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2418 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2421 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2422 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2424 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2425 "acl {{name}{arg}...}", and optional args on acl condition
2428 JH/05 Permit multiple router/transport headers_add/remove lines.
2430 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2432 JH/07 Avoid using a waiting database for a single-message-only transport.
2433 Performance patch from Paul Fisher. Bugzilla 1262.
2435 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2438 JH/09 Add $headers_added variable, with content from use of ACL modifier
2439 add_header (but not yet added to the message). Bugzilla 199.
2441 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2442 Pulled from Bugzilla 817 by Wolfgang Breyha.
2444 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2446 (nb: this is the same fix as in Exim 4.80.1)
2448 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2451 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2454 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2456 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2457 advertise SMTP AUTH mechanism to us, instead of a generic
2458 protocol violation error. Also, make Exim more robust to bad
2459 data from the Dovecot auth socket.
2461 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2463 When a queue runner is handling a message, Exim first routes the
2464 recipient addresses, during which it prunes them based on the retry
2465 hints database. After that it attempts to deliver the message to
2466 any remaining recipients. It then updates the hints database using
2469 So if a recipient address works intermittently, it can get repeatedly
2470 deferred at routing time. The retry hints record remains fresh so the
2471 address never reaches the final cutoff time.
2473 This is a fairly common occurrence when a user is bumping up against
2474 their storage quota. Exim had some logic in its local delivery code
2475 to deal with this. However it did not apply to per-recipient defers
2476 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2478 This change adds a proper retry rule check during routing so that the
2479 final cutoff time is checked against the message's age. We only do
2480 this check if there is an address retry record and there is not a
2481 domain retry record; this implies that previous attempts to handle
2482 the address had the retry_use_local_parts option turned on. We use
2483 this as an approximation for the destination being like a local
2484 delivery, as in LMTP.
2486 I suspect this new check makes the old local delivery cutoff check
2487 redundant, but I have not verified this so I left the code in place.
2489 TF/02 Correct gecos expansion when From: is a prefix of the username.
2491 Test 0254 submits a message to Exim with the header
2495 When I ran the test suite under the user fanf2, Exim expanded
2496 the header to contain my full name, whereas it should have added
2497 a Resent-Sender: header. It erroneously treats any prefix of the
2498 username as equal to the username.
2500 This change corrects that bug.
2502 GF/01 DCC debug and logging tidyup
2503 Error conditions log to paniclog rather than rejectlog.
2504 Debug lines prefixed by "DCC: " to remove any ambiguity.
2506 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2508 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2509 Bug spotted by Jeremy Harris; was flawed since initial commit.
2510 Would have resulted in OCSP responses post-SNI triggering an Exim
2511 NULL dereference and crash.
2513 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2515 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2516 Bug detection, analysis and fix by Samuel Thibault.
2517 Bugzilla 1331, Debian bug #698092.
2519 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2521 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2522 Server implementation by Todd Lyons, client by JH.
2523 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2524 config variable "prdr_enable" controls whether the server
2525 advertises the facility. If the client requests PRDR a new
2526 acl_data_smtp_prdr ACL is called once for each recipient, after
2527 the body content is received and before the acl_smtp_data ACL.
2528 The client is controlled by both of: a hosts_try_prdr option
2529 on the smtp transport, and the server advertisement.
2530 Default client logging of deliveries and rejections involving
2531 PRDR are flagged with the string "PRDR".
2533 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2534 fclose(). Diagnosis by Todd Lyons.
2536 PP/17 Update configure.default to handle IPv6 localhost better.
2537 Patch by Alain Williams (plus minor tweaks).
2540 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2541 This is now consistent with GnuTLS, and is now documented: the
2542 previous undocumented portable approach to treating the option as
2543 unset was to force an expansion failure. That still works, and
2544 an empty string is now equivalent.
2546 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2547 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2548 not performing validation itself.
2550 PP/20 Added force_command boolean option to pipe transport.
2551 Patch from Nick Koston, of cPanel Inc.
2553 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2556 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2558 PP/21 Fix eximon continuous updating with timestamped log-files.
2559 Broken in a format-string cleanup in 4.80, missed when I repaired the
2560 other false fix of the same issue.
2561 Report and fix from Heiko Schlichting.
2564 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2565 Report from Prashanth Katuri.
2567 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2568 It's SecureTransport, so affects any MacOS clients which use the
2569 system-integrated TLS libraries, including email clients.
2571 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2572 using a MIME ACL for non-SMTP local injection.
2573 Report and assistance in diagnosis by Warren Baker.
2575 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2577 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2579 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2580 last id that failed so it may be referenced in subsequent ACL's.
2582 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2585 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2586 ldap library initialization, allowing self-signed CA's to be
2587 used. Also properly sets require_cert option later in code by
2588 using NULL (global ldap config) instead of ldap handle (per
2589 session). Bug diagnosis and testing by alxgomz.
2591 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2592 the src/util/ subdirectory.
2594 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2595 renamed to Transport Post Delivery Action by Jeremy Harris, as
2598 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2599 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2600 redis_servers = needs to be configured which will be used by the redis
2601 lookup. Patch from Warren Baker, of The Packet Hub.
2603 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2605 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2606 hostname or reverse DNS when processing a host list. Used suggestions
2607 from multiple comments on this bug.
2609 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2611 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2612 Missed a few lines, added it to make the runtest require no keyboard
2615 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2616 contains upper case chars. Make router use caseful_local_part.
2618 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2619 support when GnuTLS has been built with p11-kit.
2625 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2627 This, or similar/improved, will also be change PP/11 of 4.82.
2633 PP/01 Handle short writes when writing local log-files.
2634 In practice, only affects FreeBSD (8 onwards).
2635 Bugzilla 1053, with thanks to Dmitry Isaikin.
2637 NM/01 Bugzilla 949 - Documentation tweak
2639 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2642 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2644 PP/02 Implemented gsasl authenticator.
2646 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2648 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2649 `pkg-config foo` for cflags/libs.
2651 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2652 with rest of GSASL and with heimdal_gssapi.
2654 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2655 `pkg-config foo` for cflags/libs for the TLS implementation.
2657 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2658 properties get this fed in as external SSF. A number of robustness
2659 and debugging improvements to the cyrus_sasl authenticator.
2661 PP/08 cyrus_sasl server now expands the server_realm option.
2663 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2664 Patch by Jeremy Harris.
2666 PP/10 Added dbmjz lookup type.
2668 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2670 PP/12 MAIL args handles TAB as well as SP, for better interop with
2671 non-compliant senders.
2672 Analysis and variant patch by Todd Lyons.
2674 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2675 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2676 Patch from Dirk Mueller <dmueller@suse.com>
2678 PP/13 tls_peerdn now print-escaped for spool files.
2679 Observed some $tls_peerdn in wild which contained \n, which resulted
2680 in spool file corruption.
2682 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2683 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2684 or write after TLS renegotiation, which otherwise led to messages
2687 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2688 as a tracking header (ie: a signed header comes before the signature).
2689 Patch from Wolfgang Breyha.
2691 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2692 comma-sep list; embedded commas doubled.
2694 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2696 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2698 Report and patch from Dmitry Banschikov.
2700 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2701 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2702 failures. We appear to now support TLS1.1+ with Exim.
2704 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2705 lets Exim select keys and certificates based upon TLS SNI from client.
2706 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2707 before an outbound SMTP session. New log_selector, +tls_sni.
2709 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2710 NULL dereference. Report and patch from Alun Jones.
2712 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2713 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2714 Not seeing resolver debug output on NetBSD, but suspect this is a
2715 resolver implementation change.
2717 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2718 Left warnings. Added "eximon gdb" invocation mode.
2720 PP/21 Defaulting "accept_8bitmime" to true, not false.
2722 PP/22 Added -bw for inetd wait mode support.
2724 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2725 locate the relevant includes and libraries. Made this the default.
2727 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2728 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2730 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2731 This may cause build issues on older platforms.
2733 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2734 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2735 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2736 Added SNI support via GnuTLS too.
2737 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2739 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2741 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2742 Applied second patch from Janne, implementing suggestion to default
2743 multiple-strings-in-record handling to match SPF spec.
2745 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2747 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2748 read-only, out of scope).
2749 Patch from Wolfgang Breyha, report from Stuart Northfield.
2751 PP/29 Fix three issues highlighted by clang analyser static analysis.
2752 Only crash-plausible issue would require the Cambridge-specific
2753 iplookup router and a misconfiguration.
2754 Report from Marcin Mirosław.
2756 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2758 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2759 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2760 As part of this, removing so much warning spew let me fix some minor
2761 real issues in debug logging.
2763 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2764 assignment on my part. Fixed.
2766 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2767 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2768 Janne Snabb (who went above and beyond: thank you).
2770 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2771 string otherwise requires a connection and a bunch more work and it's
2772 relatively easy to get wrong. Should also expose TLS library linkage
2775 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2776 64-bit ${eval} (JH/03).
2778 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2779 GNU libc to support some of the 64-bit stuff, should not lead to
2780 conflicts. Defined before os.h is pulled in, so if a given platform
2781 needs to override this, it can.
2783 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2784 protection layer was required, which is not implemented.
2785 Bugzilla 1254, patch from Wolfgang Breyha.
2787 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2788 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2789 tls_dhparam take prime identifiers. Also unbreak combination of
2790 OpenSSL+DH_params+TLSSNI.
2792 PP/39 Disable SSLv2 by default in OpenSSL support.
2798 PP/01 Solaris build fix for Oracle's LDAP libraries.
2799 Bugzilla 1109, patch from Stephen Usher.
2801 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2803 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2806 TF/02 Fix a couple more cases where we did not log the error message
2807 when unlink() failed. See also change 4.74-TF/03.
2809 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2810 lock up or crash if it happened to be inside a call to libc when it
2811 got a SIGUSR1 from exiwhat.
2813 The SIGUSR1 handler appends the current process status to the process
2814 log which is later printed by exiwhat. It used to use the general
2815 purpose logging code to do this, but several functions it calls are
2816 not safe for signals.
2818 The new output code in the SIGUSR1 handler is specific to the process
2819 log, and simple enough that it's easy to inspect for signal safety.
2820 Removing some special cases also simplifies the general logging code.
2821 Removing the spurious timestamps from the process log simplifies
2824 TF/04 Improved ratelimit ACL condition.
2826 The /noupdate option has been deprecated in favour of /readonly which
2827 has clearer semantics. The /leaky, /strict, and /readonly update modes
2828 are mutually exclusive. The update mode is no longer included in the
2829 database key; it just determines when the database is updated. (This
2830 means that when you upgrade Exim will forget old rate measurements.)
2832 Exim now checks that the per_* options are used with an update mode that
2833 makes sense for the current ACL. For example, when Exim is processing a
2834 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2835 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2836 must specify per_mail/readonly. If you omit the update mode it defaults to
2837 /leaky where that makes sense (as before) or /readonly where required.
2839 The /noupdate option is now undocumented but still supported for
2840 backwards compatibility. It is equivalent to /readonly except that in
2841 ACLs where /readonly is required you may specify /leaky/noupdate or
2842 /strict/noupdate which are treated the same as /readonly.
2844 A useful new feature is the /count= option. This is a generalization
2845 of the per_byte option, so that you can measure the throughput of other
2846 aggregate values. For example, the per_byte option is now equivalent
2847 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2849 The per_rcpt option has been generalized using the /count= mechanism
2850 (though it's more complicated than the per_byte equivalence). When it is
2851 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2852 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2853 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2854 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2855 Note that using per_rcpt with a non-readonly update mode in more than
2856 one ACL will cause the recipients to be double-counted. (The per_mail
2857 and per_byte options don't have this problem.)
2859 The handling of very low rates has changed slightly. If the computed rate
2860 is less than the event's count (usually one) then this event is the first
2861 after a long gap. In this case the rate is set to the same as this event's
2862 count, so that the first message of a spam run is counted properly.
2864 The major new feature is a mechanism for counting the rate of unique
2865 events. The new per_addr option counts the number of different
2866 recipients that someone has sent messages to in the last time period. It
2867 behaves like per_rcpt if all the recipient addresses are different, but
2868 duplicate recipient addresses do not increase the measured rate. Like
2869 the /count= option this is a general mechanism, so the per_addr option
2870 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2871 example, measure the rate that a client uses different sender addresses
2872 with the options per_mail/unique=$sender_address. There are further
2873 details in the main documentation.
2875 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2877 TF/06 Removed a few PCRE remnants.
2879 TF/07 Automatically extract Exim's version number from tags in the git
2880 repository when doing development or release builds.
2882 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2883 Bugzilla 879. Patch from Paul Fisher.
2885 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2886 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2889 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2891 PP/05 Variable $av_failed, true if the AV scanner deferred.
2892 Bugzilla 1078. Patch from John Horne.
2894 PP/06 Stop make process more reliably on build failure.
2895 Bugzilla 1087. Patch from Heiko Schlittermann.
2897 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2898 Bugzilla 1089. Patch from Heiko Schlittermann.
2900 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2901 Bugzilla 1131. Patch from Holger Weiß.
2903 PP/09 Handle IPv6 addresses with SPF.
2904 Bugzilla 860. Patch from Wolfgang Breyha.
2906 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2908 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2911 PP/11 match_* no longer expand right-hand-side by default.
2912 New compile-time build option, EXPAND_LISTMATCH_RHS.
2913 New expansion conditions, "inlist", "inlisti".
2915 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2917 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2919 PP/14 fix log_write() format string regression from TF/03.
2920 Bugzilla 1152. Patch from Dmitry Isaikin.
2926 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2928 PP/02 Harmonised TLS library version reporting; only show if debugging.
2929 Layout now matches that introduced for other libraries in 4.74 PP/03.
2931 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2933 PP/04 New "dns_use_edns0" global option.
2935 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2938 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2939 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2941 TK/01 Updated PolarSSL code to 0.14.2.
2942 Bugzilla 1097. Patch from Andreas Metzler.
2944 PP/07 Catch divide-by-zero in ${eval:...}.
2945 Fixes bugzilla 1102.
2947 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2950 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2951 format-string attack -- SECURITY: remote arbitrary code execution.
2953 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2954 time unintentionally subject to list matching rules, letting the header
2955 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2956 arbitrary string expansion). This allowed for information disclosure.
2958 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2959 INT_MIN/-1 -- value coerced to INT_MAX.
2965 NM/01 Workaround for PCRE version dependency in version reporting
2968 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2969 This fixes portability to compilers other than gcc, notably
2970 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2972 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2973 makefiles for portability to HP-UX and POSIX correctness.
2975 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2976 Also via indented variable definition in the Makefile.
2977 (Debugging by Oliver Heesakkers).
2979 PP/02 Restore caching of spamd results with expanded spamd_address.
2980 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2982 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2983 Improves build reliability. Fix from: Frank Elsner
2985 NM/02 Fix wide character breakage in the rfc2047 coding
2986 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2988 NM/03 Allow underscore in dnslist lookups
2989 Fixes bug 1026. Patch from Graeme Fowler
2991 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2992 Code patches from Adam Ciarcinski of NetBSD.
2994 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2997 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2998 is logged, to avoid truncation. Patch from John Horne.
3000 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3001 Patch from Jakob Hirsch.
3003 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3004 SQL string expansion failure details.
3005 Patch from Andrey Oktyabrski.
3007 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3008 Patch from Simon Arlott.
3010 PP/09 New lookups functionality failed to compile on old gcc which rejects
3011 extern declarations in function scope.
3012 Patch from Oliver Fleischmann
3014 PP/10 Use sig_atomic_t for flags set from signal handlers.
3015 Check getgroups() return and improve debugging.
3016 Fixed developed for diagnosis in bug 927 (which turned out to be
3019 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3020 Patch from Mark Zealey.
3022 PP/12 Bugzilla 1056: Improved spamd server selection.
3023 Patch from Mark Zealey.
3025 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3026 Based on patch from Heiko Schlittermann.
3028 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3029 Patch from Uwe Doering, sign-off by Michael Haardt.
3031 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3032 variable declaration deep within a block. Bug and patch from
3035 PP/15 lookups-Makefile IRIX compatibility coercion.
3037 PP/16 Make DISABLE_DKIM build knob functional.
3039 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3040 Patch by Simon Arlott
3042 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3043 variable argument macros. Our copy now differs from upstream.
3049 TF/01 Failure to get a lock on a hints database can have serious
3050 consequences so log it to the panic log.
3052 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3053 controlled using the smtp_confirmation log selector.
3055 TF/03 Include the error message when we fail to unlink a spool file.
3057 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3058 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3059 for maintaining out-of-tree patches for some time.
3061 PP/01 Bugzilla 139: Documentation and portability issues.
3062 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3063 Handle per-OS dynamic-module compilation flags.
3065 PP/02 Let /dev/null have normal permissions.
3066 The 4.73 fixes were a little too stringent and complained about the
3067 permissions on /dev/null. Exempt it from some checks.
3068 Reported by Andreas M. Kirchwitz.
3070 PP/03 Report version information for many libraries, including
3071 Exim version information for dynamically loaded libraries. Created
3072 version.h, now support a version extension string for distributors
3073 who patch heavily. Dynamic module ABI change.
3075 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3076 privilege escalation vulnerability whereby the Exim run-time user
3077 can cause root to append content of the attacker's choosing to
3080 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3083 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3084 If dropping privileges for untrusted macros, we disabled normal logging
3085 on the basis that it would fail; for the Exim run-time user, this is not
3086 the case, and it resulted in successful deliveries going unlogged.
3087 Fixed. Reported by Andreas Metzler.
3093 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3094 only prepend for the Resent-* case. Fixes regression introduced in
3095 Exim 4.70 by NM/22 for Bugzilla 607.
3097 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3098 increasing numbers of administrators be bitten by this.
3100 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3102 PP/03 Bugzilla 994: added openssl_options main configuration option.
3104 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3106 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3108 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3110 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3111 without a peer certificate, leading to a segfault because of an
3112 assumption that peers always have certificates. Be a little more
3113 paranoid. Problem reported by Martin Tscholak.
3115 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3116 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3117 NB: ClamAV planning to remove STREAM in "middle of 2010".
3118 CL also introduces -bmalware, various -d+acl logging additions and
3119 more caution in buffer sizes.
3121 PP/09 Implemented reverse_ip expansion operator.
3123 PP/10 Bugzilla 937: provide a "debug" ACL control.
3125 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3127 PP/12 Bugzilla 973: Implement --version.
3129 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3131 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3133 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3135 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3136 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3137 ignore trailing whitespace.
3139 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3141 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3144 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3145 Notification from Dr Andrew Aitchison.
3147 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3148 ExtendedDetectionInfo response format.
3149 Notification from John Horne.
3151 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3154 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3155 XSL and documented dependency on system catalogs, with examples of how
3158 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3161 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3162 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3163 configuration file which is writeable by the Exim user or group.
3165 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3166 of configuration files to cover files specified with the -C option if
3167 they are going to be used with root privileges, not just the default
3170 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3171 option (effectively making it always true).
3173 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3174 files to be used while preserving root privileges.
3176 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3177 that rogue child processes cannot use them.
3179 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3180 run-time user, instead of root.
3182 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3183 Exim run-time user without dropping privileges.
3185 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3186 result string, instead of calling string_vformat() twice with the same
3189 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3190 for other users. Others should always drop root privileges if they use
3191 -C on the command line, even for a whitelisted configure file.
3193 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3195 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3201 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3202 $data_path, and $header_path variables; fixed documentation bugs and
3205 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3206 exipick to access non-standard spools, including the "frozen" queue
3209 NM/01 Bugzilla 965: Support mysql stored procedures.
3210 Patch from Alain Williams
3212 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3214 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3215 Patch from Andreas Metzler
3217 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3218 Patch from Kirill Miazine
3220 NM/05 Bugzilla 671: Added umask to procmail example.
3222 JJ/03 installed exipick 20100323.0, fixing doc bug
3224 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3225 directory. Notification and patch from Dan Rosenberg.
3227 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3229 TK/02 Improve log output when DKIM signing operation fails.
3231 MH/01 Treat the transport option dkim_domain as a colon separated
3232 list, not as a single string, and sign the message with each element,
3233 omitting multiple occurences of the same signer.
3235 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3236 Bugzilla 985, 986. Patch by Simon Arlott
3238 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3239 Patch by Simon Arlott
3241 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3242 MBX locking. Notification from Dan Rosenberg.
3248 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3250 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3252 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3254 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3256 NM/04 Bugzilla 915: Flag broken perl installation during build.
3262 TK/01 Added patch by Johannes Berg that expands the main option
3263 "spamd_address" if it starts with a dollar sign.
3265 TK/02 Write list of recipients to X-Envelope-Sender header when building
3266 the mbox-format spool file for content scanning (suggested by Jakob
3269 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3270 (http://www.dcc-servers.net/) support via dccifd. Activated by
3271 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3273 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3274 by Mark Daniel Reidel <mr@df.eu>.
3276 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3277 When building exim an external PCRE library is now needed -
3278 PCRE is a system library on the majority of modern systems.
3279 See entry on PCRE_LIBS in EDITME file.
3281 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3282 conversation. Added nologin parameter to request.
3283 Patch contributed by Kirill Miazine.
3285 TF/01 Do not log submission mode rewrites if they do not change the address.
3287 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3289 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3290 log files in place. Contributed by Roberto Lima.
3292 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3294 TF/03 Bugzilla 615: When checking the local_parts router precondition
3295 after a local_part_suffix or local_part_prefix option, Exim now
3296 does not use the address's named list lookup cache, since this
3297 contains cached lookups for the whole local part.
3299 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3300 Robert Millan. Documentation is in experimental-spec.txt.
3302 TF/04 Bugzilla 668: Fix parallel build (make -j).
3304 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3306 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3307 Patch provided by Jan Srzednicki.
3309 TF/05 Leading white space used to be stripped from $spam_report which
3310 wrecked the formatting. Now it is preserved.
3312 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3313 that they are available at delivery time.
3315 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3317 TF/08 TLS error reporting now respects the incoming_interface and
3318 incoming_port log selectors.
3320 TF/09 Produce a more useful error message if an SMTP transport's hosts
3321 setting expands to an empty string.
3323 NM/06 Bugzilla 744: EXPN did not work under TLS.
3324 Patch provided by Phil Pennock.
3326 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3327 Patch provided by Richard Godbee.
3329 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3330 acl_smtp_notquit, added index entry.
3332 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3333 Patch provided by Eugene Bujak.
3335 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3336 accept(). Patch provided by Maxim Dounin.
3338 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3339 Patch provided by Phil Pennock.
3341 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3343 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3344 Patch provided by Brad "anomie" Jorsch.
3346 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3347 Patch provided by Dean Brooks.
3349 TK/05 Add native DKIM support (does not depend on external libraries).
3351 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3352 Patch provided by Graeme Fowler.
3354 NM/16 Bugzilla 851: Documentation example syntax fix.
3356 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3358 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3361 NM/19 Bugzilla 745: TLS version reporting.
3362 Patch provided by Phil Pennock.
3364 NM/20 Bugzilla 167: bool: condition support.
3365 Patch provided by Phil Pennock.
3367 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3368 clients. Patch provided by Phil Pennock.
3370 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3371 Patch provided by Brad "anomie" Jorsch.
3373 NM/23 Bugzilla 687: Fix misparses in eximstats.
3374 Patch provided by Heiko Schlittermann.
3376 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3377 Patch provided by Heiko Schlittermann.
3379 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3380 plus update to original patch.
3382 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3384 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3385 Patch provided by David Brownlee.
3387 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3389 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3391 NM/30 Bugzilla 888: TLS documentation bugfixes.
3393 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3395 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3396 Unlike the original bugzilla I have changed all shell scripts in src tree.
3398 NM/33 Bugzilla 898: Transport filter timeout fix.
3399 Patch by Todd Rinaldo.
3401 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3402 Patch by Serge Demonchaux.
3404 NM/35 Bugzilla 39: Base64 decode bug fixes.
3405 Patch by Jakob Hirsch.
3407 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3409 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3411 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3413 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3419 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3420 ALT-N's libdkim that I have put here:
3421 http://duncanthrax.net/exim-experimental/
3423 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3424 were called 'true' and it seems that C99 defines that as a reserved
3425 keyword to be used with 'bool' variable types. That means you could
3426 not include C99-style headers which use bools without triggering
3427 build errors in sieve.c.
3429 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3430 as mailq or other aliases. Changed the --help handling significantly
3431 to do whats expected. exim_usage() emits usage/help information.
3433 SC/01 Added the -bylocaldomain option to eximstats.
3435 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3437 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3439 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3445 PH/01 Another patch from the Sieve maintainer.
3447 PH/02 When an IPv6 address is converted to a string for single-key lookup
3448 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3449 dots are used instead of colons so that keys in lsearch files need not
3450 contain colons. This was done some time before quoting was made available
3451 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3452 (notated using the quote facility) so as to distinguish them from IPv4
3453 keys. This meant that lookups for IP addresses in host lists did not work
3454 for iplsearch lookups.
3456 This has been fixed by arranging for IPv6 addresses to be expressed with
3457 colons if the lookup type is iplsearch. This is not incompatible, because
3458 previously such lookups could never work.
3460 The situation is now rather anomalous, since one *can* have colons in
3461 ordinary lsearch keys. However, making the change in all cases is
3462 incompatible and would probably break a number of configurations.
3464 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3467 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3468 conversion specification without a maximum field width, thereby enabling
3469 a rogue spamd server to cause a buffer overflow. While nobody in their
3470 right mind would setup Exim to query an untrusted spamd server, an
3471 attacker that gains access to a server running spamd could potentially
3472 exploit this vulnerability to run arbitrary code as the Exim user.
3474 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3475 $primary_hostname instead of what libspf2 thinks the hosts name is.
3477 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3478 a directory entry by the name of the lookup key. Previously, if a
3479 symlink pointed to a non-existing file or a file in a directory that
3480 Exim lacked permissions to read, a lookup for a key matching that
3481 symlink would fail. Now it is enough that a matching directory entry
3482 exists, symlink or not. (Bugzilla 503.)
3484 PH/03 The body_linecount and body_zerocount variables are now exported in the
3487 PH/04 Added the $dnslist_matched variable.
3489 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3490 This means they are set thereafter only if the connection becomes
3493 PH/06 Added the client_condition to authenticators so that some can be skipped
3494 by clients under certain conditions.
3496 PH/07 The error message for a badly-placed control=no_multiline_responses left
3497 "_responses" off the end of the name.
3499 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3501 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3502 (without spaces) instead of just copying the configuration text.
3504 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3506 PH/11 Added $max_received_linelength.
3508 PH/12 Added +ignore_defer and +include_defer to host lists.
3510 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3511 way in which PCRE > 7.0 is built.
3513 PH/14 Implemented queue_only_load_latch.
3515 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3516 MAIL command. The effect was to mangle the value on 64-bit systems.
3518 PH/16 Another patch from the Sieve maintainer.
3520 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3522 PH/18 If a system quota error occurred while trying to create the file for
3523 a maildir delivery, the message "Mailbox is full" was not appended to the
3524 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3525 only to a quota excession during the actual writing of the file.
3527 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3528 characters?) which causes problems in log lines. The DN values are now
3529 passed through string_printing() before being added to log lines.
3531 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3532 and InterBase are left for another time.)
3534 PH/21 Added message_body_newlines option.
3536 PH/22 Guard against possible overflow in moan_check_errorcopy().
3538 PH/23 POSIX allows open() to be a macro; guard against that.
3540 PH/24 If the recipient of an error message contained an @ in the local part
3541 (suitably quoted, of course), incorrect values were put in $domain and
3542 $local_part during the evaluation of errors_copy.
3548 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3549 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3552 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3553 issue a MAIL command.
3555 PH/02 In an ACL statement such as
3557 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3559 if a client was not listed at all, or was listed with a value other than
3560 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3561 the condition was not true (as it should be), so access was not denied.
3562 The bug was that the ! inversion was incorrectly passed on to the second
3563 item. This has been fixed.
3565 PH/03 Added additional dnslists conditions == and =& which are different from
3566 = and & when the dns lookup returns more than one IP address.
3568 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3569 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3571 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3572 FSYNC, which compiles an option called disable_fsync that allows for
3573 bypassing fsync(). The documentation is heavily laced with warnings.
3575 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3577 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3578 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3579 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3580 including adding "make clean"; (3) Added -fPIC when compiling the test
3581 dynamically loaded module, to get rid of a warning.
3583 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3584 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3585 = 0s. The bug is otherwise harmless.
3587 PH/07 There was a bug in the dovecot authenticator such that the value of
3588 $auth1 could be overwritten, and so not correctly preserved, after a
3589 successful authentication. This usually meant that the value preserved by
3590 the server_setid option was incorrect.
3592 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3594 PH/09 Installed PCRE release 7.0.
3596 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3597 run for batched SMTP input. It is now run at the start of every message
3598 in the batch. While fixing this I discovered that the process information
3599 (output by running exiwhat) was not always getting set for -bs and -bS
3600 input. This is fixed, and it now also says "batched" for BSMTP.
3602 PH/11 Added control=no_pipelining.
3604 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3605 patch, slightly modified), and move the expansion of helo_data till after
3606 the connection is made in the smtp transport (so it can use these
3609 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3611 PH/14 Added log_selector = +pid.
3613 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3615 PH/16 Add ${if forany and ${if forall.
3617 PH/17 Added dsn_from option to vary the From: line in DSNs.
3619 PH/18 Flush SMTP output before performing a callout, unless control =
3620 no_callout_flush is set.
3622 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3623 was true (the default) a successful delivery failed to delete the retry
3624 item, thus causing premature timeout of the address. The bug is now
3627 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3629 PH/21 Long custom messages for fakedefer and fakereject are now split up
3630 into multiline responses in the same way that messages for "deny" and
3631 other ACL rejections are.
3633 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3634 with slight modification.
3636 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3637 draft, changing the syntax and factoring some duplicate code".
3639 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3640 for deliveries of the second and subsequent messages over the same SMTP
3643 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3644 ${reduce, with only minor "tidies".
3646 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3648 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3649 expansion side effects.
3651 PH/27 When a message times out after an over-quota error from an Exim-imposed
3652 quota, the bounce message says "mailbox is full". This message was not
3653 being given when it was a system quota that was exceeded. It now should
3656 MH/03 Made $recipients available in local_scan(). local_scan() already has
3657 better access to the recipient list through recipients_list[], but
3658 $recipients can be useful in postmaster-provided expansion strings.
3660 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3661 in the case of a MAIL command with additional options following the
3662 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3663 were accidentally chopped off.
3665 PH/29 SMTP synchronization checks are implemented when a command is read -
3666 there is a check that no more input is waiting when there shouldn't be
3667 any. However, for some commands, a delay in an ACL can mean that it is
3668 some time before the response is written. In this time, more input might
3669 arrive, invalidly. So now there are extra checks after an ACL has run for
3670 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3671 pipelining has not been advertised.
3673 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3675 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3676 returned in the SMTP response when smtp_return_error_details was false.
3677 This has been fixed.
3679 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3680 instead of the C I/O that was originally supplied, because problems were
3681 reported on Solaris.
3683 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3684 Exim which did not show up earlier: it was assuming that a call to
3685 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3686 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3687 was a macro that became an assignment, so it seemed to work. This has
3688 changed to a proper function call with a void return, hence the compile
3689 error. Exim's code has been fixed.
3691 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3694 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3696 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3698 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3699 because it turns out to be too big or there is a timeout) there is no
3700 "Completed" line in the log. When some messages of this type were
3701 selected by exigrep, they were listed as "not completed". Others were
3702 picked up by some special patterns. I have improved the selection
3703 criteria to be more general.
3705 PH/38 The host_find_failed option in the manualroute router can now be set
3706 to "ignore", to completely ignore a host whose IP address cannot be
3707 found. If all hosts are ignored, the behaviour is controlled by the new
3708 host_all_ignored option.
3710 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3711 homing or because of multiple MX records with /mx) generated more than
3712 one IP address, and the following item turned out to be the local host,
3713 all the secondary addresses of the first item were incorrectly removed
3714 from the list, along with the local host and any following hosts (which
3715 is what is supposed to happen).
3717 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3718 whoever called Exim into the -H file. In the case of the daemon it was
3719 behaving confusingly. When first started, it used values for whoever
3720 started the daemon, but after a SIGHUP it used the Exim user (because it
3721 calls itself on a restart). I have changed the code so that it now always
3724 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3725 message are rejected with the same error (e.g. no authentication or bad
3726 sender address), and a DATA command is nevertheless sent (as can happen
3727 with PIPELINING or a stupid MUA), the error message that was given to the
3728 RCPT commands is included in the rejection of the DATA command. This is
3729 intended to be helpful for MUAs that show only the final error to their
3732 PH/42 Another patch from the Sieve maintainer.
3734 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3735 Eximstats - Fixed some broken HTML links and added missing column headers
3737 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3738 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3740 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3742 PH/43 Yet another patch from the Sieve maintainer.
3744 PH/44 I found a way to check for a TCP/IP connection going away before sending
3745 the response to the final '.' that terminates a message, but only in the
3746 case where the client has not sent further data following the '.'
3747 (unfortunately, this is allowed). However, in many cases there won't be
3748 any further data because there won't be any more messages to send. A call
3749 to select() can be used: if it shows that the input is "ready", there is
3750 either input waiting, or the socket has been closed. An attempt to read
3751 the next input character can distinguish the two cases. Previously, Exim
3752 would have sent an OK response which the client would never have see.
3753 This could lead to message repetition. This fix should cure that, at
3754 least in a lot of common cases.
3756 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3757 advertised in response to EHLO.
3763 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3764 fixed by 4.65/MH/01 (is this a record?) are fixed:
3766 (i) An empty string was always treated as zero by the numeric comparison
3767 operators. This behaviour has been restored.
3769 (ii) It is documented that the numeric comparison operators always treat
3770 their arguments as decimal numbers. This was broken in that numbers
3771 starting with 0 were being interpreted as octal.
3773 While fixing these problems I realized that there was another issue that
3774 hadn't been noticed. Values of message_size_limit (both the global option
3775 and the transport option) were treated as octal if they started with 0.
3776 The documentation was vague. These values are now always treated as
3777 decimal, and I will make that clear in the documentation.
3783 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3784 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3787 MH/01 Don't check that the operands of numeric comparison operators are
3788 integers when their expansion is in "skipping" mode (fixes bug
3789 introduced by 4.64-PH/07).
3791 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3792 child addresses, Exim now panics and dies. Previously, because the count
3793 is held in a short int, deliveries were likely to be lost. As such a
3794 large number of recipients for a single message is ridiculous
3795 (performance will be very, very poor), I have chosen to impose a limit
3796 rather than extend the field.
3802 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3803 leftover -K file (the existence of which was triggered by #402).
3804 While we were at it, introduced process PID as part of the -K
3805 filename. This should rule out race conditions when creating
3808 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3809 processing considerably. Previous code took too long for large mails,
3810 triggering a timeout which in turn triggers #401.
3812 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3813 in the DK code in transports.c. sendfile() is not really portable,
3814 hence the _LINUX specificness.
3816 TF/01 In the add_headers option to the mail command in an Exim filter,
3817 there was a bug that Exim would claim a syntax error in any
3818 header after the first one which had an odd number of characters
3821 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3822 callout verification, Exim cached a "reject" for the entire domain. This
3823 is correct for most verifications, but it is not correct for a recipient
3824 verification with use_sender or use_postmaster set, because in that case
3825 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3826 case of MAIL FROM:<> rejection from other early rejections (e.g.
3827 rejection of HELO). When verifying a recipient using a non-null MAIL
3828 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3829 Whatever the result of the callout, the value of the domain cache is
3830 left unchanged (for any other kind of callout, getting as far as trying
3831 RCPT means that the domain itself is ok).
3833 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3836 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3837 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3838 session, the daemon ignores ECONNECT errors and logs others; it now
3839 ignores EPIPE as well.
3841 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3842 (quoted-printable decoding).
3844 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3845 later the small subsequent patch to fix an introduced bug.
3847 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3849 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3851 PH/08 An error is now given if message_size_limit is specified negative.
3853 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3854 to be given (somewhat) arbitrary names.
3856 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3859 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3860 miscellaneous code fixes
3862 PH/10 Added the log_reject_target ACL modifier to specify where to log
3865 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3866 hostname. This is wrong, because it relates to the incoming message (and
3867 probably the interface on which it is arriving) and not to the outgoing
3868 callout (which could be using a different interface). This has been
3869 changed to use the value of the helo_data option from the smtp transport
3870 instead - this is what is used when a message is actually being sent. If
3871 there is no remote transport (possible with a router that sets up host
3872 addresses), $smtp_active_hostname is used.
3874 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3875 tweaks were necessary in order to get it to work (see also 21 below):
3876 (a) The code assumed that strncpy() returns a negative number on buffer
3877 overflow, which isn't the case. Replaced with Exim's string_format()
3879 (b) There were several signed/unsigned issues. I just did the minimum
3880 hacking in of casts. There is scope for a larger refactoring.
3881 (c) The code used strcasecmp() which is not a standard C function.
3882 Replaced with Exim's strcmpic() function.
3883 (d) The code set only $1; it now sets $auth1 as well.
3884 (e) A simple test gave the error "authentication client didn't specify
3885 service in request". It would seem that Dovecot has changed its
3886 interface. Fortunately there's a specification; I followed it and
3887 changed what the client sends and it appears to be working now.
3889 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3892 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3893 address A is aliased to B and C, where B exists and C does not. Without
3894 -v the output is "A verified" because verification stops after a
3895 successful redirection if more than one address is generated. However,
3896 with -v the child addresses are also verified. Exim was outputting "A
3897 failed to verify" and then showing the successful verification for C,
3898 with its parentage. It now outputs "B failed to verify", showing B's
3899 parentage before showing the successful verification of C.
3901 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3902 look up a TXT record in a specific list after matching in a combined
3905 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3906 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3907 they consult the DNS. I had assumed they would set it the way they
3908 wanted; and indeed my experiments on Linux seem to show that in some
3909 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3910 To be on the safe side, however, I have now made the interface to
3911 host_find_byname() similar to host_find_bydns(), with an argument
3912 containing the DNS resolver options. The host_find_byname() function now
3913 sets these options at its start, just as host_find_bydns() does. The smtp
3914 transport options dns_qualify_single and dns_search_parents are passed to
3915 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3916 of host_find_byname() use the default settings of RES_DEFNAMES
3917 (qualify_single) but not RES_DNSRCH (search_parents).
3919 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3920 spool_read_header() do less string testing, by means of a preliminary
3921 switch on the second character of optional "-foo" lines. (This is
3922 overdue, caused by the large number of possibilities that now exist.
3923 Originally there were few.) While I was there, I also converted the
3924 str(n)cmp tests so they don't re-test the leading "-" and the first
3925 character, in the hope this might squeeze out yet more improvement.
3927 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3928 flag allowing group syntax was set by the header_syntax check but not
3929 turned off, possible causing trouble later; (2) The flag was not being
3930 set at all for the header_verify test, causing "group"-style headers to
3931 be rejected. I have now set it in this case, and also caused header_
3932 verify to ignore an empty address taken from a group. While doing this, I
3933 came across some other cases where the code for allowing group syntax
3934 while scanning a header line wasn't quite right (mostly, not resetting
3935 the flag correctly in the right place). These bugs could have caused
3936 trouble for malformed header lines. I hope it is now all correct.
3938 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3939 with the "reply" argument non-NULL. The code, however (which originally
3940 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3941 but it didn't always do it. This confused somebody who was copying the
3942 code for some other use. I have removed all the tests.
3944 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3945 feature that was used to support insecure browsers during the U.S. crypto
3946 embargo. It requires special client support, and Exim is probably the
3947 only MTA that supported it -- and would never use it because real RSA is
3948 always available. This code has been removed, because it had the bad
3949 effect of slowing Exim down by computing (never used) parameters for the
3950 RSA_EXPORT functionality.
3952 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3953 authenticator to fail if there's a tab character in the incoming data
3954 (there should never be unless someone is messing about, as it's supposed
3955 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3956 if the connection is using TLS or if the remote IP is the same as the
3957 local IP, and the "valid-client-cert option" if a client certificate has
3960 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3961 authenticators. This can be used for authorization after authentication
3962 succeeds. (In the case of plaintext, it servers for both authentication
3965 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3966 if any retry times were supplied.
3968 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3969 connection, where there is no client IP address to check. In this
3970 situation, the verify now always succeeds.
3972 PH/25 Applied John Jetmore's -Mset patch.
3974 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3976 PH/27 In a string expansion for a processed (not raw) header when multiple
3977 headers of the same name were present, leading whitespace was being
3978 removed from all of them, but trailing whitespace was being removed only
3979 from the last one. Now trailing whitespace is removed from each header
3980 before concatenation. Completely empty headers in a concatenation (as
3981 before) are ignored.
3983 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3984 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3986 PH/29 [Removed. This was a change that I later backed out, and forgot to
3987 correct the ChangeLog entry (that I had efficiently created) before
3988 committing the later change.]
3990 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3991 address errors (4xx response to RCPT) over the same connection as other
3992 messages routed to the same hosts. Such deliveries are always "forced",
3993 so retry times are not inspected. This resulted in far too many retries
3994 for the affected addresses. The effect occurred only when there were more
3995 hosts than the hosts_max_try setting in the smtp transport when it had
3996 the 4xx errors. Those hosts that it had tried were not added to the list
3997 of hosts for which the message was waiting, so if all were tried, there
3998 was no problem. Two fixes have been applied:
4000 (i) If there are any address or message errors in an SMTP delivery, none
4001 of the hosts (tried or untried) are now added to the list of hosts
4002 for which the message is waiting, so the message should not be a
4003 candidate for sending over the same connection that was used for a
4004 successful delivery of some other message. This seems entirely
4005 reasonable: after all the message is NOT "waiting for some host".
4006 This is so "obvious" that I'm not sure why it wasn't done
4007 previously. Hope I haven't missed anything, but it can't do any
4008 harm, as the worst effect is to miss an optimization.
4010 (ii) If, despite (i), such a delivery is accidentally attempted, the
4011 routing retry time is respected, so at least it doesn't keep
4012 hammering the server.
4014 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4015 in ${readsocket because some servers need this prod.
4017 PH/32 Added some extra debug output when updating a wait-xxx database.
4019 PH/33 The hint "could be header name not terminated by colon", which has been
4020 given for certain expansion errors for a long time, was not being given
4021 for the ${if def:h_colon_omitted{... case.
4023 PH/34 The spec says: "With one important exception, whenever a domain list is
4024 being scanned, $domain contains the subject domain." There was at least
4025 one case where this was not true.
4027 PH/35 The error "getsockname() failed: connection reset by peer" was being
4028 written to the panic log as well as the main log, but it isn't really
4029 panic-worthy as it just means the connection died rather early on. I have
4030 removed the panic log writing for the ECONNRESET error when getsockname()
4033 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4034 runs only) independently of the message's sender address. This meant
4035 that, if the 4xx error was in fact related to the sender, a different
4036 message to the same recipient with a different sender could confuse
4037 things. In particular, this can happen when sending to a greylisting
4038 server, but other circumstances could also provoke similar problems.
4039 I have changed the default so that the retry time for these errors is now
4040 based a combination of the sender and recipient addresses. This change
4041 can be overridden by setting address_retry_include_sender=false in the
4044 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4045 remote server are returned as part of bounce messages. This was not
4046 happening for LMTP over a pipe (the lmtp transport), but now it is the
4047 same for both kinds of LMTP.
4049 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4050 in header lines that were in fact CNAMEs. This is no longer the case.
4052 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4053 and queue runs started by the daemon processed all messages. This has
4054 been fixed so that -R and -S can now usefully be given with -q<time>.
4056 PH/40 Import PCRE release 6.7 (fixes some bugs).
4058 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4060 PH/42 Give an error if -q is specified more than once.
4062 PH/43 Renamed the variables $interface_address and $interface_port as
4063 $received_ip_address and $received_port, to make it clear that these
4064 values apply to message reception, and not to the outgoing interface when
4065 a message is delivered. (The old names remain recognized, of course.)
4067 PH/44 There was no timeout on the connect() call when using a Unix domain
4068 socket in the ${readsocket expansion. There now is.
4070 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4071 be meaningful with "accept".
4073 SC/01 Eximstats V1.43
4074 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4076 SC/02 Eximstats V1.44
4077 Use a glob alias rather than an array ref in the generated
4078 parser. This improves both readability and performance.
4080 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4081 Collect SpamAssassin and rejection statistics.
4082 Don't display local sender or destination tables unless
4083 there is data to show.
4084 Added average volumes into the top table text output.
4086 SC/04 Eximstats V1.46
4087 Collect data on the number of addresses (recipients)
4088 as well as the number of messages.
4090 SC/05 Eximstats V1.47
4091 Added 'Message too big' to the list of mail rejection
4092 reasons (thanks to Marco Gaiarin).
4094 SC/06 Eximstats V1.48
4095 Mainlog lines which have GMT offsets and are too short to
4096 have a flag are now skipped.
4098 SC/07 Eximstats V1.49 (Alain Williams)
4099 Added the -emptyok flag.
4101 SC/08 Eximstats V1.50
4102 Fixes for obtaining the IP address from reject messages.
4104 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4105 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4106 whitespace changes from 4.64-PH/27
4108 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4111 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4112 are found, allow negative numbers in numeric criteria)
4114 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4116 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4117 to match changes made in 4.64-PH/43
4119 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4121 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4122 "domain" is actually an IP address, and force a failure. This locks out
4123 those revolvers/nameservers that support "A-for-A" lookups, in
4124 contravention of the specifications.
4126 PH/48 When a host name was looked up from an IP address, and the subsequent
4127 forward lookup of the name timed out, the host name was left in
4128 $sender_host_name, contrary to the specification.
4130 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4131 restricted to single-key lookups, Exim was not diagnosing an error if
4132 * or *@ was used with a query-style lookup.
4134 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4136 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4137 long ago, but noone interested enough thought of it. Let's just say that
4138 the "1.1" means that there are some new functions that weren't there at
4139 some point in the past.
4141 PH/51 Error processing for expansion failure of helo_data from an smtp
4142 transport during callout processing was broken.
4144 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4145 tested/used via the -bh/-bhc/-bs options.
4147 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4148 bug, fixed in subsequent PCRE releases).
4150 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4151 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4153 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4159 SC/01 Use a glob alias rather than an array ref in eximstats generated
4160 parser. This improves both readability and performance.
4162 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4163 Don't display local sender or destination tables in eximstats unless
4164 there is data to show.
4165 Added average volumes into the eximstats top table text output.
4167 SC/03 Collect data on the number of addresses (recipients) as well
4168 as the number of messages in eximstats.
4170 TF/01 Correct an error in the documentation for the redirect router. Exim
4171 does (usually) call initgroups() when daemonizing.
4173 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4174 with consistent privilege compared to when running as a daemon.
4176 TF/03 Note in the spec that $authenticated_id is not set for local
4177 submissions from trusted users.
4179 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4180 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4182 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4183 by adding some example configuration directives to the default
4184 configuration file. A little bit of work is required to uncomment the
4185 directives and define how usernames and passwords are checked, but
4186 there is now a framework to start from.
4188 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4189 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4190 without this. I don't know how relevant this is to other LDAP libraries.
4192 PH/02 Add the verb name to the "unknown ACL verb" error.
4194 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4196 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4198 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4199 directory not expanded when it should be if an expanded home directory
4200 was set for the address (which is overridden by the transport).
4202 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4205 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4206 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4207 because it is too late at that time, and has no effect.
4209 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4210 security issue with \' (bugzilla #107). I could not use the
4211 PQescapeStringConn() function, because it needs a PGconn value as one of
4214 PH/08 When testing addresses using -bt, indicate those final addresses that
4215 are duplicates that would not cause an additional delivery. At least one
4216 person was confused, thinking that -bt output corresponded to deliveries.
4217 (Suppressing duplicates isn't a good idea as you lose the information
4218 about possibly different redirections that led to the duplicates.)
4220 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4221 systems where poll() doesn't work, in particular OS X.
4223 PH/10 Added more information to debugging output for retry time not reached.
4225 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4226 operations in malware.c.
4228 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4231 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4232 syslog_duplication set false, log lines that would normally be written
4233 both the the main log and to the reject log were not written to syslog at
4236 PH/14 In the default configuration, change the use of "message" in ACL warn
4237 statements to "add_header".
4239 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4240 not followed by a command (e.g. "seen endif").
4242 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4243 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4246 PH/17 Added extra conditions to the default value of delay_warning_condition
4250 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4251 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4252 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4255 The Auto-Submitted: and various List- headers are standardised, whereas I
4256 don't think Precedence: ever was.
4258 PH/18 Refactored debugging code in route_finduser() to show more information,
4259 in particular, the error code if getpwnam() issues one.
4261 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4262 This is apparently needed in addition to the PH/07 change above to avoid
4263 any possible encoding problems.
4265 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4266 but not after initializing Perl.
4268 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4269 output them only if debugging. By default they are written stderr,
4270 apparently, which is not desirable.
4272 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4275 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4278 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4280 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4281 authenticated or an ident call has been made. Suppress the default
4282 values for $authenticated_id and $authenticated_sender (but permit -oMai
4283 and -oMas) when testing with -bh.
4285 PH/24 Re-jigged the order of the tests in the default configuration so that the
4286 tests for valid domains and recipients precede the DNS black list and CSA
4287 tests, on the grounds that those ones are more expensive.
4289 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4290 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4291 This bug exists in every version of Exim that I still have, right back to
4294 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4295 However, an attempt to turn on case-sensitivity in a regex key by
4296 including (?-i) didn't work because the subject string was already
4297 lowercased, and the effects were non-intuitive. It turns out that a
4298 one-line patch can be used to allow (?-i) to work as expected.
4304 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4305 other effects) broke the use of negated acl sub-conditions.
4307 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4310 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4311 "Deny" causes Exim to reject the incoming connection with a 554 error.
4312 Unfortunately, if there is a major crisis, such as a disk failure,
4313 tcp-wrappers gives "deny", whereas what one would like would be some
4314 kind of temporary error. A kludge has been added to help with this.
4315 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4316 554 error is used if errno is still zero or contains ENOENT (which occurs
4317 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4320 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4322 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4323 errors. Otherwise a message that provokes a temporary error (when other
4324 messages do not) can cause a whole host to time out.
4326 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4327 addresses were routed directly to files or pipes from a redirect router.
4328 File deliveries just didn't batch; pipe deliveries might have suffered
4331 PH/06 A failure to get a lock for a hints database would erroneously always say
4332 "Failed to get write lock", even when it was really a read lock.
4334 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4335 of 0600. This has been changed to use the value of the lockfile_mode
4336 option (which defaults to 0600).
4338 PH/08 Applied small patch from the Sieve maintainer.
4340 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4341 folder from quota calculations, a direct delivery into this folder messed
4342 up the contents of the maildirsize file. This was because the regex was
4343 used only to exclude .Trash (or whatever) when the size of the mailbox
4344 was calculated. There was no check that a delivery was happening into an
4345 excluded directory. This bug has been fixed by ignoring all quota
4346 processing for deliveries into excluded directories.
4348 PH/10 Added the maildirfolder_create_regex option to appendfile.
4354 PH/01 The code for finding all the local interface addresses on a FreeBSD
4355 system running IPv6 was broken. This may well have applied to all BSD
4356 systems, as well as to others that have similar system calls. The broken
4357 code found IPv4 interfaces correctly, but gave incorrect values for the
4358 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4359 that it would not match correctly against @[] and not recognize the IPv6
4362 PH/02 The ipliteral router was not recognizing addresses of the form user@
4363 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4365 PH/03 Added disable_ipv6.
4367 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4368 decoded form, because it is most often used to construct To: headers
4369 lines in autoreplies, and the decoded form may well be syntactically
4370 invalid. However, $reply_address has leading white space removed, and all
4371 newlines turned into spaces so that the autoreply transport does not
4374 PH/05 If group was specified without a user on a router, and no group or user
4375 was specified on a transport, the group from the router was ignored.
4377 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4378 for visible compile-time settings that can be used to change these
4379 numbers, for those that want even more. Backwards compatibility with old
4380 spool files has been maintained. However, going back to a previous Exim
4381 release will lost any variables that are in spool files.
4383 PH/07 Two small changes when running in the test harness: increase delay when
4384 passing a TCP/IP connection to a new process, in case the original
4385 process has to generate a bounce, and remove special handling of
4386 127.0.0.2 (sic), which is no longer necessary.
4388 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4389 be the same on different OS.
4391 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4394 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4395 whether --show-vars was specified or not
4397 JJ/02 exipick: Added support for new ACL variable spool format introduced
4400 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4401 syntactically invalid From: or Reply-to: line, and a filter used this to
4402 generate an autoreply, and therefore failed to obtain an address for the
4403 autoreply, Exim could try to deliver to a non-existent relative file
4404 name, causing unrelated and misleading errors. What now happens is that
4405 it logs this as a hard delivery error, but does not attempt to create a
4408 PH/11 The exinext utility has a -C option for testing purposes, but although
4409 the given file was scanned by exinext itself; it wasn't being passed on
4410 when Exim was called.
4412 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4413 an end-of-file indication when reading a command response.
4415 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4416 compiled. In many other places in Exim, IPv6 addresses are always
4417 recognized, so I have changed this. It also means that IPv4 domain
4418 literals of the form [IPV4:n.n.n.n] are now always recognized.
4420 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4421 used if the router is not running as root, for example, when verifying at
4422 ACL time, or when using -bh. The debugging output from this situation was
4423 non-existent - all you got was a failure to exec. I have made two
4426 (a) Failures to set uid/gid, the current directory, or a process leader
4427 in a subprocess such as that created by queryprogram now generate
4428 suitable debugging output when -d is set.
4430 (b) The queryprogram router detects when it is not running as root,
4431 outputs suitable debugging information if -d is set, and then runs
4432 the subprocess without attempting to change uid/gid.
4434 PH/15 Minor change to Makefile for building test_host (undocumented testing
4437 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4438 additional section of a DNS packet that returns MX or SRV records.
4439 Instead, it always explicitly searches for A/AAAA records. This avoids
4440 major problems that occur when a DNS server includes only records of one
4441 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4442 fixed another bug: if SRV records were looked up and the corresponding
4443 address records were *not* found in the additional section, the port
4444 values from the SRV records were lost.
4446 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4447 using the correct key (the original address) when searching the retry
4448 rules in order to find which one to use for generating the retry hint.
4450 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4451 adding the default one. Similarly, if it contains a Reply-To: header, the
4452 errors_reply_to option, if set, is not used.
4454 PH/19 When calculating a retry time, Exim used to measure the "time since
4455 failure" by looking at the "first failed" field in the retry record. Now
4456 it does not use this if it is later than than the arrival time of the
4457 message. Instead it uses the arrival time. This makes for better
4458 behaviour in cases where some deliveries succeed, thus re-setting the
4459 "first failed" field. An example is a quota failure for a huge message
4460 when small messages continue to be delivered. Without this change, the
4461 "time since failure" will always be short, possible causing more frequent
4462 delivery attempts for the huge message than are intended.
4463 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4465 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4466 $1, $2, $3) because the numerical variables can be reset during some
4467 expansion items (e.g. "match"), thereby losing the authentication data.
4469 PH/21 Make -bV show the size of off_t variables so that the test suite can
4470 decide whether to run tests for quotas > 2G.
4472 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4473 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4474 filecount value is greater than 2G or if a quota value is greater than 2G
4475 on a system where the size of off_t is not greater than 4, a panic error
4478 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4479 never match. The debug and -bh output now contains an explicit error
4480 message indicating a malformed IPv4 address or mask.
4482 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4483 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4484 PH/23 above applies.
4486 PH/25 Do not write to syslog when running in the test harness. The only
4487 occasion when this arises is a failure to open the main or panic logs
4488 (for which there is an explicit test).
4490 PH/26 Added the /no_tell option to "control=freeze".
4492 PH/27 If a host name lookup failed very early in a connection, for example, if
4493 the IP address matched host_lookup and the reverse lookup yielded a name
4494 that did not have a forward lookup, an error message of the form "no IP
4495 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4496 could be logged. Now it outputs the IP address instead of "NULL".
4498 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4499 allows the sender and the authenticated sender to be set when
4500 submitting a message from within Exim. Since child_open_exim() is
4501 documented for local_scan(), the new function should be too.
4503 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4504 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4505 results in an empty string is now treated as unset.
4507 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4509 PH/31 Added sender_verify_fail logging option.
4511 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4512 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4513 tidied the source and removed it altogether.
4515 PH/33 When a queue run was abandoned because the load average was too high, a
4516 log line was always written; now it is written only if the queue_run log
4517 selector is set. In addition, the log line for abandonment now contains
4518 information about the queue run such as the pid. This is always present
4519 in "start" and "stop" lines but was omitted from the "abandon" line.
4521 PH/34 Omit spaces between a header name and the colon in the error message that
4522 is given when verify = headers_syntax fails (if there are lots of them,
4523 the message gets confusing).
4525 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4526 names, as there are now some PTR records that contain slashes. This check
4527 is only to protect against broken name servers that fall over on strange
4528 characters, so the fact that it applies to all lookups doesn't matter.
4530 PH/36 Now that the new test suite is complete, we can remove some of the
4531 special code in Exim that was needed for the old test suite. For example,
4532 sorting DNS records because real resolvers return them in an arbitrary
4533 order. The new test suite's fake resolver always returns records in the
4536 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4537 bounces) except when queue_only is set, to avoid logging races between
4538 the different processes.
4540 PH/38 Panic-die if .include specifies a non-absolute path.
4542 PH/39 A tweak to the "H" retry rule from its user.
4544 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4545 a label. They prevented compilation on older perls.
4547 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4548 a warning to be raised on newish perls.
4550 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4551 on queue. Changes to match documented behaviour of showing count of
4552 messages matching specified criteria.
4554 PH/40 Changed the default ident timeout from 30s to 5s.
4556 PH/41 Added support for the use of login_cap features, on those BSD systems
4557 that have them, for controlling the resources used by pipe deliveries.
4559 PH/42 The content-scanning code uses fopen() to create files in which to put
4560 message data. Previously it was not paying any attention to the mode of
4561 the files. Exim runs with umask(0) because the rest of the code creates
4562 files with open(), and sets the required mode explicitly. Thus, these
4563 files were ending up world-writeable. This was not a big issue, because,
4564 being within the spool directory, they were not world-accessible. I have
4565 created a function called modefopen, which takes an additional mode
4566 argument. It sets umask(777), creates the file, chmods it to the required
4567 mode, then resets the umask. All the relevant calls to fopen() in the
4568 content scanning code have been changed to use this function.
4570 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4571 to 24 hours. This avoids potential overflow problems when processing G
4572 and H retry rules. I suspect nobody ever tinkers with this value.
4574 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4576 PH/45 When the plaintext authenticator is running as a client, the server's
4577 challenges are checked to ensure they are valid base64 strings. By
4578 default, the authentication attempt is cancelled if an invalid string is
4579 received. Setting client_ignore_invalid_base64 true ignores these errors.
4580 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4581 they are received. Thus, the responses can be made to depend on the
4582 challenges. If an invalid string is ignored, an empty string is placed in
4585 PH/46 Messages that are created by the autoreply transport now contains a
4586 References: header, in accordance with RFCs 2822 and 3834.
4588 PH/47 Added authenticated_sender_force to the smtp transport.
4590 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4592 PH/49 Installed latest patch from the Sieve maintainer.
4594 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4595 was also set, the appendfile transport was unnecessarily scanning a
4596 directory of message files (e.g. for maildir delivery) to find the count
4597 of files (along with the size), even though it did not need this
4598 information. It now does the scan only if it needs to find either the
4599 size of the count of files.
4601 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4603 PH/52 Two bugs concerned with error handling when the smtp transport is
4606 (i) Exim was not creating retry information for temporary errors given
4607 for individual recipients after the DATA command when the smtp transport
4608 was used in LMTP mode. This meant that they could be retried too
4609 frequently, and not timed out correctly.
4611 (ii) Exim was setting the flag that allows error details to be returned
4612 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4613 recipients that were returned after the DATA command.
4615 PH/53 This is related to PH/52, but is more general: for any failing address,
4616 when detailed error information was permitted to be returned to the
4617 sender, but the error was temporary, then after the final timeout, only
4618 "retry timeout exceeded" was returned. Now it returns the full error as
4619 well as "retry timeout exceeded".
4621 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4622 do this, and (what is worse) MTAs that accept it.
4624 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4625 will now be deprecated.
4627 PH/56 New os.c-cygwin from the Cygwin maintainer.
4629 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4630 formats (previously only available in exim formats via -bpr, -bpru,
4631 and -bpra. Now also available in native and exiqgrep formats)
4633 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4634 with very large, slow to parse queues
4636 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4638 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4640 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4641 responses to authentication challenges, though it was showing the
4642 challenges; (ii) I've removed the CR characters from the debug output for
4645 PH/58 Allow for the insertion of a newline as well as a space when a string
4646 is turned into more than one encoded-word during RFC 2047 encoding. The
4647 Sieve code now uses this.
4649 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4650 data_4xx, lost_connection, tls_required.
4652 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4653 message was being sent as an SMTP response.
4655 PH/61 Add -l and -k options to exicyclog.
4657 PH/62 When verifying, if an address was redirected to one new address, so that
4658 verification continued, and the new address failed or deferred after
4659 having set something in $address_data, the value of $address_data was not
4660 passed back to the ACL. This was different to the case when no
4661 redirection occurred. The value is now passed back in both cases.
4663 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4664 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4665 use login_cap.h, so on its own it isn't the distinguishing feature. The
4666 new name refers directly to the setclassresources() function.
4668 PH/65 Added configuration files for NetBSD3.
4670 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4672 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4673 is preferred over IPv4.
4675 PH/68 The bounce_return_message and bounce_return_body options were not being
4676 honoured for bounces generated during the reception of non-SMTP messages.
4677 In particular, this applied to messages rejected by the ACL. This bug has
4678 been fixed. However, if bounce_return_message is true and bounce_return_
4679 body is false, the headers that are returned for a non-SMTP message
4680 include only those that have been read before the error was detected.
4681 (In the case of an ACL rejection, they have all been read.)
4683 PH/69 The HTML version of the specification is now built in a directory called
4684 spec_html instead of spec.html, because the latter looks like a path with
4685 a MIME-type, and this confuses some software.
4687 PH/70 Catch two compiler warnings in sieve.c.
4689 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4690 function verify_get_ident() calls ip_connect() to connect a socket, but
4691 if the "connect()" function timed out, ip_connect() used to close the
4692 socket. However, verify_get_ident() also closes the socket later, and in
4693 between Exim writes to the log, which may get opened at this point. When
4694 the socket was closed in ip_connect(), the log could get the same file
4695 descriptor number as the socket. This naturally causes chaos. The fix is
4696 not to close the socket in ip_connect(); the socket should be closed by
4697 the function that creates it. There was only one place in the code where
4698 this was missing, in the iplookup router, which I don't think anybody now
4699 uses, but I've fixed it anyway.
4701 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4702 well as to direct DNS lookups. Otherwise the handling of names in host
4703 lists is inconsistent and therefore confusing.
4709 PH/01 Two changes to the default runtime configuration:
4711 (1) Move the checks for relay_from_hosts and authenticated clients from
4712 after to before the (commented out) DNS black list checks.
4714 (2) Add control=submission to the relay_from_hosts and authenticated
4715 clients checks, on the grounds that messages accepted by these
4716 statements are most likely to be submissions.
4718 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4720 (1) Generate an error if the third argument for the ${prvs expansion is
4723 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4726 (3) Reset the variables that are obtained from the first argument of
4727 ${prvscheck and used in the second argument before leaving the code,
4728 because their memory is reclaimed, so using them afterwards may do
4731 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4732 one (it's much easier than Tom thought :-).
4734 (5) Because of (4), we can now allow for the use of $prvscheck_result
4735 inside the third argument.
4737 PH/03 For some reason, the default setting of PATH when running a command from
4738 a pipe transport was just "/usr/bin". I have changed it to
4741 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4742 anything to be listed in the output from -bV.
4744 PH/05 When a filter generated an autoreply, the entire To: header line was
4745 quoted in the delivery log line, like this:
4747 => >A.N.Other <ano@some.domain> <original@ddress> ...
4749 This has been changed so that it extracts the operative address. There
4750 may be more than one such address. If so, they are comma-separated, like
4753 => >ano@some.domain,ona@other.domain <original@ddress> ...
4755 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4756 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4757 not being looked up in the rDNS to get a host name, Exim was showing the
4758 IP address twice in Received: lines, even though the IP addresses were
4759 identical. For example:
4761 Received: from [1.2.3.4] (helo=[1.2.3.4])
4763 However, if the real host name was known, it was omitting the HELO data
4764 if it matched the actual IP address. This has been tidied up so that it
4765 doesn't show the same IP address twice.
4767 PH/07 When both +timestamp and +memory debugging was on, the value given by
4768 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4769 called by the debug printing, thereby overwriting the timestamp buffer.
4770 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4772 PH/08 When the original message was included in an autoreply transport, it
4773 always said "this is a copy of the message, including all the headers",
4774 even if body_only or headers_only was set. It now gives an appropriate
4777 PH/09 Applied a patch from the Sieve maintainer which:
4779 o fixes some comments
4780 o adds the (disabled) notify extension core
4781 o adds some debug output for the result of if/elsif tests
4782 o points to the current vacation draft in the documentation
4783 and documents the missing references header update
4787 o fixes a bug in processing the envelope test (when testing
4788 multiple envelope elements, the last element determined the
4791 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4792 Electronic Mail") by including:
4794 Auto-submitted: auto-generated
4796 in the messages that it generates (bounce messages and others, such as
4797 warnings). In the case of bounce messages for non-SMTP messages, there was
4798 also a typo: it was using "Auto_submitted" (underscore instead of
4799 hyphen). Since every message generated by Exim is necessarily in response
4800 to another message, thes have all been changed to:
4802 Auto-Submitted: auto-replied
4804 in accordance with these statements in the RFC:
4806 The auto-replied keyword:
4808 - SHOULD be used on messages sent in direct response to another
4809 message by an automatic process,
4811 - MUST NOT be used on manually-generated messages,
4813 - MAY be used on Delivery Status Notifications (DSNs) and Message
4814 Disposition Notifications (MDNs),
4816 - MUST NOT be used on messages generated by automatic or periodic
4817 processes, except for messages which are automatic responses to
4820 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4821 to the default Received: header definition.
4823 PH/12 Added log selector acl_warn_skipped (default on).
4825 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4826 variables because (a) they are in the wrong storage pool and (b) even if
4827 they were copied, it wouldn't work properly because of the caching.
4829 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4830 checking when decoding. Apparently there are clients that generate
4831 overlong encoded strings. Why am I not surprised?
4833 PH/15 If the first argument of "${if match_address" was not empty, but did not
4834 contain an "@" character, Exim crashed. Now it writes a panic log message
4835 and treats the condition as false.
4837 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4839 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4840 extension "envelope-auth". The code is finished and in agreement with
4841 other implementations, but there is no documentation so far and in fact,
4842 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4843 not changing the active code.
4845 Print executed "if" and "elsif" statements when debugging is used. This
4846 helps a great deal to understand what a filter does.
4848 Document more things not specified clearly in RFC3028. I had all this
4849 sorted out, when out of a sudden new issues came to my mind. Oops."
4851 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4854 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4855 canonical form (as documented). However, after a host name lookup from
4856 the IP address, check_host() was doing a simple string comparison with
4857 addresses acquired from the DNS when checking that the found name did
4858 have the original IP as one of its addresses. Since any found IPv6
4859 addresses are likely to be in abbreviated form, the comparison could
4860 fail. Luckily, there already exists a function for doing the comparison
4861 by converting both addresses to binary, so now that is used instead of
4862 the text comparison.
4864 PH/20 There was another similar case to PH/19, when a complete host name was
4865 given in a host list; looking up its IP address could give an abbreviated
4866 form, whereas the current host's name might or might not be abbreviated.
4867 The same fix has been applied.
4873 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4874 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4877 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4878 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4880 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4882 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4883 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4884 Also, for "Auto-Submitted", treat anything other than "no" as
4885 non-personal, in accordance with RFC 3834. (Previously it treated
4886 anything starting "auto-" as non-personal.)
4888 TF/01 The control=submission/name=... option had a problem with syntax
4889 errors if the name included a slash character. The /name= option
4890 now slurps the rest of the string, so it can include any characters
4891 but it must come last in the list of options (after /sender_retain
4894 PH/05 Some modifications to the interface to the fake nameserver for the new
4902 TK/01 Added the "success_on_redirect" address verification option. See
4903 NewStuff for rationale and an example.
4905 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4907 PH/02 Patch to exigrep to allow it to work on syslog lines.
4909 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4910 fread() to skip over the body file's header line, because in Cygwin the
4911 header line is locked and is inaccessible.
4913 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4914 co-exist for some time) to make it clear that it is the Exim ID that is
4915 referenced, not the Message-ID: header line.
4917 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4918 string_format() function, because snprintf() does not exist on all
4921 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4922 ${stat: expansion item.
4924 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4925 protocol synchronization error", to keep the pedants happy.
4927 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4928 well as for IRIX systems, when gcc is being used. See the host.c source
4931 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4933 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4936 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4937 transport and to the smtp transport in LMTP mode.
4939 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4941 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4942 run a filter in a subprocess. This could lead to confusion in subsequent
4943 lookups in the parent process. There should also be a search_tidyup() at
4944 the end of the subprocess.
4946 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4947 only if the host matched helo_try_verify_hosts, which caused the
4948 verification to occur when the EHLO/HELO command was issued. The ACL just
4949 tested the remembered result. Now, if a previous verification attempt has
4950 not happened, "verify = helo" does it there and then.
4952 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4954 TK/03 Fix log output including CR from clamd.
4956 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4957 exist provoked a memory error which could cause a segfault.
4959 PH/15 Installed PCRE 6.2
4961 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4963 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4964 of the problem. Specifically, suggested +O2 rather than +O1 for the
4967 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4969 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4970 record, the port was not correctly logged when the outgoing_port log
4971 selector was set (it logged the transort's default port).
4973 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4974 fallback_hosts, and "hosts" in the smtp transport.
4976 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4977 host errors such as "Connection refused".
4979 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4980 authentication with radiusclient 0.4.9:
4982 - Error returned from rc_read_config was caught wrongly
4983 - Username/password not passed on to radius server due to wrong length.
4985 The presumption is that some radiusclient API changes for 4.51/PH/17
4986 were not taken care of correctly. The code is still untested by me (my
4987 Linux distribution still has 0.3.2 of radiusclient), but it was
4988 contributed by a Radius user.
4990 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4991 expanding the "port" option of the smtp transport.
4993 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4994 while reading a MIME header. Thanks to Tom Hughes for a patch.
4996 PH/24 Include config.h inside local_scan.h so that configuration settings are
4999 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5000 that in an ACL for RCPT (for example), you can examine exactly what was
5003 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5004 commands, but it was not correctly comparing the address with the actual
5005 client host address. Thus, it would show the EHLO address in Received:
5006 header lines when this was not necessary.
5008 PH/27 Added the % operator to ${eval:}.
5010 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5011 it should be ignoring failures (because with -C, for example, it has lost
5012 privilege). It wasn't ignoring creation failures other than "already
5015 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5018 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5019 because an input line was too long, either on its own, or by virtue of
5020 too many continuations, the temporary file was not being removed, and the
5021 return code was incorrect.
5023 PH/31 Missing "BOOL" in function definition in filtertest.c.
5025 PH/32 Applied Sieve patches from the maintainer.
5027 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5029 PH/33 Added "verify = not_blind".
5031 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5032 Local/Makefile (with some defaults set). These are used in built scripts
5033 such as exicyclog, but they have never been used in the exim_install
5034 script (though there are many overriding facilities there). I have
5035 arranged that the exim_install script now takes note of these two
5038 PH/35 Installed configuration files for Dragonfly.
5040 PH/36 When a locally submitted message by a trusted user did not contain a
5041 From: header, and the sender address was obtained from -f or from an SMTP
5042 MAIL command, and the trusted user did not use -F to supply a sender
5043 name, $originator_name was incorrectly used when constructing a From:
5044 header. Furthermore, $originator_name was used for submission mode
5045 messages from external hosts without From: headers in a similar way,
5046 which is clearly wrong.
5048 PH/37 Added control=suppress_local_fixups.
5050 PH/38 When log_selector = +received_sender was set, and the addition of the
5051 sender made the log line's construction buffer exactly full, or one byte
5052 less than full, an overflow happened when the terminating "\n" was
5055 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5056 when the result of a list match is failure because a DNS lookup failed.
5058 PH/40 RM_COMMAND is now used in the building process.
5060 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5061 the "build-* directories that it finds.
5063 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5064 address in a domain literal was a prefix of an interface address.
5066 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5067 when verifying a sender address, unless rewrite_headers is false.
5069 PH/44 Wrote a long comment about why errors_to addresses are verified as
5070 recipients, not senders.
5072 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5073 the ratelimit ACL was added.
5075 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5077 PH/46 Added extra information about PostgreSQL errors to the error string.
5079 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5080 avoiding the need to install special zones in a real server. This is
5081 backwards compatible; if it can't find the fake resolver, it drops back.
5082 Thus, both old and new test suites can be run.
5084 TF/02 Added util/ratelimit.pl
5086 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5087 clock is set back in time.
5089 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5090 Candler <B.Candler@pobox.com>.
5092 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5093 OK for addresses that are the result of redirections.
5095 PH/48 A number of further additions for the benefit of the new test suite,
5096 including a fake gethostbyname() that interfaces to the fake DNS resolver
5099 TF/06 The fix for widen_domains has also been applied to qualify_single and
5100 search_parents which are the other dnslookup options that can cause
5103 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5106 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5108 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5109 service" if it encountered an MX record with an empty target hostname.
5110 The message is now "an MX or SRV record indicated no SMTP service".
5112 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5113 if the DNS of the sending site is misconfigured. This is quite a
5114 common situation. This change restores the behaviour of treating a
5115 helo verification defer as a failure.
5117 PH/51 If self=fail was set on a router, the bounce message did not include the
5118 actual error message.
5124 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5126 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5127 command itself ended in error, the underlying message about the transport
5128 filter timeout was being overwritten with the pipe command error. Now the
5129 underlying error message should be appended to the second error message.
5131 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5133 PH/02 Reduce the amount of output that "make" produces by default. Full output
5134 can still be requested.
5136 PH/03 The warning log line about a condition test deferring for a "warn" verb
5137 was being output only once per connection, rather than after each
5138 occurrence (because it was using the same function as for successful
5139 "warn" verbs). This seems wrong, so I have changed it.
5141 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5142 it should not have, which might have caused a crash in the right
5143 circumstances, but probably never did.
5145 PH/04 Installed a modified version of Tony Finch's patch to make submission
5146 mode fix the return path as well as the Sender: header line, and to
5147 add a /name= option so that you can make the user's friendly name appear
5150 TF/03 Added the control = fakedefer ACL modifier.
5152 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5153 Mark Lowes for thorough testing.
5155 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5157 TK/03 Merged latest SRS patch from Miles Wilton.
5159 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5160 with the definition in sysexits.h (which is #included earlier).
5161 Fortunately, Exim does not actually use EX_OK. The code used to try to
5162 preserve the sysexits.h value, by assuming that macro definitions were
5163 scanned for macro replacements. I have been disabused of this notion,
5164 so now the code just undefines EX_OK before #including unistd.h.
5166 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5167 in the smtp transport. When a block could not be written in a single
5168 write() function, the timeout was being re-applied to each part-write.
5169 This seems wrong - if the receiver was accepting one byte at a time it
5170 would take for ever. The timeout is now adjusted when this happens. It
5171 doesn't have to be particularly precise.
5173 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5174 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5176 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5177 without a domain if the check to <postmaster@domain> fails.
5179 SC/01 Eximstats: added -xls and the ability to specify output files
5180 (patch written by Frank Heydlauf).
5182 SC/02 Eximstats: use FileHandles for outputting results.
5184 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5186 SC/04 Eximstats: fixed display of large numbers with -nvr option
5188 SC/05 Eximstats: fixed merging of reports with empty tables.
5190 SC/06 Eximstats: added the -include_original_destination flag
5192 SC/07 Eximstats: removed tabs and trailing whitespace.
5194 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5196 TK/06 MBOX spool code: Add real "From " MBOX separator line
5197 so the .eml file is really in mbox format (even though
5198 most programs do not really care). Patch from Alex Miller.
5200 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5201 The latter is generated from $received_to and is only set if the
5202 message has one envelope recipient. SA can use these headers,
5203 obviously out-of-the-box. Patch from Alex Miller.
5205 PH/08 The ${def test on a variable was returning false if the variable's
5206 value was "0", contrary to what the specification has always said!
5207 The result should be true unless the variable is empty.
5209 PH/09 The syntax error of a character other than { following "${if
5210 def:variable_name" (after optional whitespace) was not being diagnosed.
5211 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5212 accidental colon was present, for example, could give incorrect results.
5214 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5215 result is used (not including appendfile, where other changes are about
5218 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5219 This involved changing a lot of size variables from int to off_t. It
5220 should work with maildirs and everything.
5222 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5223 spamd dying while we are connected to it.
5225 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5228 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5229 function for BDB 4.3.
5231 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5233 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5234 the sources for PCRE have been re-arranged and more files are now
5237 PH/15 The code I had for printing potentially long long variables in PH/11
5238 above was not the best (it lost precision). The length of off_t variables
5239 is now inspected at build time, and an appropriate printing format (%ld
5240 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5241 to be "long long int" or "long int". This is needed for the internal
5242 formatting function string_vformat().
5244 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5245 the configuration file to be ":syslog", then the script "guesses" where
5246 the logs files are, rather than using the compiled in default. In our
5247 case the guess is not the same as the compiled default, so the script
5248 suddenly stopped working when I started to use syslog. The patch checks
5249 to see if log_file_path is "". If so, it attempts to read it from exim
5250 with no configuration file to get the compiled in version, before it
5251 falls back to the previous guessing code."
5253 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5254 implementing BATV in an Exim configuration. See NewStuff for the gory
5257 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5258 Makefile that are specific to HP-UX.
5260 PH/18 If the "use_postmaster" option was set for a recipient callout together
5261 with the "random" option, the postmaster address was used as the MAIL
5262 FROM address for the random test, but not for the subsequent recipient
5263 test. It is now used for both.
5265 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5266 patch removes a few documentation additions to RFC 3028, because the
5267 latest draft now contains them. It adds the new en;ascii-case comparator
5268 and a new error check for 8bit text in MIME parts. Comparator and
5269 require names are now matched exactly. I enabled the subaddress
5270 extension, but it is not well tested yet (read: it works for me)."
5272 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5273 rework some of the code of TK/09 above to avoid the hardwired use of
5274 "%lld" and "long long". Replaced the call to snprintf() with a call to
5277 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5278 records point to non-existent hosts", "retry timeout exceeded", and
5279 "retry time not reached for any host after a long failure period".
5281 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5282 experimental DomainKeys support:
5284 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5285 (2) On an error such as an illegally used "control", the wrong name for
5286 the control was given.
5288 These problems did NOT occur unless DomainKeys support was compiled.
5290 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5292 PH/24 Added ${if match_ip condition.
5294 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5295 databases so that it will be absolutely obvious if a crash occurs in the
5296 DB library. This is a regular occurrence (often caused by mis-matched
5299 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5300 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5301 picked up on a user's system that detects such things. There doesn't seem
5302 to be a gcc warning option for this - only an attribute that has to be
5303 put on the function's prototype. It seems that in Fedora Core 4 they have
5304 set this on a number of new functions. No doubt there will be more in due
5307 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5308 specify a transport. However, if an address that was verified by such a
5309 router was the subject of a callout, Exim crashed because it tried to
5310 read the rcpt_include_affixes from the non-existent transport. Now it
5311 just assumes that the setting of that option is false. This bug was
5312 introduced by 4.51/PH/31.
5314 PH/28 Changed -d+all to exclude +memory, because that information is very
5315 rarely of interest, but it makes the output a lot bigger. People tend to
5316 do -d+all out of habit.
5318 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5319 code in os-type was giving problems when libc.so lives in lib64, like on
5322 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5323 aren't the modern standard, and it seems that some systems' include files
5324 don't always have them. Exim was already checking for some of the newer
5325 ones like T_AAAA, and defining it itself. I've added checks for all the
5326 record types that Exim uses.
5328 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5329 not automatically generating a new one, as it is supposed to. This
5330 prevented TLS from working. If the file did exist, but contained invalid
5331 data, a new version was generated, as expected. It was only the case of a
5332 non-existent file that was broken.
5334 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5335 with a change in libdomainkeys > 0.64.
5337 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5338 from DNS. If the selector record carries the flag, it now has
5339 precedence over the domain-wide flag.
5341 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5343 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5344 the use of a port name that isn't defined in /etc/services) occurred, the
5345 message was deferred as in a normal delivery, and thus remained on the
5346 spool, instead of being failed because of the mua_wrapper setting. This
5347 is now fixed, and I tidied up some of the mua_wrapper messages at the
5350 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5351 the messages in a hash of arrays rather than using individual hashes.
5352 This is a bit cleaner and results in dramatic memory savings, albeit
5353 at a slight CPU cost.
5355 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5356 as requested by Marc Sherman.
5358 SC/10 Eximstats: added histograms for user specified patterns as requested
5361 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5363 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5364 fopen() in the content-scanning modules that did not already have it.
5370 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5371 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5373 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5375 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5377 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5378 to test Sieve filters that use "vacation".
5380 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5381 that changes the way the GnuTLS parameters are stored in the cache file.
5382 The new format can be generated externally. For backward compatibility,
5383 if the data in the cache doesn't make sense, Exim assumes it has read an
5384 old-format file, and it generates new data and writes a new file. This
5385 means that you can't go back to an older release without removing the
5388 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5389 work if there are any delivery delays because "one_time" forces the
5390 parent to be marked "delivered", so its unseen clone is never tried
5391 again. For this reason, Exim now forbids the simultaneous setting of
5394 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5395 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5396 entry if you want to know the details. The fix, however, neglected to
5397 consider the case when local delivery batching is involved. The test for
5398 "previously delivered" was not happening when checking to see if an
5399 address could be batched with a previous (undelivered) one; under
5400 certain circumstances this could lead to multiple deliveries to the same
5403 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5404 in its include files, and this causes problems building Exim.
5406 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5407 header_syntax) but Exim was just ignoring anything given after a slash.
5408 In particular, this caused confusion with an attempt to use "verify =
5409 reverse_host_lookup/defer_ok". An error is now given when options are
5410 supplied for verify items that do not have them. (Maybe reverse_host_
5411 lookup should have a defer_ok option, but that's a different point.)
5413 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5414 defined by RFC 821) to 2048, because there were problems with some AUTH
5415 commands, and RFC 1869 says the size should be increased for extended
5416 SMTP commands that take arguments.
5418 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5421 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5422 "unknown" error; now it says that the functionality isn't in the binary.
5424 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5425 an address' error message when a string expansion fails (syntax or
5426 whatever). Otherwise the password may appear in the log. Following change
5427 PH/42 below, there is no longer a chance of it appearing in a bounce
5430 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5432 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5433 of its addresses was ever tried. (Bugzilla bug #2.)
5435 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5436 the result incorrectly in the debug output. (It correctly added a newline
5437 to what was transported.)
5439 TF/01 Added $received_time.
5441 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5442 commented out examples of how to interface to a virus scanner and to
5443 SpamAssassin. Also added commented examples of av_scanner and
5444 spamd_address settings.
5446 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5447 and controls are allowed in which ACLs. There were a couple of minor
5448 errors. Some of the entries in the conditions table (which is a table of
5449 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5450 negation of where the condition IS allowed.
5452 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5454 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5455 header file does not have a version number, so I've had to invent a new
5456 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5457 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5458 radiusclient), but it was contributed by a Radius user.
5460 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5461 files or directories, for OpenSSL.
5463 PH/19 When an Exim process that is running as root has to create an Exim log
5464 file, it does so in a subprocess that runs as exim:exim so as to get the
5465 ownership right at creation (otherwise, other Exim processes might see
5466 the file with the wrong ownership). There was no test for failure of this
5467 fork() call, which would lead to the process getting stuck as it waited
5468 for a non-existent subprocess. Forks do occasionally fail when resources
5469 run out. I reviewed all the other calls to fork(); they all seem to check
5472 PH/20 When checking for unexpected SMTP input at connect time (before writing
5473 the banner), Exim was not dealing correctly with a non-positive return
5474 from the read() function. If the client had disconnected by this time,
5475 the result was a log entry for a synchronization error with an empty
5476 string after "input=" when read() returned zero. If read() returned -1
5477 (an event I could not check), uninitialized data bytes were printed.
5478 There were reports of junk text (parts of files, etc) appearing after
5481 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5483 PH/22 Added support for macro redefinition, and (re)definition in between
5484 driver and ACL definitions.
5486 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5487 forgetting to use the resulting value; it was using the unexpanded value.
5489 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5490 hadn't been configured. The fix is from Juergen Kreileder, who
5491 understands it better than I do:
5493 "Here's what I see happening with three configured cyrus_sasl
5494 authenticators configured (plain, login, cram-md5):
5496 On startup auth_cyrus_sasl_init() gets called for each of these.
5497 This means three calls to sasl_listmech() without a specified mech_list.
5498 => SASL tests which mechs of all available mechs actually work
5499 => three warnings about OTP not working
5500 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5502 With the patch, sasl_listmech() also gets called three times. But now
5503 SASL's mech_list option is set to the server_mech specified in the the
5504 authenticator. Or in other words, the answer from sasl_listmech()
5505 gets limited to just the mech you're testing for (which is different
5507 => the return list contains just 'plain' or 'login', 'cram-md5' or
5508 nothing depending on the value of ob->server_mech.
5510 I've just tested the patch: Authentication still works fine,
5511 unavailable mechs specified in the exim configuration are still
5512 caught, and the auth.log warnings about OTP are gone."
5514 PH/25 When debugging is enabled, the contents of the command line are added
5515 to the debugging output, even when log_selector=+arguments is not
5518 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5519 answer is "GNU", and only if the return is "GNU/something" is the answer
5522 PH/27 $acl_verify_message is now set immediately after the failure of a
5523 verification in an ACL, and so is available in subsequent modifiers. In
5524 particular, the message can be preserved by coding like this:
5526 warn !verify = sender
5527 set acl_m0 = $acl_verify_message
5529 Previously, $acl_verify_message was set only while expanding "message"
5530 and "log_message" when a very denied access.
5532 PH/28 Modified OS/os.c-Linux with
5534 -#ifndef OS_LOAD_AVERAGE
5535 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5537 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5538 nomenclature these days.)
5540 PH/29 Installed patch from the Sieve maintainer that adds the options
5541 sieve_useraddress and sieve_subaddress to the redirect router.
5543 PH/30 In these circumstances:
5544 . Two addresses routed to the same list of hosts;
5545 . First host does not offer TLS;
5546 . First host accepts first address;
5547 . First host gives temporary error to second address;
5548 . Second host offers TLS and a TLS session is established;
5549 . Second host accepts second address.
5550 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5551 and peerdn, if requested) that were in fact used only for the second
5554 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5555 attention to any local part prefix or suffix that was matched by the
5556 router that accepted the address. It now behaves in the same way as it
5557 does for delivery: the affixes are removed from the local part unless
5558 rcpt_include_affixes is set on the transport.
5560 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5561 timeout during the DATA phase of an incoming message.
5563 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5564 applied a patch sanctioned by the Sieve maintainer.
5566 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5567 the uid or gid is negative. A case of a negative gid caused this to be
5568 noticed. The fix allows for either to be negative.
5570 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5571 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5574 PH/36 The change PH/12 above was broken. Fixed it.
5576 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5577 the grounds that routing the same address twice would always produce the
5578 same answer. This might have been true once, but it is certainly no
5579 longer true now. Routing a child address may depend on the previous
5580 routing that produced that child. Some complicated redirection strategies
5581 went wrong when messages had multiple recipients, and made Exim's
5582 behaviour dependent on the order in which the addresses were given.
5584 I have moved the duplicate checking until after the routing is complete.
5585 Exim scans the addresses that are assigned to local and remote
5586 transports, and removes any duplicates. This means that more work will be
5587 done, as duplicates will always all be routed, but duplicates are
5588 presumably rare, so I don't expect this is of any significance.
5590 For deliveries to pipes, files, and autoreplies, the duplicate checking
5591 still happens during the routing process, since they are not going to be
5594 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5595 It corrects a timeout issue with spamd. This is Ian's comment: "The
5596 background is that sometimes spamd either never reads data from a
5597 connection it has accepted, or it never writes response data. The exiscan
5598 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5599 blindly assumes that writes won't block so it may never time out."
5601 PH/39 Allow G after quota size as well as K and M.
5603 PH/40 The value set for $authenticated_id in an authenticator may not contain
5604 binary zeroes or newlines because the value is written to log lines and
5605 to spool files. There was no check on this. Now the value is run through
5606 the string_printing() function so that such characters are converted to
5607 printable escape sequences.
5609 PH/41 $message_linecount is a new variable that contains the total number of
5610 lines in the message. Compare $body_linecount, which is the count for the
5613 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5614 bounce and delay warning messages, except in certain special cases, which
5617 (a) An SMTP error message from a remote host;
5618 (b) A message specified in a :fail: redirection;
5619 (c) A message specified in a "fail" command in a system filter;
5620 (d) A message specified in a FAIL return from the queryprogram router;
5621 (e) A message specified by the cannot_route_message router option.
5623 In these cases only, Exim does include the error details in bounce and
5624 warning messages. There are also a few cases where bland messages such
5625 as "unrouteable address" or "local delivery error" are given.
5627 PH/43 $value is now also set for the "else" part of a ${run expansion.
5629 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5630 being worked on, but at least Exim now implements the latest version to
5633 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5634 process to complete was treated as a delivery failure, a timeout while
5635 writing the message to the pipe was logged, but erroneously treated as a
5636 successful delivery. Such timeouts include transport filter timeouts. For
5637 consistency with the overall process timeout, these timeouts are now
5638 treated as errors, giving rise to delivery failures by default. However,
5639 there is now a new Boolean option for the pipe transport called
5640 timeout_defer, which, if set TRUE, converts the failures into defers for
5641 both kinds of timeout. A transport filter timeout is now identified in
5644 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5645 systems where "make" and "gmake" are different, calling "gmake" at top
5646 level broke things. I've arranged for the value of $(MAKE) to be passed
5647 from the Makefile to this script so that it can call the same version of
5651 A note about Exim versions 4.44 and 4.50
5652 ----------------------------------------
5654 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5655 changes of various kinds. As a consequence, a big documentation update was
5656 needed. This delayed the release for rather longer than seemed good, especially
5657 in the light of a couple of (minor) security issues. Therefore, the changes
5658 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5659 release. So 4.44 and 4.50 are in effect two different branches that both start
5662 I have left the 4.50 change log unchanged; it contains all the changes since
5663 4.43. The change log for 4.44 is below; many of its items are identical to
5664 those for 4.50. This seems to be the most sensible way to preserve the
5665 historical information.
5671 1. Minor wording change to the doc/README.SIEVE file.
5673 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5674 computation of the current number of files was incorrect.
5676 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5677 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5680 4. Give more explanation in the error message when the command for a transport
5681 filter fails to execute.
5683 5. There are several places where Exim runs a non-Exim command in a
5684 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5685 was being done only for the command run by the queryprogram router. It is
5686 now done for all such subprocesses. The other cases are: ${run, transport
5687 filters, and the commands run by the lmtp and pipe transports.
5689 6. Added CONFIGURE_GROUP build-time option.
5691 7. Some older OS have a limit of 256 on the maximum number of file
5692 descriptors. Exim was using setrlimit() to set 1000 as a large value
5693 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5694 systems. I've change it so that if it can't get 1000, it tries for 256.
5696 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5697 was an oversight, and furthermore, ever since the addition of extra
5698 controls (e.g. 4.43/32), the checks on when to allow different forms of
5699 "control" were broken. There should now be diagnostics for all cases when a
5700 control that does not make sense is encountered.
5702 9. Added the /retain_sender option to "control=submission".
5704 10. $recipients is now available in the predata ACL (oversight).
5706 11. Tidy the search cache before the fork to do a delivery from a message
5707 received from the command line. Otherwise the child will trigger a lookup
5708 failure and thereby defer the delivery if it tries to use (for example) a
5709 cached ldap connection that the parent has called unbind on.
5711 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5712 of $address_data from the recipient verification was clobbered by the
5713 sender verification.
5715 13. The value of address_data from a sender verification is now available in
5716 $sender_address_data in subsequent conditions in the ACL statement.
5718 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5720 15. Added a new option "connect=<time>" to callout options, to set a different
5723 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5724 was its contents. (It was OK if the option was not defined at all.)
5726 17. A "Completed" log line is now written for messages that are removed from
5727 the spool by the -Mrm option.
5729 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5730 information about exactly what failed.
5732 19. Added -dd to debug only the daemon process.
5734 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5735 handles timeouts, both on the server side and network timeouts. Renamed the
5736 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5738 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5739 It is now set to "smtps".
5741 22. $host_address is now set to the target address during the checking of
5742 ignore_target_hosts.
5744 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5745 being passed; this would have caused $sender_host_name to have been used if
5746 matching the list had actually called for a host name (not very likely,
5747 since this list is usually IP addresses). A host name is now passed as
5750 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5751 code that specifies a non-restarting handler (typically sigaction() in
5752 modern systems) in an attempt to fix a rare and obscure crash bug.
5754 25. Narrowed the window for a race in the daemon that could cause it to ignore
5755 SIGCHLD signals. This is not a major problem, because they are used only to
5756 wake it up if nothing else does.
5758 26. A malformed maildirsize file could cause Exim to calculate negative values
5759 for the mailbox size or file count. Odd effects could occur as a result.
5760 The maildirsize information is now recalculated if the size or filecount
5763 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5764 support for a long time. Removed HAVE_SYS_VFS_H.
5766 28. Installed the latest version of exipick from John Jetmore.
5768 29. In an address list, if the pattern was not a regular expression, an empty
5769 subject address (from a bounce message) matched only if the pattern was an
5770 empty string. Non-empty patterns were not even tested. This was the wrong
5771 because it is perfectly reasonable to use an empty address as part of a
5772 database query. An empty address is now tested by patterns that are
5773 lookups. However, all the other forms of pattern expect the subject to
5774 contain a local part and a domain, and therefore, for them, an empty
5775 address still always fails if the pattern is not itself empty.
5777 30. Exim went into a mad DNS loop when attempting to do a callout where the
5778 host was specified on an smtp transport, and looking it up yielded more
5779 than one IP address.
5781 31. Re-factored the code for checking spool and log partition space into a
5782 function that finds that data and another that does the check. The former
5783 is then used to implement four new variables: $spool_space, $log_space,
5784 $spool_inodes, and $log_inodes.
5786 32. The RFC2047 encoding function was originally intended for short strings
5787 such as real names; it was not keeping to the 75-character limit for
5788 encoded words that the RFC imposes. It now respects the limit, and
5789 generates multiple encoded words if necessary. To be on the safe side, I
5790 have increased the buffer size for the ${rfc2047: expansion operator from
5793 33. It is now permitted to omit both strings after an "if" condition; if the
5794 condition is true, the result is "true". As before, when the second string
5795 is omitted, a false condition yields an empty string. This makes it less
5796 cumbersome to write custom ACL and router conditions.
5798 34. Failure to deliver a bounce message always caused it to be frozen, even if
5799 there was an errors_to setting on the router. The errors_to setting is now
5802 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5803 canonical form (fully expanded) before being placed in
5804 $sender_host_address.
5806 36. The table in the code that translates DNS record types into text (T_A to
5807 "A" for instance) was missing entries for NS and CNAME. It is just possible
5808 that this could have caused confusion if both these types were looked up
5809 for the same domain, because the text type is used as part of Exim's
5810 per-process caching. But the chance of anyone hitting this buglet seems
5813 37. The dnsdb lookup has been extended in a number of ways.
5815 (1) There is a new type, "zns", which walks up the domain tree until it
5816 finds some nameserver records. It should be used with care.
5818 (2) There is a new type, "mxh", which is like "mx" except that it returns
5819 just the host names, not the priorities.
5821 (3) It is now possible to give a list of domains (or IP addresses) to be
5822 looked up. The behaviour when one of the lookups defers can be
5823 controlled by a keyword.
5825 (4) It is now possible to specify the separator character for use when
5826 multiple records are returned.
5828 38. The dnslists ACL condition has been extended: it is now possible to supply
5829 a list of IP addresses and/or domains to be looked up in a particular DNS
5832 39. Added log_selector=+queue_time_overall.
5834 40. When running the queue in the test harness, wait just a tad after forking a
5835 delivery process, to get repeatability of debugging output.
5837 41. Include certificate and key file names in error message when GnuTLS fails
5838 to set them up, because the GnuTLS error message doesn't include the name
5839 of the failing file when there is a problem reading it.
5841 42. Allow both -bf and -bF in the same test run.
5843 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5845 44. The "Exiscan patch" is now merged into the mainline Exim source.
5847 45. Sometimes the final signoff response after QUIT could fail to get
5848 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5849 < 0 before doing a fflush(). This bug looks as though it goes back to the
5850 introduction of TLS in release 3.20, but "sometimes" must have been rare
5851 because the tests only now provoked it.
5853 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5854 (this can affect the format of dates).
5856 47. exim_tidydb, when checking for the continued existence of a message for
5857 which it has found a message-specific retry record, was not finding
5858 messages that were in split spool directories. Consequently, it was
5859 deleting retry records that should have stayed in existence.
5861 48. Steve fixed some bugs in eximstats.
5863 49. The SPA authentication driver was not abandoning authentication and moving
5864 on to the next authenticator when an expansion was forced to fail,
5865 contradicting the general specification for all authenticators. Instead it
5866 was generating a temporary error. It now behaves as specified.
5868 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5869 (the order specifies the preference for clients). The order is now AES256,
5870 AES128, 3DES, ARCFOUR128.
5872 51. Small patch to Sieve code - explicitly set From: when generating an
5875 52. Exim crashed if a remote delivery caused a very long error message to be
5876 recorded - for instance if somebody sent an entire SpamAssassin report back
5877 as a large number of 550 error lines. This bug was coincidentally fixed by
5878 increasing the size of one of Exim's internal buffers (big_buffer) that
5879 happened as part of the Exiscan merge. However, to be on the safe side, I
5880 have made the code more robust (and fixed the comments that describe what
5883 53. Now that there can be additional text after "Completed" in log lines (if
5884 the queue_time_overall log selector is set), a one-byte patch to exigrep
5885 was needed to allow it to recognize "Completed" as not the last thing in
5888 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5889 patch that reportedly fixes this has been added. I am not expert enough to
5890 create a test for it. This is what the patch creator wrote:
5892 "I found a little strange behaviour of ldap code when working with
5893 Windows 2003 AD Domain, where users was placed in more than one
5894 Organization Units. When I tried to give exim partial DN, the exit code
5895 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5896 But simultaneously result of request was absolutely normal ldap result,
5897 so I produce this patch..."
5899 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5900 REFERENCE, so I have modified the code to exclude the patch when that macro
5903 55. Some experimental protocols are using DNS PTR records for new purposes. The
5904 keys for these records are domain names, not reversed IP addresses. The
5905 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5906 leaves it alone. Component reversal etc. now happens only for IP addresses.
5909 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5911 57. Double the size of the debug message buffer (to 2048) so that more of very
5912 long debug lines gets shown.
5914 58. The exicyclog utility now does better if the number of log files to keep
5915 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5917 59. Two changes related to the smtp_active_hostname option:
5919 (1) $smtp_active_hostname is now available as a variable.
5920 (2) The default for smtp_banner uses $smtp_active_hostname instead
5921 of $primary_hostname.
5923 60. The host_aton() function is supposed to be passed a string that is known
5924 to be a valid IP address. However, in the case of IPv6 addresses, it was
5925 not checking this. This is a hostage to fortune. Exim now panics and dies
5926 if the condition is not met. A case was found where this could be provoked
5927 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5928 components; fortuitously, this particular loophole had already been fixed
5929 by change 4.50/55 above.
5931 If there are any other similar loopholes, the new check in host_aton()
5932 itself should stop them being exploited. The report I received stated that
5933 data on the command line could provoke the exploit when Exim was running as
5934 exim, but did not say which command line option was involved. All I could
5935 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5936 running as the user.
5939 61. There was a buffer overflow vulnerability in the SPA authentication code
5940 (which came originally from the Samba project). I have added a test to the
5941 spa_base64_to_bits() function which I hope fixes it.
5944 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5945 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5947 63. The daemon start-up calls getloadavg() while still root for those OS that
5948 need the first call to be done as root, but it missed one case: when
5949 deliver_queue_load_max is set with deliver_drop_privilege. This is
5950 necessary for the benefit of the queue runner, because there is no re-exec
5951 when deliver_drop_privilege is set.
5953 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5954 This has been fixed.
5956 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5957 was in use, was not putting the data itself into the right store pool;
5958 consequently, it could be overwritten for a subsequent message in the same
5959 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5962 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5964 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5965 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5966 the function were treating the return as a boolean value, which happened to
5967 work because 0=false and not-0=true, but is not correct code.
5969 68. The host_aton() function was not handling scoped IPv6 addresses (those
5970 with, for example, "%eth0" on the end) correctly.
5972 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5973 negated items (that is, ~something) in unsigned ints. Some compilers
5974 apparently mutter when there is no cast.
5976 70. If an address verification called from an ACL failed, and did not produce a
5977 user-specific message (i.e. there was only a "system" message), nothing was
5978 put in $acl_verify_message. In this situation, it now puts the system
5981 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5982 session; change 4.31/43 added the unwanted input to the log line - except
5983 that it did not do this in the start of session case. It now does.
5985 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5986 This is wrong and can cause the other end to generate a synchronization
5987 error if it is another Exim or anything else that does the synchronization
5988 check. A QUIT command is no longer sent after a timeout.
5990 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5991 during host lookups.
5993 74. The defer_ok option of callout verification was not working if it was used
5994 when verifying addresses in header lines, that is, for this case:
5996 verify = header_sender/callout=defer_ok
5998 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5999 those file descriptors could be used for SMTP connections. If anything
6000 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6001 could be sent to the SMTP client, causing chaos. The daemon now opens
6002 stdin, stdout, and stderr to /dev/null when it puts itself into the
6005 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6006 log by default. The user can override this with suitable Perl magic.
6008 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6009 the log message when discard triggers, was not working for the DATA ACL or
6010 for the non-SMTP ACL.
6012 78. Error message wording change in sieve.c.
6014 79. If smtp_accept_max_per_host was set, the number of connections could be
6015 restricted to fewer than expected, because the daemon was trying to set up
6016 a new connection before checking whether the processes handling previous
6017 connections had finished. The check for completed processes is now done
6018 earlier. On busy systems, this bug wouldn't be noticed because something
6019 else would have woken the daemon, and it would have reaped the completed
6022 80. If a message was submitted locally by a user whose login name contained one
6023 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6024 It caused a spool format error. I have fixed the spool reading code. A
6025 related problem was that the "from" clause in the Received: line became
6026 illegal because of the space(s). It is now covered by ${quote_local_part.
6028 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6031 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6033 83. Patch from Sieve maintainer to fix unterminated string problem in
6034 "vacation" handling.
6036 84. Some minor changes to the Linux configuration files to help with other
6037 OS variants using glibc.
6039 85. One more patch for Sieve to update vacation handling to latest spec.
6042 ----------------------------------------------------
6043 See the note above about the 4.44 and 4.50 releases.
6044 ----------------------------------------------------
6050 1. Change 4.43/35 introduced a bug that caused file counts to be
6051 incorrectly computed when quota_filecount was set in an appendfile
6054 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6055 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6058 3. Give more explanation in the error message when the command for a transport
6059 filter fails to execute.
6061 4. There are several places where Exim runs a non-Exim command in a
6062 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6063 was being done only for the command run by the queryprogram router. It is
6064 now done for all such subprocesses. The other cases are: ${run, transport
6065 filters, and the commands run by the lmtp and pipe transports.
6067 5. Some older OS have a limit of 256 on the maximum number of file
6068 descriptors. Exim was using setrlimit() to set 1000 as a large value
6069 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6070 systems. I've change it so that if it can't get 1000, it tries for 256.
6072 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6073 was an oversight, and furthermore, ever since the addition of extra
6074 controls (e.g. 4.43/32), the checks on when to allow different forms of
6075 "control" were broken. There should now be diagnostics for all cases when a
6076 control that does not make sense is encountered.
6078 7. $recipients is now available in the predata ACL (oversight).
6080 8. Tidy the search cache before the fork to do a delivery from a message
6081 received from the command line. Otherwise the child will trigger a lookup
6082 failure and thereby defer the delivery if it tries to use (for example) a
6083 cached ldap connection that the parent has called unbind on.
6085 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6086 of $address_data from the recipient verification was clobbered by the
6087 sender verification.
6089 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6090 was its contents. (It was OK if the option was not defined at all.)
6092 11. A "Completed" log line is now written for messages that are removed from
6093 the spool by the -Mrm option.
6095 12. $host_address is now set to the target address during the checking of
6096 ignore_target_hosts.
6098 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6099 being passed; this would have caused $sender_host_name to have been used if
6100 matching the list had actually called for a host name (not very likely,
6101 since this list is usually IP addresses). A host name is now passed as
6104 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6105 code that specifies a non-restarting handler (typically sigaction() in
6106 modern systems) in an attempt to fix a rare and obscure crash bug.
6108 15. Narrowed the window for a race in the daemon that could cause it to ignore
6109 SIGCHLD signals. This is not a major problem, because they are used only to
6110 wake it up if nothing else does.
6112 16. A malformed maildirsize file could cause Exim to calculate negative values
6113 for the mailbox size or file count. Odd effects could occur as a result.
6114 The maildirsize information is now recalculated if the size or filecount
6117 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6118 support for a long time. Removed HAVE_SYS_VFS_H.
6120 18. Updated exipick to current release from John Jetmore.
6122 19. Allow an empty sender to be matched against a lookup in an address list.
6123 Previously the only cases considered were a regular expression, or an
6126 20. Exim went into a mad DNS lookup loop when doing a callout where the
6127 host was specified on the transport, if the DNS lookup yielded more than
6130 21. The RFC2047 encoding function was originally intended for short strings
6131 such as real names; it was not keeping to the 75-character limit for
6132 encoded words that the RFC imposes. It now respects the limit, and
6133 generates multiple encoded words if necessary. To be on the safe side, I
6134 have increased the buffer size for the ${rfc2047: expansion operator from
6137 22. Failure to deliver a bounce message always caused it to be frozen, even if
6138 there was an errors_to setting on the router. The errors_to setting is now
6141 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6142 canonical form (fully expanded) before being placed in
6143 $sender_host_address.
6145 24. Updated eximstats to version 1.33
6147 25. Include certificate and key file names in error message when GnuTLS fails
6148 to set them up, because the GnuTLS error message doesn't include the name
6149 of the failing file when there is a problem reading it.
6151 26. Expand error message when OpenSSL has problems setting up cert/key files.
6154 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6155 (this can affect the format of dates).
6157 28. exim_tidydb, when checking for the continued existence of a message for
6158 which it has found a message-specific retry record, was not finding
6159 messages that were in split spool directories. Consequently, it was
6160 deleting retry records that should have stayed in existence.
6162 29. eximstats updated to version 1.35
6163 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6164 1.35 - bugfix such that pie charts by volume are generated correctly
6166 30. The SPA authentication driver was not abandoning authentication and moving
6167 on to the next authenticator when an expansion was forced to fail,
6168 contradicting the general specification for all authenticators. Instead it
6169 was generating a temporary error. It now behaves as specified.
6171 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6172 (the order specifies the preference for clients). The order is now AES256,
6173 AES128, 3DES, ARCFOUR128.
6175 31. Small patch to Sieve code - explicitly set From: when generating an
6178 32. Exim crashed if a remote delivery caused a very long error message to be
6179 recorded - for instance if somebody sent an entire SpamAssassin report back
6180 as a large number of 550 error lines. This bug was coincidentally fixed by
6181 increasing the size of one of Exim's internal buffers (big_buffer) that
6182 happened as part of the Exiscan merge. However, to be on the safe side, I
6183 have made the code more robust (and fixed the comments that describe what
6186 33. Some experimental protocols are using DNS PTR records for new purposes. The
6187 keys for these records are domain names, not reversed IP addresses. The
6188 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6189 leaves it alone. Component reversal etc. now happens only for IP addresses.
6192 34. The host_aton() function is supposed to be passed a string that is known
6193 to be a valid IP address. However, in the case of IPv6 addresses, it was
6194 not checking this. This is a hostage to fortune. Exim now panics and dies
6195 if the condition is not met. A case was found where this could be provoked
6196 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6197 components; fortuitously, this particular loophole had already been fixed
6198 by change 4.50/55 or 4.44/33 above.
6200 If there are any other similar loopholes, the new check in host_aton()
6201 itself should stop them being exploited. The report I received stated that
6202 data on the command line could provoke the exploit when Exim was running as
6203 exim, but did not say which command line option was involved. All I could
6204 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6205 running as the user.
6208 35. There was a buffer overflow vulnerability in the SPA authentication code
6209 (which came originally from the Samba project). I have added a test to the
6210 spa_base64_to_bits() function which I hope fixes it.
6213 36. The daemon start-up calls getloadavg() while still root for those OS that
6214 need the first call to be done as root, but it missed one case: when
6215 deliver_queue_load_max is set with deliver_drop_privilege. This is
6216 necessary for the benefit of the queue runner, because there is no re-exec
6217 when deliver_drop_privilege is set.
6219 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6220 was in use, was not putting the data itself into the right store pool;
6221 consequently, it could be overwritten for a subsequent message in the same
6222 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6225 38. Sometimes the final signoff response after QUIT could fail to get
6226 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6227 < 0 before doing a fflush(). This bug looks as though it goes back to the
6228 introduction of TLS in release 3.20, but "sometimes" must have been rare
6229 because the tests only now provoked it.
6235 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6236 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6237 Now it returns an integer. A number of calls were still expecting a T/F
6238 return. Fortuitously, in all cases, the tests worked in OK situations,
6239 which is the norm. However, things would have gone wrong on any write
6240 failures on the smtp file descriptor. This function is used when sending
6241 messages over SMTP and also when doing verify callouts.
6243 2. When Exim is called to do synchronous delivery of a locally submitted
6244 message (the -odf or -odi options), it no longer closes stderr before doing
6247 3. Implemented the mua_wrapper option.
6249 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6251 5. Implemented the functions header_remove(), header_testname(),
6252 header_add_at_position(), and receive_remove_recipient(), and exported them
6255 6. If an ACL "warn" statement specified the addition of headers, Exim already
6256 inserted X-ACL-Warn: at the start if there was no header name. However, it
6257 was not making this test for the second and subsequent header lines if
6258 there were newlines in the string. This meant that an invalid header could
6259 be inserted if Exim was badly configured.
6261 7. Allow an ACL "warn" statement to add header lines at the start or after all
6262 the Received: headers, as well as at the end.
6264 8. Added the rcpt_4xx retry error code.
6266 9. Added postmaster_mailfrom=xxx to callout verification option.
6268 10. Added mailfrom=xxxx to the callout verification option, for verify=
6271 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6272 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6274 12. Inserted some casts to stop certain compilers warning when using pointer
6275 differences as field lengths or precisions in printf-type calls (mostly
6276 affecting debugging statements).
6278 13. Added optional readline() support for -be (dynamically loaded).
6280 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6281 same clock tick as a message's arrival, so that its received time was the
6282 same as the "first fail" time on the retry record, and that message
6283 remained on the queue past the ultimate address timeout, every queue runner
6284 would try a delivery (because it was past the ultimate address timeout) but
6285 after another failure, the ultimate address timeout, which should have then
6286 bounced the address, did not kick in. This was a "< instead of <=" error;
6287 in most cases the first failure would have been in the next clock tick
6288 after the received time, and all would be well.
6290 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6291 being recognized when the domain list was tested by the match_domain
6292 condition in an expansion string.
6294 16. Added the ${str2b64: operator.
6296 17. Exim was always calling setrlimit() to set a large limit for the number of
6297 processes, without checking whether the existing limit was already
6298 adequate. (It did check for the limit on file descriptors.) Furthermore,
6299 errors from getrlimit() and setrlimit() were being ignored. Now they are
6300 logged to the main and panic logs, but Exim does carry on, to try to do its
6301 job under whatever limits there are.
6303 18. Imported PCRE 5.0.
6305 19. Trivial typo in log message " temporarily refused connection" (the leading
6308 20. If the log selector return_path_on_delivery was set and an address was
6309 redirected to /dev/null, the delivery process crashed because it assumed
6310 that a return path would always be set for a "successful" delivery. In this
6311 case, the whole delivery is bypassed as an optimization, and therefore no
6314 21. Internal re-arrangement: the function for sending a challenge and reading
6315 a response while authentication was assuming a zero-terminated challenge
6316 string. It's now changed to take a pointer and a length, to allow for
6317 binary data in such strings.
6319 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6321 23. Exim was not respecting finduser_retries when seeking the login of the
6322 uid under which it was called; it was always trying 10 times. (The default
6323 setting of finduser_retries is zero.) Also, it was sleeping after the final
6324 failure, which is pointless.
6326 24. Implemented tls_on_connect_ports.
6328 25. Implemented acl_smtp_predata.
6330 26. If the domain in control=submission is set empty, Exim assumes that the
6331 authenticated id is a complete email address when it generates From: or
6332 Sender: header lines.
6334 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6335 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6336 chown and chgrp in /bin and hostname in /usr/bin.
6338 28. Exim was keeping the "process log" file open after each use, just as it
6339 does for the main log. This opens the possibility of it remaining open for
6340 long periods when the USR1 signal hits a daemon. Occasional processlog
6341 errors were reported, that could have been caused by this. Anyway, it seems
6342 much more sensible not to leave this file open at all, so that is what now
6345 29. The long-running daemon process does not normally write to the log once it
6346 has entered its main loop, and it closes the log before doing so. This is
6347 so that log files can straightforwardly be renamed and moved. However,
6348 there are a couple of unusual error situations where the daemon does write
6349 log entries, and I had neglected to close the log afterwards.
6351 30. The text of an SMTP error response that was received during a remote
6352 delivery was being truncated at 512 bytes. This is too short for some of
6353 the long messages that one sometimes sees. I've increased the limit to
6356 31. It is now possible to make retry rules that apply only when a message has a
6357 specific sender, in particular, an empty sender.
6359 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6360 it possible to be selective about when SMTP synchronization is enforced.
6362 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6364 32. Implemented hosts_connection_nolog.
6366 33. Added an ACL for QUIT.
6368 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6371 35. Added mailbox_size and mailbox_filecount to appendfile.
6373 36. Added control = no_multiline_responses to ACLs.
6375 37. There was a bug in the logic of the code that waits for the clock to tick
6376 in the case where the clock went backwards by a substantial amount such
6377 that the microsecond fraction of "now" was more than the microsecond
6378 fraction of "then" (but the whole seconds number was less).
6380 38. Added support for the libradius Radius client library this is found on
6381 FreeBSD (previously only the radiusclient library was supported).
6387 1. When certain lookups returned multiple values in the form name=value, the
6388 quoting of the values was not always being done properly. Specifically:
6389 (a) If the value started with a double quote, but contained no whitespace,
6391 (b) If the value contained whitespace other than a space character (i.e.
6392 tabs or newlines or carriage returns) it was not quoted.
6393 This fix has been applied to the mysql and pgsql lookups by writing a
6394 separate quoting function and calling it from the lookup code. The fix
6395 should probably also be applied to nisplus, ibase and oracle lookups, but
6396 since I cannot test any of those, I have not disturbed their existing code.
6398 2. A hit in the callout cache for a specific address caused a log line with no
6399 reason for rejecting RCPT. Now it says "Previous (cached) callout
6400 verification failure".
6402 3. There was an off-by-one bug in the queryprogram router. An over-long
6403 return line was truncated at 256 instead of 255 characters, thereby
6404 overflowing its buffer with the terminating zero. As well as fixing this, I
6405 have increased the buffer size to 1024 (and made a note to document this).
6407 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6408 when Exim is waiting for an SMTP response from a remote server, Exim
6409 restarts its select() call on the socket, thereby resetting its timeout.
6410 This is not a problem when such interrupts are rare. Somebody set up a cron
6411 job to run exiwhat every 2 minutes, which is less than the normal select()
6412 timeout (5 or 10 minutes). This meant that the select() timeout never
6413 kicked in because it was always reset. I have fixed this by comparing the
6414 time when an interrupt arrives with the time at the start of the first call
6415 to select(). If more time than the timeout has elapsed, the interrupt is
6416 treated as a timeout.
6418 5. Some internal re-factoring in preparation for the addition of Sieve
6419 extensions (by MH). In particular, the "personal" test is moved to a
6420 separate function, and given an option for scanning Cc: and Bcc: (which is
6421 not set for Exim filters).
6423 6. When Exim created an email address using the login of the caller as the
6424 local part (e.g. when creating a From: or Sender: header line), it was not
6425 quoting the local part when it contained special characters such as @.
6427 7. Installed new OpenBSD configuration files.
6429 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6430 try to make them clearer.
6432 9. Callout options, other than the timeout value, were being ignored when
6433 verifying sender addresses in header lines. For example, when using
6435 verify = header_sender/callout=no_cache
6437 the cache was (incorrectly) being used.
6439 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6440 only the Cygwin environment.
6442 11. When return_path_on_delivery was set as a log selector, if different remote
6443 addresses in the same message used different return paths and parallel
6444 remote delivery occurred, the wrong values would sometimes be logged.
6445 (Whenever a remote delivery process finished, the return path value from
6446 the most recently started remote delivery process was logged.)
6448 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6449 lines when AUTH and/or TLS are in use. This is the "received protocol"
6450 field. Exim used to use "asmtp" for authenticated SMTP, without any
6451 indication (in the protocol name) for TLS use. Now it follows the RFC and
6452 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6453 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6454 names appear in log lines as well as in Received: header lines.
6456 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6457 extensions, and comparison tests, and to fix some bugs.
6459 14. Changes to the "personal" filter test:
6461 (1) The test was buggy in that it was just doing the equivalent of
6462 "contains" tests on header lines. For example, if a user's address was
6463 anne@some.where, the "personal" test would incorrectly be true for
6465 To: susanne@some.where
6467 This test is now done by extracting each address from the header in turn,
6468 and checking the entire address. Other tests that are part of "personal"
6469 are now done using regular expressions (for example, to check local parts
6470 of addresses in From: header lines).
6472 (2) The list of non-personal local parts in From: addresses has been
6473 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6474 taken from the Sieve specification recommendations.
6476 (3) If the message contains any header line starting with "List-" it is
6477 treated as non-personal.
6479 (4) The test for "circular" in the Subject: header line has been removed
6480 because it now seems ill-conceived.
6482 15. Minor typos in src/EDITME comments corrected.
6484 16. Installed latest exipick from John Jetmore.
6486 17. If headers_add on a router specified a text string that was too long for
6487 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6488 of string_sprintf() is now avoided.
6490 18. $message_body_size was not set (it was always zero) when running the DATA
6491 ACL and the local_scan() function.
6493 19. For the "mail" command in an Exim filter, no default was being set for
6494 the once_repeat time, causing a random time value to be used if "once" was
6495 specified. (If the value happened to be <= 0, no repeat happened.) The
6496 default is now 0s, meaning "never repeat". The "vacation" command was OK
6497 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6498 (I found it when inspecting the code).
6500 20. There is now an overall timeout for performing a callout verification. It
6501 defaults to 4 times the callout timeout, which applies to individual SMTP
6502 commands during the callout. The overall timeout applies when there is more
6503 than one host that can be tried. The timeout is checked before trying the
6504 next host. This prevents very long delays if there are a large number of
6505 hosts and all are timing out (e.g. when the network connections are timing
6506 out). The value of the overall timeout can be changed by specifying an
6507 additional sub-option for "callout", called "maxwait". For example:
6509 verify = sender/callout=5s,maxwait=20s
6511 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6512 to the end before writing, but this should make it even safer).
6514 22. Exim was forgetting that it had advertised PIPELINING for the second and
6515 subsequent messages on an SMTP connection. It was also not resetting its
6516 memory on STARTTLS and an internal HELO.
6518 23. When Exim logs an SMTP synchronization error within a session, it now
6519 records whether PIPELINING has been advertised or not.
6521 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6522 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6523 rather than long int.
6525 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6527 26. Added the never_mail option to autoreply.
6533 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6534 crash if the getsockname() call failed; this can happen if a connection is
6535 closed very soon after it is established. The problem was simply in the
6536 order in which certain operations were done, causing Exim to try to write
6537 to the SMTP stream before it had set up the file descriptor. The bug has
6538 been fixed by making things happen in the correct order.
6544 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6545 before the connection was closed, thus losing the rejection response.
6547 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6548 some early Solaris releases, but causes trouble in current releases where
6549 socklen_t is defined.
6551 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6554 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6557 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6558 that releasing the top of it at the end releases what was used for sub-
6559 expansions (unless the block got too big). However, discard this block if
6560 the first thing is a variable or header, so that we can use its block when
6561 it is dynamic (useful for very large $message_headers, for example).
6563 6. Lookups now cache *every* query, not just the most recent. A new, separate
6564 store pool is used for this. It can be recovered when all lookup caches are
6565 flushed. Lookups now release memory at the end of their result strings.
6566 This has involved some general refactoring of the lookup sources.
6568 7. Some code has been added to the store_xxx() functions to reduce the amount
6569 of flapping under certain conditions.
6571 8. log_incoming_interface used to affect only the <= reception log lines. Now
6572 it causes the local interface and port to be added to several more SMTP log
6573 lines, for example "SMTP connection from", and rejection lines.
6575 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6577 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6579 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6580 could be overwritten at the end of the current message (or the start of a
6581 new message if it was set in a HELO ACL). The value is now preserved for
6582 the duration of the SMTP connection.
6584 12. If a transport had a headers_rewrite setting, and a matching header line
6585 contained an unqualified address, that address was qualified, even if it
6586 did not match any rewriting rules. The underlying bug was that the values
6587 of the flags that permit the existence of unqualified sender and recipient
6588 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6589 non-local messages, and by -bnq for local messages) were not being
6590 preserved with the message after it was received.
6592 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6593 "next input=" as part of the text comprising the host identity instead of
6594 the correct text. The code was using the same buffer for two different
6595 strings. However, depending on which order the printing function evaluated
6596 its arguments, the bug did not always show up. Under Linux, for example, my
6597 test suite worked just fine.
6599 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6600 doesn't work with some older versions of Perl. It has been changed to "my",
6601 which in any case is probably the better facility to use.
6603 15. A really picky compiler found some instances of statements for creating
6604 error messages that either had too many or two few arguments for the format
6607 16. The size of the buffer for calls to the DNS resolver has been increased
6608 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6609 for addresses that have a lot of PTR records. This alleviates a problem; it
6610 does not fully solve it.
6612 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6613 buffer now truncates the list and logs the incident, which is the same
6614 action as happens when Exim is looking up a host name and its aliases.
6615 Previously in this situation something unpredictable would happen;
6616 sometimes it was "internal error: store_reset failed".
6618 18. If a server dropped the connection unexpectedly when an Exim client was
6619 using GnuTLS and trying to read a response, the client delivery process
6620 crashed while trying to generate an error log message.
6622 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6623 string, for example:
6625 warn message = H1: something\nH2: something
6627 the text was added as a single header line from Exim's point of view
6628 though it ended up OK in the delivered message. However, searching for the
6629 second and subsequent header lines using $h_h2: did not work. This has been
6630 fixed. Similarly, if a system filter added multiple headers in this way,
6631 the routers could not see them.
6633 20. Expanded the error message when iplsearch is called with an invalid key to
6634 suggest using net-iplsearch in a host list.
6636 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6637 ACLs are no longer actually imposed (and a message to that effect is
6640 22. If a "gecos" field in a passwd entry contained escaped characters, in
6641 particular, if it contained a \" sequence, Exim got it wrong when building
6642 a From: or a Sender: header from that name. A second bug also caused
6643 incorrect handling when an unquoted " was present following a character
6644 that needed quoting.
6646 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6647 was not being matched caselessly.
6649 24. Arranged for all hyphens in the exim.8 source to be escaped with
6652 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6653 properly. Recipient callout cache records were still being keyed to include
6654 the sender, even when use_sender was set false. This led to far more
6655 callouts that were necessary. The sender is no longer included in the key
6656 when use_sender is false.
6658 26. Added "control = submission" modifier to ACLs.
6660 27. Added the ${base62d: operator to decode base 62 numbers.
6662 28. dnsdb lookups can now access SRV records.
6664 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6665 the configuration file.
6667 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6668 (-v) mode. This makes the output for a verbose queue run more intelligible.
6670 31. Added a use_postmaster feature to recipient callouts.
6672 32. Added the $body_zerocount variable, containing the number of binary zero
6673 bytes in the message body.
6675 33. The time of last modification of the "new" subdirectory is now used as the
6676 "mailbox time last read" when there is a quota error for a maildir
6679 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6681 35. Added +ignore_unknown as a special item in host lists.
6683 36. Code for decoding IPv6 addresses in host lists is now included, even if
6684 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6685 address was recognized as an IP address, but was then not correctly decoded
6686 into binary, causing unexpected and incorrect effects when compared with
6693 1. Very minor rewording of debugging text in manualroute to say "list of
6694 hosts" instead of "hostlist".
6696 2. If verify=header_syntax was set, and a header line with an unqualified
6697 address (no domain) and a large number of spaces between the end of the
6698 name and the colon was received, the reception process suffered a buffer
6699 overflow, and (when I tested it) crashed. This was caused by some obsolete
6700 code that should have been removed. The fix is to remove it!
6702 3. When running in the test harness, delay a bit after writing a bounce
6703 message to get a bit more predictability in the log output.
6705 4. Added a call to search_tidyup() just before forking a reception process. In
6706 theory, someone could use a lookup in the expansion of smtp_accept_max_
6707 per_host which, without the tidyup, could leave open a database connection.
6709 5. Added the variables $recipient_data and $sender_data which get set from a
6710 lookup success in an ACL "recipients" or "senders" condition, or a router
6711 "senders" option, similar to $domain_data and $local_part_data.
6713 6. Moved the writing of debug_print from before to after the "senders" test
6716 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6717 problems for message scanning, either using a data ACL, or using
6718 local_scan() because the Received: header was not generated till after they
6719 were called (in order to set the time as the time of reception completion).
6720 I have revised the way this works. The header is now generated after the
6721 body is received, but before the ACL or local_scan() are called. After they
6722 are run, the timestamp in the header is updated.
6728 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6729 before starting a queue runner without re-exec. This happened only when
6730 deliver_drop_privilege was set or when the Exim user was set to root. The
6731 effect of the bug was that timeouts during subsequent deliveries caused
6732 crashes instead of being properly handled. The handler is now left at its
6733 default (and expected) setting.
6735 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6736 message, again when deliver_drop_privilege is set or Exim is run as root.
6737 The bug described in (1) was not present in this case, but the tidying up
6738 of the other signals was missing. I have made the two cases consistent.
6740 3. The ignore_target_hosts setting on a manualroute router was being ignored
6741 for hosts that were looked up using the /MX notation.
6743 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6746 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6747 operated on the sender address. After changing the $sender_address to <>
6748 for the sender address verify, Exim was re-instated it as the original
6749 (before rewriting) address, but remembering that it had rewritten it, so it
6750 wasn't rewriting it again. This bug also had the effect of breaking the
6751 sender address verification caching when the sender address was rewritten.
6753 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6754 This has been changed so that if the ip literal address matches
6755 ignore_target_hosts, the router declines.
6757 7. Added expansion conditions match_domain, match_address, and match_local_
6758 part (NOT match_host).
6760 8. The placeholder for the Received: header didn't have a length field set.
6762 9. Added code to Exim itself and to exim_lock to test for a specific race
6763 condition that could lead to file corruption when using MBX delivery. The
6764 issue is with the lockfile that is created in /tmp. If this file is removed
6765 after a process has opened it but before that process has acquired a lock,
6766 there is the potential for a second process to recreate the file and also
6767 acquire a lock. This could lead to two Exim processes writing to the file
6768 at the same time. The added code performs the same test as UW imapd; it
6769 checks after acquiring the lock that its file descriptor still refers to
6770 the same named file.
6772 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6773 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6776 11. Added the smtp_active_hostname option. If used, this will typically be made
6777 to depend on the incoming interface address. Because $interface_address is
6778 not set up until the daemon has forked a reception process, error responses
6779 that can happen earlier (such as "too many connections") no longer contain
6782 12. If an expansion in a condition on a "warn" statement fails because a lookup
6783 defers, the "warn" statement is abandoned, and the next ACL statement is
6784 processed. Previously this caused the whole ACL to be aborted.
6786 13. Added the iplsearch lookup type.
6788 14. Added ident_timeout as a log selector.
6790 15. Added tls_certificate_verified as a log selector.
6792 16. Added a global option tls_require_ciphers (compare the smtp transport
6793 option of the same name). This controls incoming TLS connections.
6795 17. I finally figured out how to make tls_require_ciphers do a similar thing
6796 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6797 before starting the TLS session.
6799 18. Tabs are now shown as \t in -bP output.
6801 19. If the log selector return_path_on_delivery was set, Exim crashed when
6802 bouncing a message because it had too many Received: header lines.
6804 20. If two routers both had headers_remove settings, and the first one included
6805 a superfluous trailing colon, the final name in the first list and the
6806 first name in the second list were incorrectly joined into one item (with a
6807 colon in the middle).
6813 1. Added -C and -D options to the exinext utility, mainly to make it easier
6814 to include in the automated testing, but these could be helpful when
6815 multiple configurations are in use.
6817 2. The exinext utility was not formatting the output nicely when there was
6818 an alternate port involved in the retry record key, nor when there was a
6819 message id as well (for retries that were specific to a specific message
6820 and a specific host). It was also confused by IPv6 addresses, because of
6821 the additional colons they contain. I have fixed the IPv4 problem, and
6822 patched it up to do a reasonable job for IPv6.
6824 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6825 delivery, the log line now contains "pipelined" if PIPELINING was used.
6827 4. An SMTP transport process used to panic and die if the bind() call to set
6828 an explicit outgoing interface failed. This has been changed; it is now
6829 treated in the same way as a connect() failure.
6831 5. A reference to $sender_host_name in the part of a conditional expansion
6832 that was being skipped was still causing a DNS lookup. This no longer
6835 6. The def: expansion condition was not recognizing references to header lines
6836 that used bh_ and bheader_.
6838 7. Added the _cache feature to named lists.
6840 8. The code for checking quota_filecount in the appendfile transport was
6841 allowing one more file than it should have been.
6843 9. For compatibility with Sendmail, the command line option
6851 and sets the incoming protocol and host name (for trusted callers). The
6852 host name and its colon can be omitted when only the protocol is to be set.
6853 Note the Exim already has two private options, -pd and -ps, that refer to
6854 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6855 "s", but I don't think that's a major issue.
6857 10. A number of refactoring changes to the code, none of which should affect
6860 (a) The number of logging options was getting close to filling up the
6861 32-bit word that was used as a bit map. I have split them into two classes:
6862 those that are passed in the argument to log_write(), and those that are
6863 only ever tested independently outside of that function. These are now in
6864 separate 32-bit words, so there is plenty of room for expansion again.
6865 There is no change in the user interface or the logging behaviour.
6867 (b) When building, for example, log lines, the code previously used a
6868 macro that called string_cat() twice, in order to add two strings. This is
6869 not really sufficiently general. Furthermore, there was one instance where
6870 it was actually wrong because one of the argument was used twice, and in
6871 one call a function was used. (As it happened, calling the function twice
6872 did not affect the overall behaviour.) The macro has been replaced by a
6873 function that can join an arbitrary number of extra strings onto a growing
6876 (c) The code for expansion conditions now uses a table and a binary chop
6877 instead of a serial search (which was left over from when there were very
6878 few conditions). Also, it now recognizes conditions like "pam" even when
6879 the relevant support is not compiled in: a suitably worded error message is
6880 given if an attempt is made to use such a condition.
6882 11. Added ${time_interval:xxxxx}.
6884 12. A bug was causing one of the ddress fields not to be passed back correctly
6885 from remote delivery subprocesses. The field in question was not being
6886 subsequently used, so this caused to problems in practice.
6888 13. Added new log selectors queue_time and deliver_time.
6890 14. Might have fixed a bug in maildirsizefile handling that threw up
6891 "unexpected character" debug warnings, and recalculated the data
6892 unnecessarily. In any case, I expanded the warning message to give more
6895 15. Added the message "Restricted characters in address" to the statements in
6896 the default ACL that block characters like @ and % in local parts.
6898 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6899 Three changes have been made:
6901 (a) There was a serious bug; a negative response to MAIL caused the whole
6902 recipient domain to be cached as invalid, thereby blocking all messages
6903 to all local parts at the same domain, from all senders. This bug has
6904 been fixed. The domain is no longer cached after a negative response to
6905 MAIL if the sender used is not empty.
6907 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6910 (c) A new callout option, "use_sender" has been added for people who want
6911 the modified behaviour.
6917 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6920 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6921 indeed breaks things for older releases.
6923 3. Added additional logging to the case where there is a problem reading data
6924 from a filter that is running in a subprocess using a pipe, in order to
6925 try to track down a specific problem.
6927 4. Testing facility fudge: when running in the test harness and attempting
6928 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6929 getting "No route to host". Convert this to a timeout.
6931 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6934 6. Some OS don't have socklen_t but use size_t instead. This affects the
6935 fifth argument of getsockopt() amongst other things. This is now
6936 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6937 can be set for individual OS. I have set it for SunOS5, OSF1, and
6938 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6939 some earlier ones do not.
6941 7. Change 4.30/15 was not doing the test caselessly.
6943 8. The standard form for an IPv6 address literal was being rejected by address
6944 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6945 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6946 this, as well as the form without the "IPv6" on the front (but only when
6947 address literals are enabled, of course).
6949 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6951 10. Exim crashed if a message with an empty sender address specified by -f
6952 encountered a router with an errors_to setting. This could be provoked only
6953 by a command such as
6957 where an empty string was supplied; "<>" did not hit this bug.
6959 11. Installed PCRE release 4.5.
6961 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6962 remained set. It is now erased.
6964 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6965 times from message ids (which are base 36 rather than the normal 62).
6967 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6968 were being counted as actual protocol errors, and logged if the log
6969 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6970 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6971 and DATA following a set of rejected RCPTs do not count as protocol errors.
6972 In other words, Exim assumes they were pipelined, though this may not
6973 actually be the case. Of course, in all cases the client gets an
6974 appropriate error code.
6976 15. If a lookup fails in an ACL condition, a message about the failure may
6977 be available; it is used if testing the ACL cannot continue, because most
6978 such messages specify what the cause of the deferral is. However, some
6979 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6980 that caused an old message to be retained and used if a later statement
6981 caused a defer, replacing the real cause of the deferral.
6983 16. If an IP address had so many PTR records that the DNS lookup buffer
6984 was not large enough to hold them, Exim could crash while trying to process
6985 the truncated data. It now detects and logs this case.
6987 17. Further to 4.21/58, another change has been made: if (and only if) the
6988 first line of a message (the first header line) ends with CRLF, a bare LF
6989 in a subsequent header line has a space inserted after it, so as not to
6990 terminate the header.
6992 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6993 unnecessarily, used atoi() instead of strtol(), and didn't check the
6994 termination when getting file sizes from file names by regex.
6996 19. Completely re-implemented the support for maildirsize files, in the light
6997 of a number of problems with the previous contributed implementation
6998 (4.30/29). In particular:
7000 . If the quota is zero, the maildirsize file is maintained, but no quota is
7003 . If the maildir directory does not exist, it is created before any attempt
7004 to write a maildirsize file.
7006 . The quota value in the file is just a cache; if the quota is changed in
7007 the transport, the new value overrides.
7009 . A regular expression is available for excluding directories from the
7012 20. The autoreply transport checks the characters in options that define the
7013 message's headers; it allows continued headers, but it was checking with
7014 isspace() after an embedded newline instead of explicitly looking for a
7017 21. If all the "regular" hosts to which an address was routed had passed their
7018 expiry times, and had not reached their retry times, the address was
7019 bounced, even if fallback hosts were defined. Now Exim should go on to try
7022 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7023 equivalent code in the SMTP transport. Some hosts send humungous responses
7024 to HELO/EHLO, more than 1024 it seems.
7026 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7027 gives compiler warnings in some environments. I've now done it "properly",
7030 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7031 (because of problems with the built-in one) was declared to return uschar *
7032 instead of char *, causing compiler failure.
7034 25. Fixed a file descriptor leak when processing alias/forward files.
7036 26. Fixed a minor format string issue in dbfn.c.
7038 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7040 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7041 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7044 29. When a sender address is verified, it is cached, to save repeating the test
7045 when there is more than one recipient in a message. However, when the
7046 verification involves a callout, it is possible for different callout
7047 options to be set for different recipients. It is too complicated to keep
7048 track of this in the cache, so now Exim always runs a verification when a
7049 callout is required, relying on the callout cache for the optimization.
7050 The overhead is duplication of the address routing, but this should not be
7053 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7054 to be verified with callout=postmaster, and the main callout worked but the
7055 postmaster check failed, the verification correctly failed. However, if a
7056 subsequent RCPT command asked for sender verification *without* the
7057 postmaster check, incorrect caching caused this verification also to fail,
7060 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7061 it was not caching the DNS options (qualify_single, search_parents) that
7062 were used when the lookup failed. A subsequent lookup with different
7063 options therefore always gave the same answer, though there were cases
7064 where it should not have. (Example: a "domains = !$mx_any" option on a
7065 dnslookup router: the "domains" option is always processed without any
7066 widening, but the router might have qualify_single set.) Now Exim uses the
7067 cached value only when the same options are set.
7069 32. Added John Jetmore's "exipick" utility to the distribution.
7071 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7072 than a timeout (e.g. a certificate is required, and is not provided), an
7073 Exim server now closes the connection immediately. Previously it waited for
7074 the client to close - but if the client is SSL, it seems that they each
7075 wait for each other, leading to a delay before one of them times out.
7077 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7078 maintained 0.8.x compatibility because I don't think many are using it, and
7079 it is clearly obsolete.
7081 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7084 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7085 client certificate was expired. A simple patch fixes this, though I don't
7086 understand the full logic of why the verify callback is called multiple
7089 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7090 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7091 which causes problems with some clients (such as the Certicom SSL Plus
7092 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7093 disables the coutermeasure allowing Eudora to connect."
7095 38. Exim was not checking that a write() to a log file succeeded. This could
7096 lead to Bad Things if a log got too big, in particular if it hit a file
7097 size limit. Exim now panics and dies if it cannot write to a log file, just
7098 as it does if it cannot open a log file.
7100 39. Modified OS/Makefile-Linux so that it now contains
7102 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7104 The two -D definitions ensure that Exim is compiled with large file
7105 support, which makes it possible to handle log files that are bigger than
7108 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7109 instance) a domain was checked against a named list that involved a lookup,
7110 causing $domain_data to be set, then another domain was checked against the
7111 same list, then the first domain was re-checked, the value of $domain_data
7112 after the final check could be wrong. In particular, if the second check
7113 failed, it could be set empty. This bug probably also applied to
7116 41. The strip_trailing_dot option was not being applied to the address given
7117 with the -f command-line option.
7119 42. The code for reading a message's header from the spool was incrementing
7120 $received_count, but never initializing it. This meant that the value was
7121 incorrect (doubled) while delivering a message in the same process in which
7122 it was received. In the most common configuration of Exim, this never
7123 happens - a fresh exec is done - but it can happen when
7124 deliver_drop_privilege is set.
7126 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7127 it now includes up to 150 characters of the unexpected data in the log
7130 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7131 and building data strings. The size of both of these buffers was 10 000
7132 bytes - far larger than anybody would *ever* want, thought I. Needless to
7133 say, somebody hit the limit. I have increased the maximum line length to
7134 20 000 and the maximum data length of concatenated lines to 100 000. I have
7135 also fixed two bugs, because there was no checking on these buffers. Tsk,
7136 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7137 buffer is too small.
7139 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7140 lsearch lookups. Now it does.
7142 46. When parsing a route_list item in a manualroute router, a fixed-length
7143 buffer was used for the list of hosts. I made this 1024 bytes long,
7144 thinking that nobody would ever have a list of hosts that long. Wrong.
7145 Somebody had a whole pile of complicated expansion conditions, and the
7146 string was silently truncated, leading to an expansion error. It turns out
7147 that it is easier to change to an unlimited length (owing to other changes
7148 that have happened since this code was originally written) than to build
7149 structure for giving a limitation error. The length of the item that
7150 expands into the list of hosts is now unlimited.
7152 47. The lsearch lookup could not handle data where the length of text line was
7153 more than 4095 characters. Such lines were truncated, leading to shortened
7154 data being returned. It should now handle lines of any length.
7156 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7157 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7160 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7161 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7162 no longer happens in comment lines. A list of replacements is now placed
7163 at the head of all of the source files, except those whose only change is
7164 to replace PERL_COMMAND in the very first #! line.
7166 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7167 messages on the queue, with a bottom-up merge sort, using code contributed
7168 by Michael Haardt. This should make operations like -bp somewhat faster on
7169 large queues. It won't affect queue runners, except when queue_run_in_order
7172 51. Installed eximstats 1.31 in the distribution.
7174 52. Added support for SRV lookups to the dnslookup router.
7176 53. If an ACL referred to $message_body or $message_body_end, the value was not
7177 reset for any messages that followed in the same SMTP session.
7179 54. The store-handling optimization for building very long strings was not
7180 differentiating between the different store pools. I don't think this
7181 actually made any difference in practice, but I've tidied it.
7183 55. While running the routers to verify a sender address, $sender_address
7184 was still set to the sender address. This is wrong, because when routing to
7185 send a bounce to the sender, it would be empty. Therefore, I have changed
7186 it so that, while verifying a sender address, $sender_address is set to <>.
7187 (There is no change to what happens when verifying a recipient address.)
7189 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7190 target A or AAAA records (if not already returned) without resetting the
7191 qualify_single or search_parents options of the DNS resolver. These are
7192 inappropriate in this case because the targets of MX and SRV records must
7193 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7194 target that, when qualified, matched something in the local domain. These
7195 two options are now turned off when doing these lookups.
7197 57. It seems that at least some releases of Reiserfs (which does not have the
7198 concept of a fixed number of inodes) returns zero and not -1 for the
7199 number of available inodes. This interacted badly with check_spool_inodes,
7200 which assumed that -1 was the "no such thing" setting. What I have done is
7201 to check that the total number of inodes is greater than zero before doing
7202 the test of how many are available.
7204 58. When a "warn" ACL statement has a log_message modifier, the message is
7205 remembered, and not repeated. This is to avoid a lot of repetition when a
7206 message has many recipients that cause the same warning to be written.
7207 However, Exim was preserving the list of already written lines for an
7208 entire SMTP session, which doesn't seem right. The memory is now reset if a
7209 new message is started.
7211 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7212 debugging output unless log_rewrite was also set.
7214 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7215 of (int)(handle) when we know that handle contains (void *)(-1).
7217 61. The Exim daemon panic-logs an error return when it closes the incoming
7218 connection. However "connection reset by peer" seems to be common, and
7219 isn't really an error worthy of noting specially, so that particular error
7222 62. When Exim is trying to find all the local interfaces, it used to panic and
7223 die if the ioctl to get the interface flags failed. However, it seems that
7224 on at least one OS (Solaris 9) it is possible to have an interface that is
7225 included in the list of interfaces, but for which you get a failure error
7226 for this call. This happens when the interface is not "plumbed" into a
7227 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7228 failure of the "get flags" call assumes that the interface is down.
7230 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7231 makes life easier for people who are doing arithmetic on fields extracted
7232 from dates, where you often get leading zeros that should not be
7233 interpreted as octal.
7235 64. Added qualify_domain to the redirect router, to override the global
7238 65. If a pathologically long header line contained very many addresses (the
7239 report of this problem mentioned 10 000) and each of them was rewritten,
7240 Exim could use up a very large amount of memory. (It kept on making new
7241 copies of the header line as it rewrote, and never released the old ones.)
7242 At the expense of a bit more processing, the header rewriting function has
7243 been changed so that it no longer eats memory in this way.
7245 66. The generation of the Received: header has been moved from the time that a
7246 message starts to be received, to the time that it finishes. The timestamp
7247 in the Received: header should now be very close to that of the <= log
7248 line. There are two side-effects of this change:
7250 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7251 logged header lines no longer include the local Received: line, because
7252 it has not yet been created. The same applies to a copy of the message
7253 that is returned to a non-SMTP sender when a message is rejected.
7255 (b) When a filter file is tested using -bf, no additional Received: header
7256 is added to the test message. After some thought, I decided that this
7259 This change does not affect the value of $received_for. It is still set
7260 after address rewriting, but before local_scan() is called.
7262 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7264 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7265 gave an unhelpful panic error message, and a defer error. I have managed to
7266 change this behaviour so that it now rejects any supplied certificate,
7267 which seems right, as the list of acceptable certificates is empty.
7269 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7270 gave an unhelpful defer error. I have not managed to make this reject any
7271 supplied certificates, but the error message it gives is "no certificate
7272 supplied", which is not helpful.
7274 70. exigrep's output now also includes lines that are not associated with any
7275 message, but which match the given pattern. Implemented by a patch from
7276 Martin Sluka, which also tidied up the Perl a bit.
7278 71. Recipient callout verification, like sender verification, was using <> in
7279 the MAIL FROM command. This isn't really the right thing, since the actual
7280 sender may affect whether the remote host accepts the recipient or not. I
7281 have changed it to use the actual sender in the callout; this means that
7282 the cache record is now keyed on a recipient/sender pair, not just the
7283 recipient address. There doesn't seem to be a real danger of callout loops,
7284 since a callout by the remote host to check the sender would use <>.
7285 [SEE ABOVE: changed after hitting problems.]
7287 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7288 temporary errors. However, in the case of such a code being given after
7289 the end of a data transmission (i.e. after ".") Exim was failing to write
7290 a retry record for the message. (Yes, there was some broken host that was
7291 actually sending 8xx at this point.)
7293 73. An unknown lookup type in a host list could cause Exim to panic-die when
7294 the list was checked. (An example that provoked this was putting <; in the
7295 middle of a list instead of at the start.) If this happened during a DATA
7296 ACL check, a -D file could be left lying around. This kind of configuration
7297 error no longer causes Exim to die; instead it causes a defer error. The
7298 incident is still logged to the main and panic logs.
7300 74. Buglet left over from Exim 3 conversion. The message "too many messages
7301 in one connection" was written to the rejectlog but not the mainlog, except
7302 when address rewriting (yes!) was being logged.
7304 75. Added write_rejectlog option.
7306 76. When a system filter was run not as root (that is, when system_filter_user
7307 was set), the values of the $n variables were not being returned to the
7308 main process; thus, they were not subsequently available in the $sn
7311 77. Added +return_path_on_delivery log selector.
7313 78. A connection timeout was being treated differently from recipients deferred
7314 when testing hosts_max_try with a message that was older than the host's
7315 retry timeout. (The host should not be counted, thus allowing all hosts to
7316 be tried at least once before bouncing.) This may have been the cause of an
7317 occasionally reported bug whereby a message would remain on the queue
7318 longer than the retry timeout, but would be bounced if a delivery was
7319 forced. I say "may" because I never totally pinned down the problem;
7320 setting up timeout/retry tests is difficult. See also the next item.
7322 79. The ultimate address timeout was not being applied to errors that involved
7323 a combination of host plus message (for example, a timeout on a MAIL
7324 command). When an address resolved to a number of possible hosts, and they
7325 were not all tried for each delivery (e.g. because of hosts_max_try), a
7326 message could remain on the queue longer than the retry timeout.
7328 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7331 81. Fixed an obscure SMTP outgoing bug which required at least the following
7332 conditions: (a) there was another message waiting for the same server;
7333 (b) the server returned 5xx to all RCPT commands in the first message so
7334 that the message was not completed; (c) the server dropped the connection
7335 or gave a negative response to the RSET that Exim sends to abort the
7336 transaction. The observed case was a dropped connection after DATA that had
7337 been sent in pipelining mode. That is, the server had advertised PIPELINING
7338 but was not implementing it correctly. The effect of the bug was incorrect
7339 behaviour, such as trying another host, and this could lead to a crash.
7345 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7346 and daemon.c were passed as pointers to ints; they should have been
7347 pointers to socklen_t variables (which are typically unsigned ints).
7349 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7352 3. Fixed a really odd bug that affected only the testing scheme; patching a
7353 certain fixed string in the binary changed the value of another string that
7354 happened to be identical to the end of the original first string.
7356 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7357 name", it returns that address as the IP address. On some operating
7358 systems (e.g. Solaris), it also passes back the IP address string as the
7359 "host name". However, on others (e.g. Linux), it passes back an empty
7360 string. Exim wasn't checking for this, and was changing the host name to an
7361 empty string, assuming it had been canonicalized.
7363 5. Although rare, it is permitted to have more than one PTR record for a given
7364 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7365 all the names associated with an address, because they do in Solaris.
7366 However, it seems that they do not in Linux for data that comes from the
7367 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7368 I found this out when I moved to a new Linux workstation and tried to run
7369 the Exim test suite.
7371 To get round this problem I have changed the code so that it now does its
7372 own call to the DNS to look up PTR records when searching for a host name.
7373 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7374 addresses that are only in /etc/hosts are still found.
7376 This behaviour is, however, controlled by an option called host_lookup_
7377 order, which defaults to "bydns:byaddr". If people want to use the other
7378 order, or indeed, just use one or the other means of lookup, they can
7379 specify it in this variable.
7381 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7382 some operating systems, this comes back from gethostbyaddr() as an empty
7383 string, and this is what Exim used to test for. However, it seems that in
7384 other systems, "." is yielded. Exim now tests for this case too.
7386 7. The values of check_spool_space and check_log_space are now held internally
7387 as a number of kilobytes instead of an absolute number of bytes. If a
7388 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7389 kilobyte. This means that much larger values can be stored.
7391 8. Exim monitor: an attempt to get the action menu when not actually pointing
7392 at a message produces an empty menu entitled "No message selected". This
7393 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7394 no entries in it ("Shell widget menu has zero width and/or height"). So I
7395 have added a single, blank menu entry in this case.
7397 9. Added ${quote_local_part.
7399 10. MIME decoding is now applied to the contents of Subject: header lines when
7402 11. Now that a reference to $sender_host_address automatically causes a reverse
7403 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7404 host lookup before query-style lookups in lists that might use this
7405 variable. This has therefore been abolished, and the "net-" prefix is no
7406 longer necessary for query-style lookups.
7408 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7409 have been a typo for LFLAGS, so it has been changed.
7411 13. The install script calls Exim with "-C /dev/null" in order to find the
7412 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7413 to be output. However, since Exim outputs its version number before the
7414 error, it didn't break the script. It just looked ugly. I fixed this by
7415 always allowing "-C /dev/null" if the caller is root.
7417 14. Ignore overlarge ACL variable number when reading spool file - insurance
7418 against a later release with more variables having written the file.
7420 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7421 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7422 this, as well as the form without the "IPv6" on the front.
7424 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7425 OS/Makefile-Darwin file.
7427 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7428 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7430 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7431 variable. However, some OS use ints for inodes. I've added cast to long int
7432 to get rid of the compiler warning.
7434 19. I had forgotten to lock out "/../" in configuration file names when
7435 ALT_CONFIG_PREFIX was set.
7437 20. Routers used for verification do not need to specify transports. However,
7438 if such a router generated a host list, and callout was configured, Exim
7439 crashed, because it could not find a port number from the (non-existent)
7440 transport. It now assumes port 25 in this circumstance.
7442 21. Added the -t option to exigrep.
7444 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7445 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7446 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7447 them is set, LOOKUP_LSEARCH is forced.
7449 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7451 24. Added sender and host information to the "rejected by local_scan()" log
7452 line; previously there was no indication of these.
7454 25. Added .include_if_exists.
7456 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7457 Linux. It turns out that not all file systems support this. Apparently some
7458 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7459 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7460 not supported on the file descriptor, is now ignored when Exim is trying to
7461 sync a directory. This applies only to Linux.
7463 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7465 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7466 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7469 29. Added support for maildirsize files from supplied patch (modified a bit).
7471 30. The use of :fail: followed by an empty string could lead Exim to respond to
7472 sender verification failures with (e.g.):
7474 550 Verification failed for <xxx>
7475 550 Sender verify failed
7477 where the first response line was missing the '-' that indicates it is not
7478 the final line of the response.
7480 31. The loop for finding the name of the user that called Exim had a hardwired
7481 limit of 10; it now uses the value of finduser_retries, which is used for
7482 all other user lookups.
7484 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7487 33. Exim was neglecting to zero errno before one call of strtol() when
7488 expanding a string and expecting an integer value. On some systems this
7489 resulted in spurious "integer overflow" errors. Also, it was casting the
7490 result into an int without checking.
7492 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7493 did not work. The code looks as if it has *never* worked, though it appears
7494 to have been documented since at least release 1.62. I have made it work.
7496 35. The "timeout_DNS" error in retry rules, also documented since at least
7497 1.62, also never worked. As it isn't clear exactly what this means, and
7498 clearly it isn't a major issue, I have abolished the feature by treating it
7499 as "timeout", and writing a warning to the main and panic logs.
7501 36. The display of retry rules for -brt wasn't always showing the error code
7504 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7505 timeout_connect_A, timeout_connect_MX.
7507 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7508 to the empty sender.
7510 39. The daemon was not analysing the content of -oX till after it had closed
7511 stderr and disconnected from the controlling terminal. This meant that any
7512 syntax errors were only noted on the panic log, and the return code from
7513 the command was 0. By re-arranging the code a little, I've made the
7514 decoding happen first, so such errors now appear on stderr, and the return
7515 code is 1. However, the actual setting up of the sockets still happens in
7516 the disconnected process, so errors there are still only recorded on the
7519 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7520 connections (as happens on some IP stacks) was logged at start up time as
7521 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7522 it from "IPv6 and IPv4", which means that two separate sockets are being
7525 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7526 says whether AF_INET or AF_INET6 was passed as an argument.
7528 42. Exiwhat output was messed up when time zones were included in log
7531 43. Exiwhat now gives more information about the daemon's listening ports,
7532 and whether -tls-on-connect was used.
7534 44. The "port" option of the smtp transport is now expanded.
7536 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7537 silently ignored. Now an error message is written to the main and panic
7540 46. There's a new ACL modifier called "logwrite" which writes to a log file
7541 as soon as it is encountered.
7543 47. Added $local_user_uid and $local_user_gid at routing time.
7545 48. Exim crashed when trying to verify a sender address that was being
7548 49. Exim was recognizing only a space character after ".include". It now also
7549 recognizes a tab character.
7551 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7552 extracting the relevant information from the specification. The man page no
7553 longer contains scrambled data for the -d option, and I've added a section
7554 at the front about calling Exim under different names.
7556 51. Added "extra_headers" argument to the "mail" command in filter files.
7558 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7561 53. Installed eximstats 1.29.
7563 54. Added transport_filter_timeout as a generic transport option.
7565 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7566 Cc: header lines. This was required by RFC 822, but it not required by RFC
7569 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7570 incoming messages that did not have them. Now it does so only if the
7571 message originates locally, that is, if there is no associated remote host
7572 address. When Resent- header lines are present, this applies to the Resent-
7573 lines rather than the non-Resent- lines.
7575 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7576 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7578 58. Messages for configuration errors now include the name of the main
7579 configuration file - useful now that there may be more than one file in a
7580 list (.included file names were always shown).
7582 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7583 for those rare installations that do not start the daemon as root or run it
7584 setuid root. I've cut out the call to initgroups() if the daemon is not
7587 60. The Exim user and group can now be bound into the binary as text strings
7588 that are looked up at the start of Exim's processing.
7590 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7592 62. Added $mailstore_basename variable.
7594 63. Installed patch to sieve.c from Michael Haardt.
7596 64. When Exim failed to open the panic log after failing to open the main log,
7597 the original message it was trying to log was written to stderr and debug
7598 output, but if they were not available (the usual case in production), it
7599 was lost. Now it is written to syslog before the two lines that record the
7600 failures to open the logs.
7602 65. Users' Exim filters run in subprocesses under the user's uid. It is
7603 possible for a "deliver" command or an alias in a "personal" command to
7604 provoke an address rewrite. If logging of address rewriting is configured,
7605 this fails because the process is not running as root or exim. There may be
7606 a better way of dealing with this, but for the moment (because 4.30 needs
7607 to be released), I have disabled address rewrite logging when running a
7608 filter in a non-root, non-exim process.
7614 1. The buildconfig auxiliary program wasn't quoting the value set for
7615 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7616 not defined. This bug was masked in 4.22 by the effect that was fixed in
7619 2. Some messages that were rejected after a message id was allocated were
7620 shown as "incomplete" by exigrep. It no longer does this for messages that
7621 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7623 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7624 have allow_domain_literals set, the ID did not get logged in the <= line.
7625 Domain literals are now always recognized in Message-ID: header lines.
7627 4. The first argument for a ${extract expansion item is the key name or field
7628 number. Leading and trailing spaces in this item were not being ignored,
7629 causing some misleading effects.
7631 5. When deliver_drop_privilege was set, single queue runner processes started
7632 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7633 same command in the process it spins off) were not dropping privilege.
7635 6. When the daemon running as "exim" started a queue runner, it always
7636 re-executed Exim in the spun-off process. This is a waste of effort when
7637 deliver_drop_privilege is set. The new process now just calls the
7638 queue-runner function directly.
7644 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7647 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7648 ignored. Though the use of -oP was forcing the writing of a pid file, it
7649 was always written to the default place.
7651 3. If the message "no IP address found for host xxxx" is generated during
7652 incoming verification, it is now followed by identification of the incoming
7653 connection (so you can more easily find what provoked it).
7655 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7657 5. Added some features to "harden" Exim a bit more against certain attacks:
7659 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7660 be put in Local/Makefile. This is like the never_users runtime option,
7661 but it cannot be overridden. The default setting is "root".
7663 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7664 prefix string with which any file named in a -C command line option
7667 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7668 is retained for -C and -D only if the caller of Exim is root. Without
7669 it, the exim user may also use -C and -D and retain privilege.
7671 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7672 command line option is disabled.
7674 6. Macro names set by the -D option must start with an upper case letter, just
7675 like macro names defined in the configuration file.
7677 7. Added "dereference=" facility to LDAP.
7679 8. Two instances of the typo "uknown" in the source files are fixed.
7681 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7682 the Configure-Makefile script screwed up while processing it.
7684 10. Incorporated PCRE 4.4.
7686 11. The SMTP synchronization check was not operating right at the start of an
7687 SMTP session. For example, it could not catch a HELO sent before the client
7688 waited for the greeting. There is now a check for outstanding input at the
7689 point when the greeting is written. Because of the duplex, asynchronous
7690 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7691 way, but not yet received, when the check is performed.
7693 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7694 on TCP/IP sockets, because this apparently causes some broken clients to
7697 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7698 unchanged) from the Cygwin maintainer.
7700 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7701 was supported instead of testing for mbx. Effectively a typo.
7703 15. The spa authenticator server code was not checking that the input it
7704 received was valid base64.
7706 16. The debug output line for the "set" modifier in ACLs was not showing the
7707 name of the variable that was being set.
7709 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7711 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7712 lookup on its own. Something else was needed to trigger the lookup. For
7713 example, a match in host_lookup or the need for a host name in a host list.
7714 Now, if $sender_host_name is referenced and the host name has not yet been
7715 looked up, a lookup is performed. If the lookup fails, the variable remains
7716 empty, and $host_lookup_failed is set to "1".
7718 19. Added "eqi" as a case-independent comparison operator.
7720 20. The saslauthd authentication condition could segfault if neither service
7721 nor realm was specified.
7723 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7724 error message that was logged was misleading, and incoming SMTP
7725 connections were dropped. The message is now more accurate, and temporary
7726 errors are given to SMTP connections.
7728 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7729 (except RSET) with a 421 error, until QUIT is received. However, it was
7730 failing to send a response to QUIT.
7732 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7733 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7734 ACL is now run after the helo_try_verify_hosts code.
7736 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7737 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7738 case-independent, so other case variants are also recognized). Apparently
7739 some systems use these upper case variants.
7741 25. If more than two messages were waiting for the same host, and a transport
7742 filter was specified for the transport, Exim sent two messages over the
7743 same TCP/IP connection, and then failed with "socket operation on non-
7744 socket" when it tried to send the third.
7746 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7748 27. The extern definition of crypt16() in expand.c was not being excluded when
7749 the OS had its own crypt16() function.
7751 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7752 as a preferred synonym for return_size_limit, both as an option and as an
7755 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7757 30. Changed the default configuration ACL to relax the local part checking rule
7758 for addresses that are not in any local domains. For these addresses,
7759 slashes and pipe symbols are allowed within local parts, but the sequence
7760 /../ is explicitly forbidden.
7762 31. SPA server authentication was not clearing the challenge buffer before
7765 32. log_message in a "warn" ACL statement was writing to the reject log as
7766 well as to the main log, which contradicts the documentation and doesn't
7767 seem right (because no rejection is happening). So I have stopped it.
7769 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7770 However, I am unable to do any testing of this.
7772 34. Fixed an infelicity in the appendfile transport. When checking directories
7773 for a mailbox, to see if any needed to be created, it was accidentally
7774 using path names with one or more superfluous leading slashes; tracing
7775 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7777 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7778 contents are added to the log line that is written for every discarded
7779 recipient. (Previously a log_message setting was ignored.)
7781 36. The ${quote: operator now quotes the string if it is empty.
7783 37. The install script runs exim in order to find its version number. If for
7784 some reason other than non-existence or emptiness, which it checks, it
7785 could not run './exim', it was installing it with an empty version number,
7786 i.e. as "exim-". This error state is now caught, and the installation is
7789 38. An argument was missing from the function that creates an error message
7790 when Exim fails to connect to the socket for saslauthd authentication.
7791 This could cause Exim to crash, or give a corrupted message.
7793 39. Added isip, isip4, and isip6 to ${if conditions.
7795 40. The ACL variables $acl_xx are now saved with the message, and can be
7796 accessed later in routers, transports, and filters.
7798 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7799 strings in the file are not string-expanded.
7801 42. If a MAIL command specified a SIZE value that was too large to fit into an
7802 int variable, the check against message_size_limit failed. Such values are
7803 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7804 day this will have to be increased, but I don't think I want to be around
7805 when emails are that large.
7812 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7813 iconv() is not standard in FreeBSD.
7815 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7816 IPv6 enabled. The observed symptom was a segmentation fault on return from
7817 the function os_common_find_running_interfaces() in src/os.c.
7819 3. In the check_special_case() function in daemon.c I had used "errno" as an
7820 argument name, which causes warnings on some systems. This was basically a
7821 typo, since it was named "eno" in the comments!
7823 4. The code that waits for the clock to tick (at a resolution of some fraction
7824 of a second) so as to ensure message-id uniqueness was always waiting for
7825 at least one whole tick, when it could have waited for less. [This is
7826 almost certainly not relevant at current processor speeds, where it is
7827 unlikely to ever wait at all. But we try to future-proof.]
7829 5. The function that sleeps for a time interval that includes fractions of a
7830 second contained a race. It did not block SIGALRM between setting the
7831 timer, and suspending (a couple of lines later). If the interval was short
7832 and the sigsuspend() was delayed until after it had expired, the suspension
7833 never ended. On busy systems this could lead to processes getting stuck for
7836 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7837 process, before it forks any delivery processes. The open lookup caching
7838 mechanism meant that the open file or database connection was passed into
7839 the delivery process. The problem was that delivery processes always tidy
7840 up cached lookup data. This could cause a problem for the next delivery
7841 process started by the queue runner, because the external queue runner
7842 process does not know about the closure. So the next delivery process
7843 still has data in the lookup cache. In the case of a file lookup, there was
7844 no problem because closing a file descriptor in a subprocess doesn't affect
7845 the parent. However, if the lookup was caching a connection to a database,
7846 the connection was closed, and the second delivery process was likely to
7847 see errors such as "PGSQL: query failed: server closed the connection
7848 unexpectedly". The problem has been fixed by closing all cached lookups
7849 in a queue runner before running a delivery process.
7851 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7852 seem to have the "const" qualifier which it has on other OS. I've
7855 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7856 *of the same type* (client or server) with the same public name that an
7857 error should be diagnosed.
7859 9. When Exim looked up a host name for an IP address, but failed to find the
7860 original IP address when looking up the host name (a safety check), it
7861 output the message "<ip address> does not match any IP for NULL", which was
7862 confusing, to say the least. The bug was that the host name should have
7863 appeared instead of "NULL".
7865 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7866 user that is built into the binary, and the -C or -D options is used, root
7867 privilege is dropped before the configuration file is read. In addition,
7868 logging is switched to stderr instead of the normal log files. If the
7869 configuration then re-defines the Exim user, the unprivileged environment
7870 is probably not what is expected, so Exim logs a panic warning message (but
7873 However, if deliver_drop_privilege is set, the unprivileged state may well
7874 be exactly what is intended, so the warning has been cut out in that case,
7875 and Exim is allowed to try to write to its normal log files.
7881 1. smtp_return_error_details was not giving details for temporary sender
7882 or receiver verification errors.
7884 2. Diagnose a configuration error if two authenticators have the same public
7887 3. Exim used not to create the message log file for a message until the first
7888 delivery attempt. This could be confusing when incoming messages were held
7889 for policy or load reasons. The message log file is now created at the time
7890 the message is received, and an initial "Received" line is written to it.
7892 4. The automatically generated man page for command line options had a minor
7893 bug that caused no ill effects; however, a more serious problem was that
7894 the procedure for building the man page automatically didn't always
7895 operate. Consequently, release 4.20 contains an out-of-date version. This
7896 shouldn't happen again.
7898 5. When building Exim with embedded Perl support, the script that builds the
7899 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7900 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7902 6. The freeze_tell option was not being used for messages that were frozen on
7903 arrival, either by an ACL or by local_scan().
7905 7. Added the smtp_incomplete_transaction log selector.
7907 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7908 was accepting AUTH without a new EHLO.
7910 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7911 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7914 10. Logging of TCP/IP connections (when configured) now happens in the main
7915 daemon process instead of the child process, so that the TCP/IP connection
7916 count is more accurate (but it can never be perfect).
7918 11. The use of "drop" in a nested ACL was not being handled correctly in the
7919 outer ACL. Now, if condition failure induced by the nested "drop" causes
7920 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7921 or "require"), the connection is dropped.
7923 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7924 that yield "discard" can now be used with an "accept" or a "discard" verb,
7925 but an error is generated for any others (because I can't see a useful way
7926 to define what should happen).
7928 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7929 are now processed in the same way as lines in the Exim configuration file.
7930 In particular, continuation lines are supported.
7932 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7934 15. Added -ti meaning -t -i.
7936 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7937 domains, and warn by logging if others are found.
7939 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7940 returned from GIFCONF when Exim is trying to find the list of interfaces on
7941 a host. The code in os.c has been modified to copy each ifreq to an aligned
7942 structure in all cases.
7944 Also, in some cases, the returned ifreq's were being copied to a 'struct
7945 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7946 means the last couple of bytes of an IPv6 address could be chopped if the
7947 ifreq contained only a normal sockaddr (14 bytes storage).
7949 18. Named domain lists were not supported in the hosts_treat_as_local option.
7950 An entry such as +xxxx was not recognized, and was treated as a literal
7953 19. Ensure that header lines added by a DATA ACL are included in the reject log
7954 if the ACL subsequently rejects the message.
7956 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7957 MD5 (which is deprecated).
7959 21. When testing a filter file using -bf, Exim was writing a message when it
7960 took the sender from a "From " line in the message, but it was not doing so
7961 when it took $return_path from a Return-Path: header line. It now does.
7963 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7964 with a valid header line field name (a series of printing characters
7965 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7967 23. Changed "disc" in the source to "disk" to conform to the documentation and
7968 the book and for uniformity.
7970 24. Ignore Sendmail's -Ooption=value command line item.
7972 25. When execve() failed while trying to run a command in a pipe transport,
7973 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7974 could be confused with a return value of 69 from the command itself. This
7975 has been changed to 127, the value the shell returns if it is asked to run
7976 a non-existent command. The wording for the related log line suggests a
7977 non-existent command as the problem.
7979 26. If received_header_text expands to an empty string, do not add a Received:
7980 header line to the message. (Well, it adds a token one on the spool, but
7981 marks it "old" so that it doesn't get used or transmitted.)
7983 27. Installed eximstats 1.28 (addition of -nt option).
7985 28. There was no check for failure on the call to getsockname() in the daemon
7986 code. This can fail if there is a shortage of resources on the system, with
7987 ENOMEM, for example. A temporary error is now given on failure.
7989 29. Contrary to the C standard, it seems that in some environments, the
7990 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7991 program. Exim now does this explicitly; it affects the formatting of
7992 timestamps using strftime().
7994 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7995 complaints. I've now initialized all the variables, to avoid this.
7997 32. Header lines added by a system filter were not being "seen" during
7998 transport-time rewrites.
8000 33. The info_callback() function passed to OpenSSL is set up with type void
8001 (*)(SSL *, int, int), as described somewhere. However, when calling the
8002 function (actually a macro) that sets it up, the type void(*)() is
8003 expected. I've put in a cast to prevent warnings from picky compilers.
8005 34. If a DNS black list lookup found a CNAME record, but there were no A
8006 records associated with the domain it pointed at, Exim crashed.
8008 35. If a DNS black list lookup returned more than one A record, Exim ignored
8009 all but the first. It now scans all returned addresses if a particular IP
8010 value is being sought. In this situation, the contents of the
8011 $dnslist_value variable are a list of all the addresses, separated by a
8014 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8015 to accept a host name and all its aliases if the forward lookup for any of
8016 them yielded the IP address of the incoming connection. Now it accepts only
8017 those names whose forward lookup yields the correct IP address. Any other
8018 names are discarded. This closes a loophole whereby a rogue DNS
8019 administrator could create reverse DNS records to break through a
8020 wildcarded host restriction in an ACL.
8022 37. If a user filter or a system filter that ran in a subprocess used any of
8023 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8024 the wrong values were passed to the pipe command ($thisaddress had the
8025 value of $0, $0 had the value of $1, etc). This bug was introduced by
8026 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8028 38. Improved the line breaking for long SMTP error messages from ACLs.
8029 Previously, if there was no break point between 40 and 75 characters, Exim
8030 left the rest of the message alone. Two changes have been made: (a) I've
8031 reduced the minimum length to 35 characters; (b) if it can't find a break
8032 point between 35 and 75 characters, it looks ahead and uses the first one
8033 that it finds. This may give the occasional overlong line, but at least the
8034 remaining text gets split now.
8036 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8037 file descriptors might be low, and that setting 1000 would always raise it.
8038 It turns out that in some environments, the limit is already over 1000 and
8039 that lowering it causes trouble. So now Exim takes care not to decrease it.
8041 40. When delivering a message, the value of $return_path is set to $sender_
8042 address at the start of routing (routers may change the value). By an
8043 oversight, this default was not being set up when an address was tested by
8044 -bt or -bv, which affected the outcome if any router or filter referred to
8047 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8048 log only when "message" or "log_message" are set. However, if one of the
8049 conditions was an address verification, or a call to a nested ACL, the
8050 messages generated by the underlying test were being passed through. This
8051 no longer happens. The underlying message is available in $acl_verify_
8052 message for both "message" and "log_message" expansions, so it can be
8053 passed through if needed.
8055 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8056 new expansion $bh_ to give the encoded byte string without charset
8057 translation. Translation happens only if iconv() is available; HAVE_ICONV
8058 indicates this at build time. HEADERS_CHARSET gives the charset to
8059 translate to; headers_charset can change it in the configuration, and
8060 "headers charset" can change it in an individual filter file.
8062 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8063 that creates RFC 2047 encoded "words" labels them as that charset instead
8064 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8065 expansion operator; (ii) when Exim creates a From: line for a local
8066 message; (iii) when a header line is rewritten to include a "phrase" part.
8068 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8069 buggy, causing it to skip the first lines of messages whose message ID
8070 ended in 'D'. This would not have bitten before Exim release 4.14, because
8071 message IDs were unlikely to end in 'D' before then. The effect was to have
8072 incorrect size information for certain domains.
8074 45. #include "config.h" was missing at the start of the crypt16.c module. This
8075 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8078 46. If there was a timeout during a "random" callout check, Exim treated it as
8079 a failure of the random address, and carried on sending RSET and the real
8080 address. If the delay was just some slowness somewhere, the response to the
8081 original RCPT would be taken as a response to RSET and so on, causing
8082 mayhem of various kinds.
8084 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8085 when I implemented it. It didn't allow for the fact that some option values
8086 may legitimately be negative (e.g. size_addition), and it didn't even do
8087 the right test for positive values.
8089 48. Domain names in DNS records are case-independent. Exim always looks them up
8090 in lower case. Some resolvers return domain names in exactly the case they
8091 appear in the zone file, that is, they may contain uppercase letters. Not
8092 all resolvers do this - some return always lower case. Exim was treating a
8093 change of case by a resolver as a change of domain, similar to a widening
8094 of a domain abbreviation. This triggered its re-routing code and so it was
8095 trying to route what was effectively the same domain again. This normally
8096 caused routing to fail (because the router wouldn't handle the domain
8097 twice). Now Exim checks for this case specially, and just changes the
8098 casing of the domain that it ultimately uses when it transmits the message
8101 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8104 50. If a filter generated a file delivery with a non-absolute name (possible if
8105 no home directory exists for the router), the forbid_file option was not
8108 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8109 the existing equality matching.
8111 52. Exim was using ints instead of ino_t variables in some places where it was
8112 dealing with inode numbers.
8114 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8115 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8116 called TMPDIR, and if it finds it is different, it changes its value.
8118 54. The smtp_printf() function is now made available to local_scan() so
8119 additional output lines can be written before returning. There is also an
8120 smtp_fflush() function to enable the detection of a dropped connection.
8121 The variables smtp_input and smtp_batched_input are exported to
8124 55. Changed the default runtime configuration: the message "Unknown user"
8125 has been removed from the ACL, and instead placed on the localuser router,
8126 using the cannot_route_message feature. This means that any verification
8127 failures that generate their own messages won't get overridden. Similarly,
8128 the "Unrouteable address" message that was in the ACL for unverifiable
8129 relay addresses has also been removed.
8131 56. Added hosts_avoid_esmtp to the smtp transport.
8133 57. The exicyclog script was not checking for the esoteric option
8134 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8135 will work only if exicyclog is run under the appropriate euid.
8137 58. Following a discussion on the list, the rules by which Exim recognises line
8138 endings on incoming messages have been changed. The -dropcr and drop_cr
8139 options are now no-ops, retained only for backwards compatibility. The
8140 following line terminators are recognized: LF CRLF CR. However, special
8141 processing applies to CR:
8143 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8144 nor a local message in the state where . is a terminator.
8146 (ii) If a bare CR is encountered in a header line, an extra space is added
8147 after the line terminator so as not to end the header. The reasoning
8148 behind this is that bare CRs in header lines are most likely either
8149 to be mistakes, or people trying to play silly games.
8151 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8152 was being incorrectly given as 18 bytes larger than it should have been.
8153 This is a VOB (very old bug).
8155 60. This may never have affected anything current, but just in case it has:
8156 When the local host is found other than at the start of a list of hosts,
8157 the local host, those with the same MX, and any that follow, are discarded.
8158 When the list in question was part of a longer list of hosts, the following
8159 hosts (not currently being processed) were also being discarded. This no
8160 longer happens. I'm not sure if this situation could ever has previously
8163 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8166 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8167 header. This is something that is recommended in a new Internet Draft, and
8168 is something that is documented as being done by Sendmail. There are two
8169 possible values. For messages generated by the autoreply transport, Exim
8172 Auto-Submitted: auto-replied
8174 whereas for all other generated messages (e.g. bounces) it adds
8176 Auto-Submitted: auto-generated
8178 63. The "personal" condition in filters now includes a test for the
8179 Auto-Submitted: header. If it contains the string "auto-" the message it
8180 not considered personal.
8182 64. Added rcpt_include_affixes as a generic transport option.
8184 65. Added queue_only_override (default true).
8186 66. Added the syslog_duplication option.
8188 67. If what should have been the first header line of a message consisted of
8189 a space followed by a colon, Exim was mis-interpreting it as a header line.
8190 It isn't of course - it is syntactically invalid and should therefore be
8191 treated as the start of the message body. The misbehaviour could have
8192 caused a number of strange effects, including loss of data in subsequent
8193 header lines, and spool format errors.
8195 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8196 client host had authenticated. This control can now be exercised by an ACL
8197 for more flexibility.
8199 69. By default, callouts do not happen when testing with -bh. There is now a
8200 variant, -bhc, which does actually run the callout code, including
8201 consulting and updating the callout cache.
8203 70. Added support for saslauthd authentication, courtesy of Alexander
8206 71. If statvfs() failed on the spool or log directories while checking their
8207 size for availability, Exim confusingly gave the error "space shortage".
8208 Furthermore, in debugging mode it crashed with a floating point exception.
8209 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8210 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8211 serious problem, Exim now writes to the main and panic logs when this
8212 happens, with details of the failure. It then refuses to accept the
8213 incoming message, giving the message "spool directory problem" or "log
8214 directory problem" with a 421 code for SMTP messages.
8216 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8217 0, 1, and 2 exist, because some OS complain for execs without them (see
8218 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8219 descriptors. However, the code omitted to check that the open succeeded,
8220 causing mysterious errors if for some reason the permissions on /dev/null
8221 got screwed. Now Exim writes a message to the main and panic logs, and
8222 bombs out if it can't open /dev/null.
8224 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8225 interact so that it is all more flexible. It is supposed to remain
8226 backwards compatible. Also added extra_local_interfaces.
8228 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8229 to bomb out with an assertion failure - to the client this appears as a
8230 connection drop. This problem occurs in the part of the code that was taken
8231 from the Samba project. Fortunately, the assertion is in a very simple
8232 function, so I have fixed this by reproducing the function inline in the
8233 one place where it is called, and arranging for authentication to fail
8234 instead of killing the process with assert().
8236 75. The SPA client code was not working when the server requested OEM rather
8237 than Unicode encoding.
8239 76. Added code to make require_files with a specific uid setting more usable in
8240 the case where statting the file as root fails - usually a non-root-mounted
8241 NFS file system. When this happens and the failure is EACCES, Exim now
8242 forks a subprocess and does the per-uid checking as the relevant uid.
8244 77. Added process_log_path.
8246 78. If log_file_path was not explicitly set, a setting of check_log_space or
8247 check_log_inodes was ignored.
8249 79. If a space check for the spool or log partitions fails, the incident is now
8250 logged. Of course, in the latter case the data may get lost...
8252 80. Added the %p formatting code to string_format() so that it can be used to
8253 print addresses in debug_print(). Adjusted all the address printing in the
8254 debugging in store.c to use %p rather than %d.
8256 81. There was a concern that a line of code in smtp_in.c could overflow a
8257 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8258 initially expressed, the concern was not well-founded, because trailing
8259 spaces are removed early. However, if the trailing spaces were followed by
8260 a NULL, they did not get removed, so the overflow was possible. Two fixes
8263 (a) I re-wrote the offending code in a cleaner fashion.
8264 (b) If an incoming SMTP command contains a NULL character, it is rejected
8267 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8268 runs initgroups(), so that if the Exim user is in any additional groups,
8269 they will be used during message reception.
8275 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob