1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
9 after reception to before a subsequent reception. This should
10 mean slightly faster delivery, and also confirmation of reception
13 JH/02 Move from using the pcre library to pcre2. The former is no longer
14 being developed or supported (by the original developer).
16 JH/03 Constification work in the filters module required a major version
17 bump for the local-scan API. Specifically, the "headers_charset"
18 global which is visible via the API is now const and may therefore
19 not be modified by local-scan code.
21 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
22 sendfile() didi not account for the way the ClamAV driver code called it.
24 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
25 time check was being done for every character; replace that with one
28 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
29 sent was prefixed with a length byte.
31 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
32 RFC 5321. Previously Dovecot (at least) would log errors during
35 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
36 in 4.61 without a reason logged. Binaries will be bigger, which might
37 matter on diskspace-constrained systems, but debug is easier.
39 JH/09 Fix macro-definition during "-be" expansion testing. The move to
40 write-protected store for macros had not accounted for these runtime
41 additions; fix by removing this protection for "-be" mode.
43 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
44 handing out large-numbered file descriptors, violating the usual Unix
45 assumption (and required by Posix) that the lowest possible number will be
46 allocated by the kernel when a new one is needed. In the daemon, and any
47 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
48 are not useable for FD_SET() [and hence select()] and overwrite the stack.
49 Assorted crashes happen.
51 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
52 main-section options or in a connect ACL, the value from the first ever
53 connection was never replaced for subsequent connections. Found by
56 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
57 though only once PCRE2 was introduced: the memory accounting used under
58 debug offset allocations by an int, giving a hard trap in early startup.
59 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
61 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
62 with underbars is given. The write-protection of configuration introduced
63 in 4.95 trapped when normalisation was applied to an option not needing
66 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
68 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
69 failing to close the certificates directory, every hour or any time it
72 JH/16 Debugging initiated by an ACL control now continues through into routing
73 and transport processes, when delivery is immediate. Previously debugging
74 stopped any time Exim re-execs.
76 JH/17 The "expand" debug selector now gives more detail, specifically on the
77 result of expansion operators and items.
79 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
80 bad comparison between the option value and the name of the file to
81 be included was done, and a mismatch was wrongly identified.
82 4.88 to 4.95 are affected.
84 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
86 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
87 supplied as the db file-pair basename. Previously, if a directory
88 path was given, for example via the autoreply "once" option, the DB
89 file.pag and file.dir files would be created in that directory's
92 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
93 "taint" log_selector. These were previously deprecated.
95 JH/22 Fix static address-list lookups to properly return the matched item.
96 Previously only the domain part was returned.
98 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
99 the call into OpenSSL to send a TLS Close was being repeated; this
100 resulted in the library waiting for the peer's Close. If that was never
101 sent we waited forever. Fix by tracking send calls.
103 JH/24 The ${run} expansion item now expands its command string elements after
104 splitting. Previously it was before; the new ordering makes handling
105 zero-length arguments simpler. The old ordering can be obtained by
106 appending a new option "preexpand", after a comma, to the "run".
108 JH/25 Taint-check exec arguments for transport-initiated external processes.
109 Previously, tainted values could be used. This affects "pipe", "lmtp" and
110 "queryprogram" transport, transport-filter, and ETRN commands.
111 The ${run} expansion is also affected: in "preexpand" mode no part of
112 the command line may be tainted, in default mode the executable name
115 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
116 the the facility was not passed across execs, and only the first message
117 passed over a connection could use BDAT; any further ones using DATA.
119 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
120 uses $sending_ip_address and an interface is specified.
121 Previously any use of the local address in the EHLO name disabled
122 PIPECONNECT, the common case being to use the rDNS of it.
128 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
129 related applications. Previously an "H" was used where available info
130 says that "M" should be, so change to match.
132 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
133 as arguments, so an implementation trying to copy these into a local
134 buffer was taking a taint-enforcement trap. Fix by using dynamically
135 created buffers. Similar fix for radius expansion condition.
137 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
138 reasonable, eg. to count headers. Fix by using dynamically created
139 buffers rather than a local. Do similar fixes for ACL actions "dcc",
140 "log_reject_target", "malware" and "spam"; the arguments are expanded
141 so could be handling tainted values.
143 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
144 broken the (no-op) support for this sendmail command. Restore it
145 to doing nothing, silently, and returning good status.
147 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
148 record path was given (or the default used) without a leading directory
149 path, an error occurred on trying to open it. Use the transport's working
152 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
153 transport. Previously it was the name on the DNS A-record; use instead
154 the head of the CNAME chain leading there (if there is one). This seems
155 to align better with RFC 6125.
157 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
158 smtp_accept_max_per_host allocated resources which were not released
159 when the limit was exceeded. This eventually crashed the daemon. Fix
160 by adding a release action in that path.
162 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
163 expanded; previously using tainted values was rejected. Fix by using
164 dynamically-created buffers.
166 JH/09 Relax restrictions on ACL verify condition needing access to message
167 headers. Previously they were only permitted in data and non-smtp ACLs;
168 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
169 not_blind, header_sender and header_names_ascii verification.
171 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
172 Previously a macro used one argument twice; when called with the
173 argument as an expression having side-effects, incorrect operation
174 resulted. Use an inlineable function.
176 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
177 held open for a verify callout. Previously this wan not accounted for
178 and a corrupt onward SMTP conversation resulted.
180 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
181 Previously they were embedded naively in the constructed address; when
182 needed, strip the quoting and quote the entire local_part.
183 Also make the inbound_srs expansion condition handle quoting.
185 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
186 excluded, not matching the documentation.
188 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
189 was given for the sqlite_dbfile a trap resulted.
191 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
192 "name" argument resulted in a trap. There is no reason to disallow such;
193 this was a coding error.
195 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
196 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
197 spent suspended, ignoring the POSIX definition. Previously we assumed
198 it did and a constant offset from real time could be used as a correction.
199 Change to using the same clock source for the start-of-message and the
200 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
201 exists, just to get a clock slightly more aligned to reality.
203 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
204 RFC says it is optional some validators care. The missing char was not
205 intended but triggered by a line-wrap alignment. Discovery and fix by
206 Guillaume Outters, hacked on by JH.
208 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
209 name being quoted was tainted a trap would be taken. Fix by using
210 dynamically created buffers. The routine could have been called by a
211 rewrite with the "h" flag, by using the "-F" command-line option, or
212 by using a "name=" option on a control=submission ACL modifier.
214 JH/19 SPF: change the Authentication-Results expansion component to give
215 smtp.helo when the sender domain is empty. Previously it gave
218 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
219 not in 127.0.0.0/8 to help in spotting list domains taken over by a
220 domain-parking registrar.
222 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
223 Previously when a whitespace character was specified it was not inserted
224 after removing the newline.
226 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
227 the domain part of the recipient address. This overrides any tls_sni
228 option set, which was previously used.
230 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
233 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
234 is_tainted() had an off-by-one error in the overenthusiastic direction.
235 Find and fix by Gavan. Although NetBSD is not a supported platform for
236 4.94 this bug could affect other platforms.
238 PP/01 Fix default prime selection to be consistent.
239 One path used ike23 still, instead of exim.dev.20160529.3; now both
240 execution flows will use the same DH primes (currently
241 exim.dev.20160529.3).
243 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
244 option in smtp transport, to match the documentation. Previously
245 verification was not being done in some cases where it should have been.
247 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
248 than one server was defined and depending on the platform memory layout
249 details, an internal consistency trap could be hit while walking the list
252 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
253 files. The value is used by the authresults expansion item. Previously
254 if this was used in a router or transport, a crash could result.
256 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
257 have a sufficient incidence of EINTR returns from select that an
258 interaction with other operations done by the main daemon loop exposed
259 a bug in the error-handling. This was benign apart from the log
262 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
263 for consistency with delivered "=>" and failed "**" lines. While we're
264 there, handle PRX and TFO.
266 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
267 applied. This resulted, if any header-line rewrite rules were configured,
268 in a panic-log triggerable by sending a message with a long address in
269 a header. Fix by increasing the arbitrary limit to larger than a single
270 (dewrapped) 5322 header line maximum size.
272 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
273 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
274 RFC 6648 which deprecates X- options in protocols as a general practice.
275 Changeover between the implementations is handled by the mechanisms
278 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
279 a remote which uses callout/hold. Previously the local was queued.
281 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
284 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
285 log line. The asterisk is now consistently appended to the remote IP
286 (and port, if given), and will also be provided on defer and fail log
287 lines. Previously it could be placed on the local IP if that was being
288 logged, and was only provided on delivery lines.
290 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
292 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
293 authenticator client_send option. Previously the next char, after a pair
294 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
295 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
296 documentation. There is still no way to get a leading ^ immediately
297 after a NUL (ie. for the password of a PLAIN method authenticator.
299 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
300 files. For bad sizes read, delete the record and whine to paniclog.
302 JH/38 When logging an AUTH failure, as server, do not include sensitive
303 information. Previously, the credentials would be included if given
304 as part of the AUTH command line and an ACL denied authentication.
306 JH/39 Bug 2691: fix $local_part_data. When the matching list element
307 referred to a file, bad data was returned. This likely also affected
310 JH/40 The gsasl authenticator now supports caching of the salted password
311 generated by the client-side implementation. This required the addition
312 of a new variable: $auth4.
314 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
315 left undeleted; the attempt to re-create it then failed - resulting in
316 the usual "SIGHUP tp have daemon reload configuration" to not work.
317 This affected any platform not supporting "abstract" Unix-domain
318 sockets (i.e. not Linux).
320 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
321 recipients" response to RCPT in a later response, with a 250. The
322 previous coding assumed this would not happen, and under PIPELINING
323 would result in both lost and duplicate recipients for a message.
325 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
326 Previously the weighting was incorrectly applied. Similar fix for socks
327 proxies. Found and fixed by Heiko Schlichting.
329 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
330 not handle sub-lists included using the +namedlist syntax. While
331 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
332 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
335 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
336 Previously the permanent pool was used, so the sore could not be freed.
337 This meant a connection with many messages would use continually-growing
340 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
341 per-pool so as not to waste too much space. Previously a constant size
342 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
343 DOS attacks harder. The cost is wasted memory use in the larger blocks.
345 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
346 same space repeatedly, and a smaller process footprint.
348 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
349 for GnuTLS. Previously the name started "%s" which, while not a bug,
350 looked as if if might be one.
352 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
353 "continued connection") make the $proxy_* variables available. Previously
354 the information was not passed across the exec() call for subsequent
355 transport executions. This also mean that the log lines for the
356 messages can show the proxy information.
358 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
359 receive time. With modern systems the difference is significant.
360 The historical behaviour can be restored by disabling (a new) log_selector
361 "queue_time_exclusive".
363 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
364 content) were specifically tested for. Now, also cover expansions
365 resulting in ACL names and inline ACL content.
367 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
368 making it unusable in complex expressions.
370 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
371 fail with a taint-check on the spoolfile name, and leave the message
374 HS/01 Enforce absolute PID file path name.
376 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
378 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
380 PP/02 Bug 2643: Correct TLS DH constants.
381 A missing NUL termination in our code-generation tool had led to some
382 incorrect Diffie-Hellman constants in the Exim source.
383 Reported by kylon94, code-gen tool fix by Simon Arlott.
385 PP/03 Impose security length checks on various command-line options.
386 Fixes CVE-2020-SPRSS reported by Qualys.
388 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
389 better. Reported by Qualys.
391 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
392 providing a particularly obnoxious sender full name.
395 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
397 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
400 PP/08 Change default for recipients_max from unlimited to 50,000.
402 PP/09 Fix security issue with too many recipients on a message (to remove a
403 known security problem if someone does set recipients_max to unlimited,
404 or if local additions add to the recipient list).
405 Fixes CVE-2020-RCPTL reported by Qualys.
407 PP/10 Fix security issue in SMTP verb option parsing
408 Fixes CVE-2020-EXOPT reported by Qualys.
410 PP/11 Fix security issue in BDAT state confusion.
411 Ensure we reset known-good where we know we need to not be reading BDAT
412 data, as a general case fix, and move the places where we switch to BDAT
413 mode until after various protocol state checks.
414 Fixes CVE-2020-BDATA reported by Qualys.
416 HS/03 Die on "/../" in msglog file names
418 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
419 the Exim runtime user are allowed to create files.
421 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
424 QS/03 When reading the output from interpreted forward files we do not
425 pass the pipe between the parent and the interpreting process to
426 executed child processes (if any).
428 QS/04 Always die if requested from internal logging, even is logging is
431 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
432 the API; compilation noo longer completes with DMARC support included.
433 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
434 been reported on other platforms.
436 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
438 JH/56 Make the majority of info read from config files readonly, for defence-in-
439 depth against exploits. Suggestion by Qualys.
440 Not supported on Solaris 10.
442 JH/57 Fix control=fakreject for a custom message containing tainted data.
443 Previously this resulted in a log complaint, due to a re-expansion present
444 since fakereject was originally introduced.
446 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
447 modifier was given, a loop resulted.
449 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
450 pipelined SMTP command followed the BDAT LAST then it would be
451 incorrectly treated as part of the message body, causing a verification
454 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
455 require looser validation rules than those for 821-level addresses,
456 which only permit IP addresses.
462 JH/01 Avoid costly startup code when not strictly needed. This reduces time
463 for some exim process initialisations. It does mean that the logging
464 of TLS configuration problems is only done for the daemon startup.
466 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
468 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
469 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
472 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
473 DKIM signing is being done. Previously a transport_filter would always
474 disable CHUNKING, falling back to traditional DATA.
476 JH/05 Regard command-line recipients as tainted.
478 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
480 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
481 PAM library frees one of the arguments given to it, despite the
482 documentation. Therefore a plain malloc must be used.
484 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
485 on-stack buffers were used, resulting in a taint trap when DSN information
486 copied from a received message was written into the buffer.
488 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
489 the ordering of its ARC headers. This caused a crash.
491 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
492 a new record was being constructed with information from the peer, a trap
495 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
496 installation would get error messages from DMARC verify, when it hit the
497 nonexistent file indicated by the default. Distros wanting DMARC enabled
498 should both provide the file and set the option.
499 Also enforce no DMARC verification for command-line sourced messages.
501 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
502 could, depending on the platform, hang at the STARTTLS response.
504 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
505 message on a connection. Previously if one message had ARC headers and
506 the following one did not, a crash could result when adding an
507 Authentication-Results: header.
509 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
510 Exim main code and Exim-related utities. The introduction of taint
511 tracking also did many adjustments to string handling. Since then, eximon
512 frequently terminated with an assert failure.
514 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
515 check for 452 responses. This slightly helps the inefficieny of doing
516 a large alias-expansion into a recipient-limited target. The max_rcpt
517 transport option still applies (and at the current default, will override
518 the new feature). The check is done for either cause of synch, and forces
519 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
520 connection. The new facility is not tunable at this time.
522 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
523 library live data was being used, so the results became garbage. Make
524 copies while it is still usable.
526 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
527 on delivery deferred (==) and failed (**) lines (if a delivery was
528 attemtped). Previously it was only on completion (=>) lines.
530 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
531 for the expansion of the server_scram_iter and server_scram_salt options.
533 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
534 are now specifically given a NO_DATA response without hitting the system
535 resolver. The library goes on to do the now-standard TXT lookup.
536 Use of dnsdb lookups is not affected.
538 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
539 only retrieve the errormessage once. Previously two calls to dlerror()
540 were used, and the second one (for mainlog/paniclog) retrieved null
543 JH/20 Taint checking: disallow use of tainted data for
544 - the appendfile transport file and directory options
545 - the pipe transport command
546 - the autoreply transport file, log and once options
547 - file names used by the redirect router (including filter files)
549 - paths used by single-key lookups
550 Previously this was permitted.
552 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
553 adjusted the size of a major service buffer; this failed because the
554 buffer was in use at the time. Change to a compile-time increase in the
555 buffer size, when this authenticator is compiled into exim.
557 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
558 previous fast-mode was untenable in the face of glibs using mmap to
559 support larger malloc requests.
561 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
562 New values supported, if defined on system where compiled:
563 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
564 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
566 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
567 running a limited number of proceses in parallel, a benefit is gained. The
568 amount varies with the platform hardware and load. The use of the option
569 queue_run_in_order means we cannot do this, as ordering becomes
572 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
573 had introduced a string-copy (for ensuring NUL-termination) which was not
574 appropriate for that case, which can include embedded NUL bytes in the
575 block of data. Investigation showed the copy to actually be needless, the
576 data being length-specified.
578 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
579 done during a receiving connection, and both used TLS, global info was
580 used rather than per-connection info for tracking the state of data
581 queued for transmission. This could result in a connection hang.
583 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
584 Previously, when delivering serveral messages down a single connection
585 only the first would provide a SIZE. This was due to the size information
586 not being properly tracked.
588 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
589 TAI (at 37 seconds currently), pretend to be in UTC for time-related
590 expansion and logging. Previously, spurious values such as a future
591 minute could be seen.
593 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
594 it could crash from a null-deref. This could also affect the
595 ${addresses: } operator and ${readsock } item.
597 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
598 message following a mime one, the variable was not reset.
600 JH/30 When an pipelined-connect fails at the first response, assume incorrect
601 cached capability (perhaps the peer reneged?) and immediately retry in
604 JH/31 Fix spurious detection of timeout while writing to transport filter.
606 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
607 an attempt to copy the string was made before checking it.
609 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
610 taint of the lookup key was maintained; we now regard the presence in the
611 filesystem as sufficient validation.
613 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
614 argument is supplied.
616 JH/35 The dsearch lookup now requires that the directory is an absolute path.
617 Previously this was not checked, and nonempty relative paths made an
618 access under Exim's current working directory.
620 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
621 Previously no event was raised.
623 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
624 parameter supplied by the sender MAIL FROM command. Previously it was
625 ignored, and only the check_spool_space option value for the required
628 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
629 the size of the signing public-key. Previously it was instead giving
630 the size of the signature hash.
632 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
633 the default. See the (new) dkim_verify_min_keysizes option.
635 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
636 and an ACL use a lookup for checking either the local_part or domain,
637 stale data could be accessed. Ensure that variable references are
638 dropped between messages.
640 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
641 by the client was not checked as pointing within response data before
642 being used. A malicious client could thus cause an out-of-bounds read and
643 possibly gain authentication. Fix by adding the check.
645 JH/42 Internationalisation: change the default for downconversion in the smtp
646 transport to be "if needed". Previously it was "as previously set" for
647 the message, which usually meant "if needed" for message-submission but
648 "no" for everything else. However, MTAs have been seen using SMTPUTF8
649 even when the envelope addresses did not need it, resulting in forwarding
650 failures to non-supporting MTAs. A downconvert in such cases will be
651 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
652 transport. The change does mean that addresses needing conversion will
653 be converted when previously a delivery failure would occur.
655 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
656 response was received it would be used unchecked in a fail-DSN, violating
657 standards on line-length limits. Truncate if needed.
659 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
660 doesn't work. (Additionally add a new main config option to configure the
667 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
668 side, to decode a TLS 1.3 packet capture.
670 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
671 Previously the default library behaviour applied, sending two, each in
674 JH/03 Debug output for ACL now gives the config file name and line number for
677 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
679 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
681 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
682 buffer overrun for (non-chunking) other transports.
684 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
685 TLS1.3, means that a server rejecting a client certificate is not visible
686 to the client until the first read of encrypted data (typically the
687 response to EHLO). Add detection for that case and treat it as a failed
688 TLS connection attempt, so that the normal retry-in-clear can work (if
689 suitably configured).
691 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
692 and/or domain. Found and fixed by Jason Betts.
694 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
695 configuration). If a CNAME target was not a wellformed name pattern, a
698 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
699 the OS reports them interleaved with other addresses.
701 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
702 used both for input and for a verify callout, both encrypted, SMTP
703 responses being sent by the server could be lost. This resulted in
704 dropped connections and sometimes bounces generated by a peer sending
707 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
708 string. Previously it was possible to cause undefined behaviour in a
709 library routine (usually a crash). Found by "zerons".
711 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
714 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
715 API was removed, so update to use the newer ones.
717 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
718 any timeout set, is taking a long time. Previously we would hang on to a
719 rotated logfile "forever" if the input was arriving with long gaps
720 (a previous attempt to fix addressed lack, for a long time, of initial
723 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
724 shared (NFS) environment. The length of the tempfile name is now
725 4 + 16 ("hdr.$message_exim_id") which might break on file
726 systems which restrict the file name length to lower values.
729 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
730 shared (NFS) environment.
732 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
733 did for all versions <4.90). Notably -M, -m, --invert, -I may be
736 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
737 on some platforms for bit 31.
739 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
740 to changes apparently associated with TLS1.3 handling some of the APIs
741 previously used were either nonfunctional or inappropriate. Strings
742 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
743 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
744 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
745 This affects log line X= elements, the $tls_{in,out}_cipher variables,
746 and the use of specific cipher names in the encrypted= ACL condition.
748 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
750 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
751 verification result was not updated unless hosts_require_ocsp applied.
753 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
754 queue_list_requires_admin set to false, non-admin users were denied the
757 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
758 directory-of-certs mode. Previously they were advertised despite the
761 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
762 A single TCP connection by a client will now hold a TLS connection open
763 for multiple message deliveries, by default. Previously the default was to
766 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
767 default. If built with the facility, DANE will be used. The facility
768 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
770 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
771 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
772 must be defined and you must still, unless you define DISABLE_TLS, manage
773 the the include-dir and library-file requirements that go with that
774 choice. Non-TLS builds are still supported.
776 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
779 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
780 default. If the platform supports and has the facility enabled, it will
781 be requested on all coneections.
783 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
784 controlled by the build-time option SUPPORT_PIPE_CONNECT.
786 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
788 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
789 success-DSN messages. Previously the From: header was always the default
790 one for these; the option was ignored.
792 JH/28 Fix the timeout on smtp response to apply to the whole response.
793 Previously it was reset for every read, so a teergrubing peer sending
794 single bytes within the time limit could extend the connection for a
795 long time. Credit to Qualsys Security Advisory Team for the discovery.
797 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
798 delivery address, which leaked information of the results of local
799 forwarding. Change to the original envelope recipient address, per
802 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
803 requested. Previously not bounce was generated and a log entry of
804 error ignored was made.
806 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
808 JH/32 Introduce a general tainting mechanism for values read from the input
809 channel, and values derived from them. Refuse to expand any tainted
810 values, to catch one form of exploit.
812 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
813 was unused and the unexpanded text used for the test. Found and
814 fixed by Ruben Jenster.
816 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
817 an attempt to use a TLS library read routine dereffed a nul pointer,
820 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
821 them in our smtp response.
823 JH/36 Have the general router option retry_use_local_part default to true when
824 any of the restrictive preconditions are set (to anything). Previously it
825 was only for check_local user. The change removes one item of manual
826 configuration which is required for proper retries when a remote router
827 handles a subset of addresses for a domain.
829 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
830 link count into consideration.
832 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
833 caused the extension of big_buffer, the following lines were ignored.
835 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
836 accordance with RFC 2308. Previously there was no expiry, so a longlived
837 receive process (eg. due to ACL delays) versus a short SOA value could
840 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
842 JH/39 Promote DMARC support to mainline.
844 JH/40 Bug 2452: Add a References: header to DSNs.
846 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
847 parameters. The relevant library call is documented as "Deprecated: This
848 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
849 3.6.0, DH parameters are negotiated following RFC7919."
851 HS/06 Change the default of dnssec_request_domains to "*"
853 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
854 carried on and emitted a BDAT command, even when PIPELINING was not
857 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
858 buffer was used for the filename, resulting in a trap when tainted
859 arguments (eg. $domain) were used.
861 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
862 recommended to avoid a possible server-load attack. The feature can be
863 re-enabled via the openssl_options main cofiguration option.
865 JH/45 local_scan API: documented the current smtp_printf() call. This changed
866 for version 4.90 - adding a "more data" boolean to the arguments.
867 Bumped the ABI version number also, this having been missed previously;
868 release versions 4.90 to 4.92.3 inclusive were effectively broken in
869 respect of usage of smtp_printf() by either local_scan code or libraries
870 accessed via the ${dlfunc } expansion item. Both will need coding
871 adjustment for any calls to smtp_printf() to match the new function
872 signature; a FALSE value for the new argument is always safe.
874 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
875 the file-offset (which the Linux syscall does, and exim expects); this
876 resulted in an indefinite loop.
878 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
879 to do ARC verification. The Authentication-Results: header line added
880 by the configuration then had no ARC item.
882 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
883 an incorrect mode was used for reading the result, resulting in it being
890 JH/01 Remove code calling the customisable local_scan function, unless a new
891 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
893 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
894 non-signal-safe functions being used.
896 JH/03 Bug 2269: When presented with a received message having a stupidly large
897 number of DKIM-Signature headers, disable DKIM verification to avoid
898 a resource-consumption attack. The limit is set at twenty.
900 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
901 report of oldest_pass in ${authres } in consequence, and separate out
902 some descriptions of reasons for verification fail.
904 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
905 files in the spool were present and unlocked. A queue-runner could spot
906 them, resulting in a duplicate delivery. Fix that by doing the unlock
907 after the unlink. Investigation by Tim Stewart. Take the opportunity to
908 add more error-checking on spoolfile handling while that code is being
911 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
912 No known attacks, no CVE, this is defensive hardening.
914 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
915 a queue-runner could start a delivery while other operations were ongoing.
916 Cutthrough delivery was a common victim, resulting in duplicate delivery.
917 Found and investigated by Tim Stewart. Fix by using the open message data
918 file handle rather than opening another, and not locally closing it (which
919 releases a lock) for that case, while creating the temporary .eml format
920 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
922 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
923 $sender_verify_failure/$recipient_verify_failure to "random".
925 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
928 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
929 Previously this would segfault.
931 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
934 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
935 like zero, since the resolver should be doing this for us, But we need one
936 as a CNAME but no MX presence gets the CNAME returned; we need to check
937 that doesn't point to an MX to declare it "no MX returned" rather than
938 "error, loop". A new main option is added so the older capability of
939 following some limited number of chain links is maintained.
941 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
943 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
944 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
945 now-deprecated earlier definitions used only specified the range up to TLS
946 1.2 (in the older-version library docs).
948 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
950 JH/15 Rework TLS client-side context management. Stop using a global, and
951 explicitly pass a context around. This enables future use of TLS for
952 connections to service-daemons (eg. malware scanning) while a client smtp
953 connection is using TLS; with cutthrough connections this is quite likely.
955 JH/16 Fix ARC verification to do AS checks in reverse order.
957 JH/17 Support a "tls" option on the ${readsocket } expansion item.
959 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
960 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
961 Previously the "utf8" would be re-prepended for every additional message.
963 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
964 Previously thery were accepted, resulting in issues when attempting to
965 forward messages to a non-supporting MTA.
967 PP/02 Let -n work with printing macros too, not just options.
969 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
970 one parent address was copied, and bogus data was used at delivery-logging
971 time. Either a crash (after delivery) or bogus log data could result.
972 Discovery and analysis by Tim Stewart.
974 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
975 Previously if the string ended mid-character, we did not insert the
976 promised '?' replacement.
978 PP/04 Documentation: current string operators work on bytes, not codepoints.
980 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
981 should pack well giving a smaller memory footprint so better caching and
982 therefore performance. Group the declarations where this can't be done so
983 that the byte-sized flag variables are not interspersed among pointer
984 variables, giving a better chance of good packing by the compiler.
986 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
987 non-null, to avoid issues with sites running BATV. Previously reports were
988 sent with an empty envelope sender so looked like bounces.
990 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
991 The ignore_error flag wasn't being returned from the filter subprocess so
992 was not set for later routers. Investigation and fix by Matthias Kurz.
994 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
995 and a msg:complete for the whole, when a message is manually removed using
996 -Mrm. Developement by Matthias Kurz, hacked on by JH.
998 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
999 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1000 hope that is portable enough.
1002 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1003 requiring a known-CA anchor certificate; make it now rely entirely on the
1004 TLSA as an anchor. Checking the name on the leaf cert against the name
1005 on the A-record for the host is still done for TA (but not for EE mode).
1007 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1008 included in delivery lines for non-proxied connections, when compiled with
1009 SUPPORT_SOCKS and running with proxy logging enabled.
1011 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1012 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1013 move the existing event to fire before the normal logging of message
1014 failure so that custom logging is bracketed by normal logging.
1016 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1017 msg:fail:internal event. Developement by Matthias Kurz.
1019 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1020 far too small for todays use of crypto signatures stored there. Go all
1021 the way to the max DNS message size of 64kB, even though this might be
1022 overmuch for IOT constrained device use.
1024 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1025 copy a string over itself. The library routine is documented as not
1026 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1028 JH/32 For main options check_spool_space and check_inode_space, where the
1029 platform supports 64b integers, support more than the previous 2^31 kB
1030 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1031 the previous G, M, k.
1033 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1034 $authenticated_fail_id variable on authentication failure. Previously
1037 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1038 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1039 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1040 GNUTLS_SEC_PARAM_MEDIUM.
1042 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1043 side. Previously we would continue as if no SNI had been received.
1045 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1046 "<" character, which should be a list-separator specification, we walked
1047 off past the nul-terimation.
1049 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1050 causes) even when the retry time is not yet met. Previously they were
1051 not, meaning that when (say) an account was over-quota and temp-rejecting,
1052 and multiple senders' messages were queued, only one sender would get
1053 notified on each configured delay_warning cycle.
1055 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1057 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1058 cipher-suites, an error can be left on the stack even for a succeeding
1059 accept; this results in impossible error messages when a later operation
1062 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1063 return error codes indicating retry. Under TLS1.3 this becomes required.
1065 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1066 it only wrote the new authenticators, resulting in a lack of tracking of
1067 peer changes of ESMTP extensions until the next cache flush.
1069 JH/41 Fix the loop reading a message header line to check for integer overflow,
1070 and more-often against header_maxsize. Previously a crafted message could
1071 induce a crash of the recive process; now the message is cleanly rejected.
1073 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1074 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1080 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1081 When redis_servers is set to a list of > 1 element, and the Redis servers
1082 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1083 case of MOVED into a DEFER case instead, thus moving the query onto the
1084 next server in the list. For a cluster of N elements, all N servers must
1085 be defined in redis_servers.
1087 GF/02 Catch and remove uninitialized value warning in exiqsumm
1088 Check for existence of @ARGV before looking at $ARGV[0]
1090 JH/01 Replace the store_release() internal interface with store_newblock(),
1091 which internalises the check required to safely use the old one, plus
1092 the allocate and data copy operations duplicated in both (!) of the
1093 extant use locations.
1095 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1096 modifier. This matches the restriction on the commandline.
1098 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1099 Previously only the last row was returned.
1101 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1102 we assumed that tags in the header were well-formed, and parsed the
1103 element content after inspecting only the first char of the tag.
1104 Assumptions at that stage could crash the receive process on malformed
1107 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1108 While running the DKIM ACL we operate on the Permanent memory pool so that
1109 variables created with "set" persist to the DATA ACL. Also (at any time)
1110 DNS lookups that fail create cache records using the Permanent pool. But
1111 expansions release any allocations made on the current pool - so a dnsdb
1112 lookup expansion done in the DKIM ACL releases the memory used for the
1113 DNS negative-cache, and bad things result. Solution is to switch to the
1114 Main pool for expansions.
1115 While we're in that code, add checks on the DNS cache during store_reset,
1116 active in the testsuite.
1117 Problem spotted, and debugging aided, by Wolfgang Breyha.
1119 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1120 When none of the hosts presented to a transport match an already-open
1121 connection, close it and proceed with the list. Previously we would
1122 queue the message. Spotted by Lena with Yahoo, probably involving
1125 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1126 Previously a spurious "250 OK id=" response was appended to the proper
1129 JH/08 The "support for" informational output now, which built with Content
1130 Scanning support, has a line for the malware scanner interfaces compiled
1131 in. Interface can be individually included or not at build time.
1133 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1134 by the template makefile "src/EDITME". The "STREAM" support for an older
1135 ClamAV interface method is removed.
1137 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1138 rows affected is given instead).
1140 JH/11 The runtime Berkeley DB library version is now additionally output by
1141 "exim -d -bV". Previously only the compile-time version was shown.
1143 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1144 SMTP connection. Previously, when one had more recipients than the
1145 first, an abortive onward connection was made. Move to full support for
1146 multiple onward connections in sequence, handling cutthrough connection
1147 for all multi-message initiating connections.
1149 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1150 routers. Previously, a multi-recipient message would fail to match the
1151 onward-connection opened for the first recipient, and cause its closure.
1153 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1154 a timeout on read on a GnuTLS initiating connection, resulting in the
1155 initiating connection being dropped. This mattered most when the callout
1156 was marked defer_ok. Fix to keep the two timeout-detection methods
1159 JH/15 Relax results from ACL control request to enable cutthrough, in
1160 unsupported situations, from error to silently (except under debug)
1161 ignoring. This covers use with PRDR, frozen messages, queue-only and
1164 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1166 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1167 metadata, resulting in a crash in free().
1169 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1170 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1171 Broken also in d185889f4, with init system revamp.
1173 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1174 Previously we abruptly closed the connection after reading a malware-
1175 found indication; now we go on to read the "scan ok" response line,
1178 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1179 ACL. Previously, a crash would result.
1181 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1182 macro text after a replacement (previously it was only once per line) and
1183 by skipping builtin macros when searching for an uppercase lead character.
1185 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1186 for the build is renamed.
1188 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1189 was allocated for every new TLS startup, meaning one per message. Fix
1190 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1192 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1193 reported the original. Fix to report (as far as possible) the ACL
1194 result replacing the original.
1196 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1197 OpenSSL. Certificate information is loaded for every new TLS startup,
1198 and the resources needed to be freed.
1200 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1202 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1205 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1206 DATA response info to the (existing) per-recipient response info for
1207 the "C=" log element. It can have useful tracking info from the
1208 destination system. Patch from Simon Arlott.
1210 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1211 length value. Previously this would segfault.
1213 HS/02 Support Avast multiline protoocol, this allows passing flags to
1214 newer versions of the scanner.
1216 JH/28 Ensure that variables possibly set during message acceptance are marked
1217 dead before release of memory in the daemon loop. This stops complaints
1218 about them when the debug_store option is enabled. Discovered specifically
1219 for sender_rate_period, but applies to a whole set of variables.
1220 Do the same for the queue-runner and queue-list loops, for variables set
1221 from spool message files. Do the same for the SMTP per-message loop, for
1222 certain variables indirectly set in ACL operations.
1224 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1225 as a multi-recipient message from a mailinglist manager). The coding had
1226 an arbitrary cutoff number of characters while checking for more input;
1227 enforced by writing a NUL into the buffer. This corrupted long / fast
1228 input. The problem was exposed more widely when more pipelineing of SMTP
1229 responses was introduced, and one Exim system was feeding another.
1230 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1231 receiving system, and refused recipients seen by the sending system
1232 (propating to people being dropped from mailing lists).
1233 Discovered and pinpointed by David Carter.
1235 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1236 replaced by the ${authresults } expansion.
1238 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1240 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1241 allows proper process termination in container environments.
1243 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1244 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1246 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1247 and "err_perm", deprecated since 4.83 when the RFC-defined words
1248 "temperror" and "permerror" were introduced.
1250 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1251 transport-filters or DKIM-signing. The restriction was lost in the
1252 consolidation of verify-callout and delivery SMTP handling.
1253 Extend the restriction to also cover ARC-signing.
1255 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1256 in defer=pass mode supply a 450 to the initiator. Previously the message
1259 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1260 tls_require_ciphers is used as before.
1262 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1263 "pass_unscanned". Only tmpfails from the scanner are written to
1264 the paniclog, as they may require admin intervention (permission
1265 denied, license issues). Other scanner errors (like decompression
1266 bombs) do not cause a paniclog entry.
1268 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1269 Previously it was possible to log spurious information in receive log
1272 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1273 triggered odd behaviour from Outlook Express clients.
1275 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1278 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1279 since the IETF WG has not yet settled on that versus the original
1280 "bare" representation.
1282 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1283 Previously the millisecond value corrupted the output.
1284 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1285 corrupted the output.
1291 JH/01 Rework error string handling in TLS interface so that the caller in
1292 more cases is responsible for logging. This permits library-sourced
1293 string to be attached to addresses during delivery, and collapses
1294 pairs of long lines into single ones.
1296 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1297 during configuration. Wildcards are allowed and expanded.
1299 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1300 This permits better logging.
1302 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1303 do not close it down and have the child transport start it up again on
1304 the passed-on TCP connection. Instead, proxy the child (and any
1305 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1306 affected: the continued delivery log lines do not have any DNSSEC, TLS
1307 Certificate or OCSP information. TLS cipher information is still logged.
1309 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1310 identical IP addresses on different listening ports. Will also affect
1313 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1314 add noisy ifdef guards to special-case this sillyness.
1315 Patch from Bernd Kuhls.
1317 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1318 than 255 are no longer allowed.
1320 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1321 Disable the session-cache too, which might reduce our load. Since we
1322 currrectly use a new context for every connection, both as server and
1323 client, there is no benefit for these.
1324 GnuTLS appears to not support tickets server-side by default (we don't
1325 call gnutls_session_ticket_enable_server()) but client side is enabled
1326 by default on recent versions (3.1.3 +) unless the PFS priority string
1329 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1330 <https://reproducible-builds.org/specs/source-date-epoch/>.
1332 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1333 the check for any unsuccessful recipients did not notice the limit, and
1334 erroneously found still-pending ones.
1336 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1337 MSG_MORE. Only in-clear (not on TLS connections).
1339 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1340 transport-filter is involved we can buffer the headers in memory for
1341 creating the signature, and read the spool data file once for the
1342 signature and again for transmission.
1344 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1345 4.77 as the kernel support then wasn't solid, having issues in 64bit
1346 mode. Now, it's been long enough. Add support for FreeBSD also.
1348 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1349 case where the routing stage had gathered several addresses to send to
1350 a host before calling the transport for the first, we previously failed
1351 to close down TLS in the old transport process before passing the TCP
1352 connection to the new process. The new one sent a STARTTLS command
1353 which naturally failed, giving a failed delivery and bloating the retry
1354 database. Investigation and fix prototype from Wolfgang Breyha.
1356 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1357 false-negatives in the check that the sender had not preempted a response
1358 or prompt from Exim (running as a server), due to that code's lack of
1359 awareness of the SMTP input buffering.
1361 PP/04 Add commandline_checks_require_admin option.
1362 Exim drops privileges sanely, various checks such as -be aren't a
1363 security problem, as long as you trust local users with access to their
1364 own account. When invoked by services which pass untrusted data to
1365 Exim, this might be an issue. Set this option in main configuration
1366 AND make fixes to the calling application, such as using `--` to stop
1369 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1370 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1371 responses to those, into a single TLS record each way (this usually means
1372 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1375 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1376 affects you only if you're dancing at the edge of the param size limits.
1377 If you are, and this message makes sense to you, then: raise the
1378 configured limit or use OpenSSL 1.1. Nothing we can do for older
1381 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1382 cmdline element to get the documented default one. Previously it was
1385 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1386 are unsuable for non-message-specific reasons - eg. connection timeout,
1387 banner-time rejection.
1389 JH/16 Fix logging of delivery remote port, when specified by router, under
1392 PP/06 Repair manualroute's ability to take options in any order, even if one
1393 is the name of a transport.
1396 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1398 JH/17 Change the list-building routines interface to use the expanding-string
1399 triplet model, for better allocation and copying behaviour.
1401 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1402 Previously it was constructed the first time a possibly-matching string
1403 was met in the configuration file input during startup; now it is done
1406 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1407 compatible one, to avoid the (poorly documented) possibility of a config
1408 file in the working directory redirecting the DB files, possibly correpting
1409 some existing file. CVE-2017-10140 assigned for BDB.
1411 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1412 cache-hot. Previously, although the result was properly cached, the
1413 initial verify call returned a defer.
1415 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1416 the main verify for receipient in uncached-mode.
1418 JH/22 Retire historical build files to an "unsupported" subdir. These are
1419 defined as "ones for which we have no current evidence of testing".
1421 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1422 if present. Previously it was ignored.
1424 JH/24 Start using specified-initialisers in C structure init coding. This is
1425 a C99 feature (it's 2017, so now considered safe).
1427 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1428 if was a fixed-sized field and bitmask ops via macros; it is now more
1431 PP/07 GitHub PR 56: Apply MariaDB build fix.
1432 Patch provided by Jaroslav Škarvada.
1434 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1435 during Coverity cleanups [4.87 JH/47]
1436 Diagnosis and fix provided by Michael Fischer v. Mollard.
1438 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1439 the right size to place the terminating semicolon on its own folded
1440 line, the header hash was calculated to an incorrect value thanks to
1441 the (relaxed) space the fold became.
1443 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1444 and confused the parent.
1446 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1447 which could crash as a result. This could lead to undeliverable messages.
1449 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1452 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1453 matters on fast-turnover and PID-randomising systems, which were getting
1454 out-of-order delivery.
1456 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1457 a possibly-overlapping copy. The symptom was that "Remote host closed
1458 connection in response to HELO" was logged instead of the actual 4xx
1461 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1462 Previously only that bufferd was discarded, resulting in SYMTP command
1465 JH/32 DKIM: when a message has multiple signatures matching an identity given
1466 in dkim_verify_signers, run the dkim acl once for each. Previously only
1467 one run was done. Bug 2189.
1469 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1470 "panic the current process" to "deliberately defer". The panic log is
1471 still written with the problem list name; the mail and reject logs now
1472 get a temp-reject line for the message that was being handled, saying
1473 something like "domains check lookup or other defer". The SMTP 451
1474 message is still "Temporary local problem".
1476 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1477 A crafted sequence of BDAT commands could result in in-use memory beeing
1478 freed. CVE-2017-16943.
1480 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1481 from SMTP input. Previously it was always done; now only done for DATA
1482 and not BDAT commands. CVE-2017-16944.
1484 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1485 to the message (such as an overlong header line). Previously this was
1486 not done and we did not exit BDAT mode. Followon from the previous item
1487 though a different problem.
1493 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1494 than -2003 did; needs libidn2 in addition to libidn.
1496 JH/02 The path option on a pipe transport is now expanded before use.
1498 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1499 Patch provided by "Björn", documentation fix added too.
1501 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1502 missing a wire-to-host endian conversion.
1504 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1505 close after a BDAT command line could be taken as a following command,
1506 giving a synch failure. Fix by only checking for synch immediately
1507 before acknowledging the chunk.
1509 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1510 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1511 macro. Patches provided by Josh Soref.
1513 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1514 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1515 by IANA as a service extension. However, John Klensin suggests that we
1518 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1519 the dkim code may be unix-mode line endings rather than smtp wire-format
1520 CRLF, so prepend a CR to any bare LF.
1522 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1523 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1525 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1526 the first were themselves being wrongly included in the feed into dkim
1527 processing; with most chunk sizes in use this resulted in an incorrect
1528 body hash calculated value.
1530 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1531 DKIM signature block, for verification. Although advised against by
1532 standards it is specifically not ruled illegal.
1534 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1536 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1537 missing a body hash (the bh= tag).
1539 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1540 It seems that HAProxy sends the Proxy Protocol information in clear and
1541 only then does a TLS startup, so do the same.
1543 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1544 TCP connections (such as for Spamd) unless the daemon successfully set
1545 Fast Open mode on its listening sockets. This fixes breakage seen on
1546 too-old kernels or those not configured for Fast Open, at the cost of
1547 requiring both directions being enabled for TFO, and TFO never being used
1548 by non-daemon-related Exim processes.
1550 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1551 endings, at least on the first header line. Try to canonify any that get
1552 past that check, despite the cost.
1554 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1555 now limited to an arbitrary five deep, while parsing addresses with the
1556 strip_excess_angle_brackets option enabled.
1558 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1559 instead leave the unprompted TLS handshake in socket buffer for the
1560 TLS library to consume.
1562 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1564 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1566 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1567 region in one operation, for speed, and this leaves assigned pointers
1568 dangling. Add checks run only under the testsuite which checks all
1569 variables at a store-reset and panics on a dangling pointer; add code
1570 explicitly nulling out all the variables discovered. Fixes one known
1571 bug: a transport crash, where a dangling pointer for $sending_ip_address
1572 originally assigned in a verify callout, is re-used.
1574 PP/06 Drop '.' from @INC in various Perl scripts.
1576 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1578 PP/08 Reduce a number of compilation warnings under clang; building with
1579 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1580 should be warning-free.
1582 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1584 HS/01 Fix portability problems introduced by PP/08 for platforms where
1585 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1587 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1588 chunk. This allows us to accept broken chunked messages. We need a more
1589 general solution here.
1591 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1592 already-broken messages in the queue.
1594 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1596 JH/19 Fix reference counting bug in routing-generated-address tracking.
1602 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1603 supports it and a size is available (ie. the sending peer gave us one).
1605 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1606 years of being deprecated). The replacements are the ACLs
1607 acl_smtp_mime and acl_not_smtp_mime.
1609 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1610 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1611 or even an in-clear connection were permitted. Now, if the host lookup
1612 was dnssec and dane was requested then the host is only used if the
1613 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1614 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1615 if one fails this test.
1616 This means that a poorly-configured remote DNS will make it incommunicado;
1617 but it protects against a DNS-interception attack on it.
1619 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1620 non-noisy when a race steals the message being considered.
1622 JH/05 If main configuration option tls_certificate is unset, generate a
1623 self-signed certificate for inbound TLS connections.
1625 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1626 in rewrites and routers.
1628 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1629 and logged a warning sing 4.83; now they are a configuration file error.
1631 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1632 (lacking @domain). Apply the same qualification processing as RCPT.
1634 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1636 JH/10 Support ${sha256:} applied to a string (as well as the previous
1639 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1640 a cutthrough deliver is pending, as we always want to make a connection.
1641 This also avoids re-routing the message when later placing the cutthrough
1642 connection after a verify cache hit.
1643 Do not update it with the verify result either.
1645 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1646 when routing results in more than one destination address.
1648 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1649 signing (which inhibits the cutthrough capability). Previously only
1650 the presence of an option was tested; now an expansion evaluating as
1651 empty is permissible (obviously it should depend only on data available
1652 when the cutthrough connection is made).
1654 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1655 the relevant preceding SMTP command did not note the pipelining mode.
1657 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1658 Previously they were not counted.
1660 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1661 as one having no matching records. Previously we deferred the message
1662 that needed the lookup.
1664 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1665 distinguished as "(=".
1667 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1668 for missing MX records. Previously it only worked for missing A records.
1670 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1672 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1673 after the data-go-ahead and data-ack. Patch from Jason Betts.
1675 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1676 even for a "none" policy. Patch from Tony Meyer.
1678 JH/22 Fix continued use of a connection for further deliveries. If a port was
1679 specified by a router, it must also match for the delivery to be
1682 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1683 When one of the recipients of a message was unsuitable for the connection
1684 (has no matching addresses), we lost track of needing to mark it
1685 deferred. As a result mail would be lost.
1687 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1689 JH/25 Decoding ACL controls is now done using a binary search; the source code
1690 takes up less space and should be simpler to maintain. Merge the ACL
1691 condition decode tables also, with similar effect.
1693 JH/26 Fix problem with one_time used on a redirect router which returned the
1694 parent address unchanged. A retry would see the parent address marked as
1695 delivered, so not attempt the (identical) child. As a result mail would
1698 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1699 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1700 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1703 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1704 Main config options check_{log,spool}_{inodes,space} are now
1705 100 inodes, 10MB unless set otherwise in the configuration.
1707 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1708 Previously it only applied to the main-section connection policy
1711 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1713 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1714 by me. Added RFC7919 DH primes as an alternative.
1716 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1717 are not in the system include path.
1719 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1720 GnuTLS, when a session startup failed (eg because the client disconnected)
1721 Exim did stdio operations after fclose. This was exposed by a recent
1722 change which nulled out the file handle after the fclose.
1724 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1725 signed directly by the cert-signing cert, rather than an intermediate
1726 OCSP-signing cert. This is the model used by LetsEncrypt.
1728 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1730 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1731 an incoming connection.
1733 HS/02 Bug 1802: Do not half-close the connection after sending a request
1736 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1737 fallback to "prime256v1".
1739 JH/34 SECURITY: Use proper copy of DATA command in error message.
1740 Could leak key material. Remotely exploitable. CVE-2016-9963.
1746 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1747 and 3.4.4 - once the server is enabled to respond to an OCSP request
1748 it does even when not requested, resulting in a stapling non-aware
1749 client dropping the TLS connection.
1751 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1752 support variable-length bit vectors. No functional change.
1754 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1755 The I= interface field on outgoing lines is now after the H= remote
1756 host field, same as incoming lines. There is a separate
1757 outgoing_interface log selector which allows you to disable the
1760 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1761 If not running log_selector +smtp_connection the mainlog would be held
1762 open indefinitely after a "too many connections" event, including to a
1763 deleted file after a log rotate. Leave the per net connection logging
1764 leaving it open for efficiency as that will be quickly detected by the
1765 check on the next write.
1767 HS/01 Bug 1671: Fix post transport crash.
1768 Processing the wait-<transport> messages could crash the delivery
1769 process if the message IDs didn't exist for some reason. When
1770 using 'split_spool_directory=yes' the construction of the spool
1771 file name failed already, exposing the same netto behaviour.
1773 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1774 mime_regex ACL conditions.
1776 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1777 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1778 to HELO, local diagnostic string.
1780 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1781 log line to debug. Even when configured with a tls authenticator many
1782 client connections are expected to not authenticate in this way, so
1783 an authenticate fail is not an error.
1785 HS/02 Add the Exim version string to the process info. This way exiwhat
1786 gives some more detail about the running daemon.
1788 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1789 matter for fast-change records such as DNSBLs.
1791 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1792 retry record, even if constant. There may be multiple transports with
1793 different interface settings and the retry behaviour needs to be kept
1796 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1798 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1800 JH/10 Bug 840: fix log_defer_output option of pipe transport
1802 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1803 hosts_randomize. This matters a lot when combined with mua_wrapper.
1805 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1806 ${quote_pgsql:<string>} operator.
1808 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1810 JH/14 Change header file name for freeradius-client. Relevant if compiling
1811 with Radius support; from the Gentoo tree and checked under Fedora.
1813 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1815 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1816 option transport_filter, meaning no filtering.
1818 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1820 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1821 defaults to "*" (all hosts). The variable is now available when not built
1822 with TLS, default unset, mainly to enable keeping the testsuite sane.
1823 If a server certificate is not supplied (via tls_certificate) an error is
1824 logged, and clients will find TLS connections fail on startup. Presumably
1825 they will retry in-clear.
1826 Packagers of Exim are strongly encouraged to create a server certificate
1827 at installation time.
1829 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1830 with the $config_file variable.
1832 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1833 in transport context, after the attempt, and per-recipient. The latter type
1834 is per host attempted. The event data is the error message, and the errno
1835 information encodes the lookup type (A vs. MX) used for the (first) host,
1836 and the trailing two digits of the smtp 4xx response.
1838 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1839 to write to mainlog (or rejectlog, paniclog) in the window between file
1840 creation and permissions/ownership being changed. Particularly affects
1841 installations where exicyclog is run as root, rather than exim user;
1842 result is that the running daemon panics and dies.
1844 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1846 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1847 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1848 "pri" and "weight". Note that the previous implicit priority given by the
1849 list order is no longer honoured.
1851 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1852 for DKIM processing.
1854 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1855 by defining SUPPORT_SOCKS.
1857 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1858 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1859 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1860 variables are renamed to proxy_{local,external}_{address,port}.
1862 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1863 for a build by defining SUPPORT_I18N
1865 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1866 of the query string, and make ${quote_redis:} do that quoting.
1868 JH/29 Move Events support from Experimental to mainline, enabled by default
1869 and removable for a build by defining DISABLE_EVENT.
1871 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1873 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1874 cached by the daemon.
1876 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1877 by defining LOOKUP_REDIS. The libhiredis library is required.
1879 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1880 keys are given for lookup.
1882 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1883 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1884 only supported when built with TLS support. The PolarSSL SHA routines
1885 are still used when the TLS library is too old for convenient support.
1887 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1888 openssl_options), for security. OpenSSL forces this from version 1.1.0
1889 server-side so match that on older versions.
1891 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1892 allocation for $value could be released as the expansion processing
1893 concluded, but leaving the global pointer active for it.
1895 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1896 and to use the domains and local_parts ACL conditions.
1898 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1899 incorrectly not doubled on cutthrough transmission, hence seen as a
1900 body-termination at the receiving system - resulting in truncated mails.
1901 Commonly the sender saw a TCP-level error, and retransmitted the message
1902 via the normal store-and-forward channel. This could result in duplicates
1903 received - but deduplicating mailstores were liable to retain only the
1904 initial truncated version.
1906 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1908 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1910 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1911 we're in there, support oversigning also; bug 1309.
1913 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1915 HS/04 Add support for keep_environment and add_environment options.
1917 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1918 either intentional arithmetic overflow during PRNG, or testing config-
1921 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1922 delivery resulted in actual delivery. Cancel cutthrough before DATA
1925 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1926 rejects a recipient: pass the reject to the originator.
1928 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1929 Many were false-positives and ignorable, but it's worth fixing the
1932 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1933 for the new environment-manipulation done at startup. Move the routines
1934 from being local to tls.c to being global via the os.c file.
1936 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1937 an extract embedded as result-arg for a map, the first arg for extract
1938 is unavailable so we cannot tell if this is a numbered or keyed
1939 extraction. Accept either.
1945 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1948 JH/02 The smtp transport option "multi_domain" is now expanded.
1950 JH/03 The smtp transport now requests PRDR by default, if the server offers
1953 JH/04 Certificate name checking on server certificates, when exim is a client,
1954 is now done by default. The transport option tls_verify_cert_hostnames
1955 can be used to disable this per-host. The build option
1956 EXPERIMENTAL_CERTNAMES is withdrawn.
1958 JH/05 The value of the tls_verify_certificates smtp transport and main options
1959 default to the word "system" to access the system default CA bundle.
1960 For GnuTLS, only version 3.0.20 or later.
1962 JH/06 Verification of the server certificate for a TLS connection is now tried
1963 (but not required) by default. The verification status is now logged by
1964 default, for both outbound TLS and client-certificate supplying inbound
1967 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1970 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1971 Status Notification (bounce) messages are now MIME format per RFC 3464.
1972 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1973 under the control of the dsn_advertise_hosts option, and routers may
1974 have a dsn_lasthop option.
1976 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1977 default, modifiable by a malware= option. The list separator for
1978 the options can now be changed in the usual way. Bug 68.
1980 JH/10 The smtp_receive_timeout main option is now expanded before use.
1982 JH/11 The incoming_interface log option now also enables logging of the
1983 local interface on delivery outgoing connections.
1985 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1986 if the interface and destination host and port all match.
1988 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1991 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1992 Patch from Andrew Lewis.
1994 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1995 now supports optional time-restrictions, weighting, and priority
1996 modifiers per server. Patch originally by <rommer@active.by>.
1998 JH/16 The spamd_address main option now supports a mixed list of local
1999 and remote servers. Remote servers can be IPv6 addresses, and
2000 specify a port-range.
2002 JH/17 Bug 68: The spamd_address main option now supports an optional
2003 timeout value per server.
2005 JH/18 Bug 1581: Router and transport options headers_add/remove can
2006 now have the list separator specified.
2008 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2011 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2014 JH/21 Support for the A6 type of dns record is withdrawn.
2016 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2017 rather than the verbs used.
2019 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2020 from 255 to 1024 chars.
2022 JH/24 Verification callouts now attempt to use TLS by default.
2024 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2025 are generic router options now. The defaults didn't change.
2027 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2028 Original patch from Alexander Shikoff, worked over by JH.
2030 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2031 files if ZCAT_COMMAND is not executable.
2033 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2035 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2037 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2038 Normally benign, it bites when the pair was led to by a CNAME;
2039 modern usage is to not canonicalize the domain to a CNAME target
2040 (and we were inconsistent anyway for A-only vs AAAA+A).
2042 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2044 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2045 when evaluating $sender_host_dnssec.
2047 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2048 $sender_helo_dnssec variable.
2050 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2052 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2054 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2056 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2057 documented as working, but never had. Support all but $spam_report.
2059 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2060 added for tls authenticator.
2062 HS/03 Add perl_taintmode main config option
2068 TL/01 When running the test suite, the README says that variables such as
2069 no_msglog_check are global and can be placed anywhere in a specific
2070 test's script, however it was observed that placement needed to be near
2071 the beginning for it to behave that way. Changed the runtest perl
2072 script to read through the entire script once to detect and set these
2073 variables, reset to the beginning of the script, and then run through
2074 the script parsing/test process like normal.
2076 TL/02 The BSD's have an arc4random API. One of the functions to induce
2077 adding randomness was arc4random_stir(), but it has been removed in
2078 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2079 function when detected.
2081 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2082 cause callback expansion.
2084 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2085 syntax errors in an expansion can be treated as a string instead of
2086 logging or causing an error, due to the internal use of bool_lax
2087 instead of bool when processing it.
2089 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2090 server certificates when making smtp deliveries.
2092 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2094 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2096 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2098 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2099 Merged patch from Sebastian Wiedenroth.
2101 JH/05 Fix results-pipe from transport process. Several recipients, combined
2102 with certificate use, exposed issues where response data items split
2103 over buffer boundaries were not parsed properly. This eventually
2104 resulted in duplicates being sent. This issue only became common enough
2105 to notice due to the introduction of connection certificate information,
2106 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2108 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2109 size buffer was used, resulting in syntax errors when an expansion
2112 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2113 version 3.3.6 or later.
2115 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2116 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2117 both become "event_action", the variables become $event_name, $event_data
2118 and $event_defer_errno. There is a new variable $verify_mode, usable in
2119 routers, transports and related events. The tls:cert event is now also
2120 raised for inbound connections, if the main configuration event_action
2123 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2124 early OCSP support, but no stapling (appears to be less than 1.0.0).
2126 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2127 server certificate names available under the smtp transport option
2128 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2131 JH/10 Time-related extraction expansions from certificates now use the main
2132 option "timezone" setting for output formatting, and are consistent
2133 between OpenSSL and GnuTLS compilations. Bug 1541.
2135 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2136 encoded parameter in the incoming message. Bug 1558.
2138 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2139 include certificate info, eximon was claiming there were spoolfile
2142 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2144 JH/14 Log delivery-related information more consistently, using the sequence
2145 "H=<name> [<ip>]" wherever possible.
2147 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2148 are problematic for Debian distribution, omit them from the release
2151 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2153 JH/16 Fix string representation of time values on 64bit time_t architectures.
2156 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2157 output list separator was used.
2162 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2163 checkers that were complaining about end of non-void function with no
2166 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2167 This was a regression introduced in 4.83 by another bugfix.
2169 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2171 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2172 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2178 TF/01 Correctly close the server side of TLS when forking for delivery.
2180 When a message was received over SMTP with TLS, Exim failed to clear up
2181 the incoming connection properly after forking off the child process to
2182 deliver the message. In some situations the subsequent outgoing
2183 delivery connection happened to have the same fd number as the incoming
2184 connection previously had. Exim would try to use TLS and fail, logging
2185 a "Bad file descriptor" error.
2187 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2188 utilities have not been installed.
2190 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2191 temporary space as the ACL may create new global variables.
2193 TL/01 LDAP support uses per connection or global context settings, depending
2194 upon the detected version of the libraries at build time.
2196 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2197 to extract and use the src ip:port in logging and expansions as if it
2198 were a direct connection from the outside internet. PPv2 support was
2199 updated based on HAProxy spec change in May 2014.
2201 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2203 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2204 Properly escape header and check for NULL return.
2206 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2209 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2211 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2212 characters in header names, implemented as a verify condition.
2213 Contributed by Michael Fischer v. Mollard.
2215 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2216 results permerror and temperror. Previous values are deprecated but
2217 still accepted. In a future release, err_perm and err_temp will be
2218 completely removed, which will be a backward incompatibility if the
2219 ACL tests for either of these two old results. Patch contributed by
2220 user bes-internal on the mailing list.
2222 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2224 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2225 selectors, in both main and reject logs.
2227 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2230 JH/07 Add malware type "sock" for talking to simple daemon.
2232 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2234 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2235 routers/transports under cutthrough routing.
2237 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2238 numbers. Touch up "bool" conditional to keep the same definition.
2240 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2242 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2244 JH/12 Expand items in router/transport headers_add or headers_remove lists
2245 individually rather than the list as a whole. Bug 1452.
2247 Required for reasonable handling of multiple headers_ options when
2248 they may be empty; requires that headers_remove items with embedded
2249 colons must have them doubled (or the list-separator changed).
2251 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2252 view the policy declared in the DMARC record. Currently, $dmarc_status
2253 is a combined value of both the record presence and the result of the
2256 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2258 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2259 dnslookup router and the smtp transport (applying to the forward
2262 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2263 of ldap servers used for a specific lookup. Patch provided by Heiko
2266 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2267 New variable $lookup_dnssec_authenticated for observability.
2269 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2270 Patch submitted by Lars Timman.
2272 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2274 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2275 Requires trusted mode and valid format message id, aborts otherwise.
2276 Patch contributed by Heiko Schlichting.
2278 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2279 certextract with support for various fields. Bug 1358.
2281 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2282 is requested by default, modifiable by smtp transport option
2285 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2286 operate on certificate variables to give certificate fingerprints
2287 Also new ${sha256:cert_variable}.
2289 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2291 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2292 Christian Aistleitner.
2294 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2296 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2297 file. Patch from Wolfgang Breyha.
2299 JH/25 Expand the coverage of the delivery $host and $host_address to
2300 client authenticators run in verify callout. Bug 1476.
2302 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2303 align with daemon_smtp_ports. Bug 72.
2305 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2306 support and error reporting did not work properly.
2308 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2309 and is readable. Patch from Andrew Colin Kissa.
2311 TL/14 Enhance documentation of ${run expansion and how it parses the
2312 commandline after expansion, particularly in the case when an
2313 unquoted variable expansion results in an empty value.
2315 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2317 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2318 by renaming away from it. Addresses GH issue 15, reported by
2321 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2323 TL/15 SECURITY: prevent double expansion in math comparison functions
2324 (can expand unsanitized data). Not remotely exploitable.
2331 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2333 PP/02 Make -n do something, by making it not do something.
2334 When combined with -bP, the name of an option is not output.
2336 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2339 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2340 $sender_host_name and config options to manage this, and basic check
2343 PP/05 DSCP support for outbound connections and control modifier for inbound.
2345 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2346 (Only plugin which currently uses this is kerberos4, which nobody should
2347 be using, but we should make it available and other future plugins might
2348 conceivably use it, even though it would break NAT; stuff *should* be
2349 using channel bindings instead).
2351 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2352 name; added for Sendmail compatibility; requires admin caller.
2353 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2354 just ignore it); requires trusted caller.
2355 Also parse but ignore: -Ac -Am -X<logfile>
2358 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2360 TL/02 Add +smtp_confirmation as a default logging option.
2362 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2363 Patch by Magnus Holmgren from 2007-02-20.
2365 TL/04 Bugzilla 1281 - Spec typo.
2366 Bugzilla 1283 - Spec typo.
2367 Bugzilla 1290 - Spec grammar fixes.
2369 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2371 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2373 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2374 mailing list by by Dmitry Isaikin.
2376 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2378 JH/02 Support "G" suffix to numbers in ${if comparisons.
2380 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2382 NM/01 Bugzilla 1197 - Spec typo
2383 Bugzilla 1196 - Spec examples corrections
2385 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2387 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2388 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2391 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2392 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2394 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2395 "acl {{name}{arg}...}", and optional args on acl condition
2398 JH/05 Permit multiple router/transport headers_add/remove lines.
2400 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2402 JH/07 Avoid using a waiting database for a single-message-only transport.
2403 Performance patch from Paul Fisher. Bugzilla 1262.
2405 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2408 JH/09 Add $headers_added variable, with content from use of ACL modifier
2409 add_header (but not yet added to the message). Bugzilla 199.
2411 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2412 Pulled from Bugzilla 817 by Wolfgang Breyha.
2414 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2416 (nb: this is the same fix as in Exim 4.80.1)
2418 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2421 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2424 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2426 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2427 advertise SMTP AUTH mechanism to us, instead of a generic
2428 protocol violation error. Also, make Exim more robust to bad
2429 data from the Dovecot auth socket.
2431 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2433 When a queue runner is handling a message, Exim first routes the
2434 recipient addresses, during which it prunes them based on the retry
2435 hints database. After that it attempts to deliver the message to
2436 any remaining recipients. It then updates the hints database using
2439 So if a recipient address works intermittently, it can get repeatedly
2440 deferred at routing time. The retry hints record remains fresh so the
2441 address never reaches the final cutoff time.
2443 This is a fairly common occurrence when a user is bumping up against
2444 their storage quota. Exim had some logic in its local delivery code
2445 to deal with this. However it did not apply to per-recipient defers
2446 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2448 This change adds a proper retry rule check during routing so that the
2449 final cutoff time is checked against the message's age. We only do
2450 this check if there is an address retry record and there is not a
2451 domain retry record; this implies that previous attempts to handle
2452 the address had the retry_use_local_parts option turned on. We use
2453 this as an approximation for the destination being like a local
2454 delivery, as in LMTP.
2456 I suspect this new check makes the old local delivery cutoff check
2457 redundant, but I have not verified this so I left the code in place.
2459 TF/02 Correct gecos expansion when From: is a prefix of the username.
2461 Test 0254 submits a message to Exim with the header
2465 When I ran the test suite under the user fanf2, Exim expanded
2466 the header to contain my full name, whereas it should have added
2467 a Resent-Sender: header. It erroneously treats any prefix of the
2468 username as equal to the username.
2470 This change corrects that bug.
2472 GF/01 DCC debug and logging tidyup
2473 Error conditions log to paniclog rather than rejectlog.
2474 Debug lines prefixed by "DCC: " to remove any ambiguity.
2476 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2478 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2479 Bug spotted by Jeremy Harris; was flawed since initial commit.
2480 Would have resulted in OCSP responses post-SNI triggering an Exim
2481 NULL dereference and crash.
2483 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2485 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2486 Bug detection, analysis and fix by Samuel Thibault.
2487 Bugzilla 1331, Debian bug #698092.
2489 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2491 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2492 Server implementation by Todd Lyons, client by JH.
2493 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2494 config variable "prdr_enable" controls whether the server
2495 advertises the facility. If the client requests PRDR a new
2496 acl_data_smtp_prdr ACL is called once for each recipient, after
2497 the body content is received and before the acl_smtp_data ACL.
2498 The client is controlled by both of: a hosts_try_prdr option
2499 on the smtp transport, and the server advertisement.
2500 Default client logging of deliveries and rejections involving
2501 PRDR are flagged with the string "PRDR".
2503 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2504 fclose(). Diagnosis by Todd Lyons.
2506 PP/17 Update configure.default to handle IPv6 localhost better.
2507 Patch by Alain Williams (plus minor tweaks).
2510 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2511 This is now consistent with GnuTLS, and is now documented: the
2512 previous undocumented portable approach to treating the option as
2513 unset was to force an expansion failure. That still works, and
2514 an empty string is now equivalent.
2516 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2517 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2518 not performing validation itself.
2520 PP/20 Added force_command boolean option to pipe transport.
2521 Patch from Nick Koston, of cPanel Inc.
2523 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2526 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2528 PP/21 Fix eximon continuous updating with timestamped log-files.
2529 Broken in a format-string cleanup in 4.80, missed when I repaired the
2530 other false fix of the same issue.
2531 Report and fix from Heiko Schlichting.
2534 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2535 Report from Prashanth Katuri.
2537 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2538 It's SecureTransport, so affects any MacOS clients which use the
2539 system-integrated TLS libraries, including email clients.
2541 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2542 using a MIME ACL for non-SMTP local injection.
2543 Report and assistance in diagnosis by Warren Baker.
2545 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2547 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2549 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2550 last id that failed so it may be referenced in subsequent ACL's.
2552 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2555 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2556 ldap library initialization, allowing self-signed CA's to be
2557 used. Also properly sets require_cert option later in code by
2558 using NULL (global ldap config) instead of ldap handle (per
2559 session). Bug diagnosis and testing by alxgomz.
2561 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2562 the src/util/ subdirectory.
2564 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2565 renamed to Transport Post Delivery Action by Jeremy Harris, as
2568 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2569 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2570 redis_servers = needs to be configured which will be used by the redis
2571 lookup. Patch from Warren Baker, of The Packet Hub.
2573 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2575 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2576 hostname or reverse DNS when processing a host list. Used suggestions
2577 from multiple comments on this bug.
2579 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2581 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2582 Missed a few lines, added it to make the runtest require no keyboard
2585 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2586 contains upper case chars. Make router use caseful_local_part.
2588 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2589 support when GnuTLS has been built with p11-kit.
2595 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2597 This, or similar/improved, will also be change PP/11 of 4.82.
2603 PP/01 Handle short writes when writing local log-files.
2604 In practice, only affects FreeBSD (8 onwards).
2605 Bugzilla 1053, with thanks to Dmitry Isaikin.
2607 NM/01 Bugzilla 949 - Documentation tweak
2609 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2612 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2614 PP/02 Implemented gsasl authenticator.
2616 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2618 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2619 `pkg-config foo` for cflags/libs.
2621 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2622 with rest of GSASL and with heimdal_gssapi.
2624 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2625 `pkg-config foo` for cflags/libs for the TLS implementation.
2627 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2628 properties get this fed in as external SSF. A number of robustness
2629 and debugging improvements to the cyrus_sasl authenticator.
2631 PP/08 cyrus_sasl server now expands the server_realm option.
2633 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2634 Patch by Jeremy Harris.
2636 PP/10 Added dbmjz lookup type.
2638 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2640 PP/12 MAIL args handles TAB as well as SP, for better interop with
2641 non-compliant senders.
2642 Analysis and variant patch by Todd Lyons.
2644 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2645 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2646 Patch from Dirk Mueller <dmueller@suse.com>
2648 PP/13 tls_peerdn now print-escaped for spool files.
2649 Observed some $tls_peerdn in wild which contained \n, which resulted
2650 in spool file corruption.
2652 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2653 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2654 or write after TLS renegotiation, which otherwise led to messages
2657 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2658 as a tracking header (ie: a signed header comes before the signature).
2659 Patch from Wolfgang Breyha.
2661 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2662 comma-sep list; embedded commas doubled.
2664 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2666 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2668 Report and patch from Dmitry Banschikov.
2670 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2671 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2672 failures. We appear to now support TLS1.1+ with Exim.
2674 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2675 lets Exim select keys and certificates based upon TLS SNI from client.
2676 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2677 before an outbound SMTP session. New log_selector, +tls_sni.
2679 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2680 NULL dereference. Report and patch from Alun Jones.
2682 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2683 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2684 Not seeing resolver debug output on NetBSD, but suspect this is a
2685 resolver implementation change.
2687 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2688 Left warnings. Added "eximon gdb" invocation mode.
2690 PP/21 Defaulting "accept_8bitmime" to true, not false.
2692 PP/22 Added -bw for inetd wait mode support.
2694 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2695 locate the relevant includes and libraries. Made this the default.
2697 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2698 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2700 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2701 This may cause build issues on older platforms.
2703 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2704 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2705 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2706 Added SNI support via GnuTLS too.
2707 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2709 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2711 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2712 Applied second patch from Janne, implementing suggestion to default
2713 multiple-strings-in-record handling to match SPF spec.
2715 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2717 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2718 read-only, out of scope).
2719 Patch from Wolfgang Breyha, report from Stuart Northfield.
2721 PP/29 Fix three issues highlighted by clang analyser static analysis.
2722 Only crash-plausible issue would require the Cambridge-specific
2723 iplookup router and a misconfiguration.
2724 Report from Marcin Mirosław.
2726 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2728 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2729 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2730 As part of this, removing so much warning spew let me fix some minor
2731 real issues in debug logging.
2733 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2734 assignment on my part. Fixed.
2736 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2737 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2738 Janne Snabb (who went above and beyond: thank you).
2740 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2741 string otherwise requires a connection and a bunch more work and it's
2742 relatively easy to get wrong. Should also expose TLS library linkage
2745 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2746 64-bit ${eval} (JH/03).
2748 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2749 GNU libc to support some of the 64-bit stuff, should not lead to
2750 conflicts. Defined before os.h is pulled in, so if a given platform
2751 needs to override this, it can.
2753 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2754 protection layer was required, which is not implemented.
2755 Bugzilla 1254, patch from Wolfgang Breyha.
2757 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2758 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2759 tls_dhparam take prime identifiers. Also unbreak combination of
2760 OpenSSL+DH_params+TLSSNI.
2762 PP/39 Disable SSLv2 by default in OpenSSL support.
2768 PP/01 Solaris build fix for Oracle's LDAP libraries.
2769 Bugzilla 1109, patch from Stephen Usher.
2771 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2773 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2776 TF/02 Fix a couple more cases where we did not log the error message
2777 when unlink() failed. See also change 4.74-TF/03.
2779 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2780 lock up or crash if it happened to be inside a call to libc when it
2781 got a SIGUSR1 from exiwhat.
2783 The SIGUSR1 handler appends the current process status to the process
2784 log which is later printed by exiwhat. It used to use the general
2785 purpose logging code to do this, but several functions it calls are
2786 not safe for signals.
2788 The new output code in the SIGUSR1 handler is specific to the process
2789 log, and simple enough that it's easy to inspect for signal safety.
2790 Removing some special cases also simplifies the general logging code.
2791 Removing the spurious timestamps from the process log simplifies
2794 TF/04 Improved ratelimit ACL condition.
2796 The /noupdate option has been deprecated in favour of /readonly which
2797 has clearer semantics. The /leaky, /strict, and /readonly update modes
2798 are mutually exclusive. The update mode is no longer included in the
2799 database key; it just determines when the database is updated. (This
2800 means that when you upgrade Exim will forget old rate measurements.)
2802 Exim now checks that the per_* options are used with an update mode that
2803 makes sense for the current ACL. For example, when Exim is processing a
2804 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2805 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2806 must specify per_mail/readonly. If you omit the update mode it defaults to
2807 /leaky where that makes sense (as before) or /readonly where required.
2809 The /noupdate option is now undocumented but still supported for
2810 backwards compatibility. It is equivalent to /readonly except that in
2811 ACLs where /readonly is required you may specify /leaky/noupdate or
2812 /strict/noupdate which are treated the same as /readonly.
2814 A useful new feature is the /count= option. This is a generalization
2815 of the per_byte option, so that you can measure the throughput of other
2816 aggregate values. For example, the per_byte option is now equivalent
2817 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2819 The per_rcpt option has been generalized using the /count= mechanism
2820 (though it's more complicated than the per_byte equivalence). When it is
2821 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2822 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2823 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2824 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2825 Note that using per_rcpt with a non-readonly update mode in more than
2826 one ACL will cause the recipients to be double-counted. (The per_mail
2827 and per_byte options don't have this problem.)
2829 The handling of very low rates has changed slightly. If the computed rate
2830 is less than the event's count (usually one) then this event is the first
2831 after a long gap. In this case the rate is set to the same as this event's
2832 count, so that the first message of a spam run is counted properly.
2834 The major new feature is a mechanism for counting the rate of unique
2835 events. The new per_addr option counts the number of different
2836 recipients that someone has sent messages to in the last time period. It
2837 behaves like per_rcpt if all the recipient addresses are different, but
2838 duplicate recipient addresses do not increase the measured rate. Like
2839 the /count= option this is a general mechanism, so the per_addr option
2840 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2841 example, measure the rate that a client uses different sender addresses
2842 with the options per_mail/unique=$sender_address. There are further
2843 details in the main documentation.
2845 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2847 TF/06 Removed a few PCRE remnants.
2849 TF/07 Automatically extract Exim's version number from tags in the git
2850 repository when doing development or release builds.
2852 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2853 Bugzilla 879. Patch from Paul Fisher.
2855 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2856 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2859 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2861 PP/05 Variable $av_failed, true if the AV scanner deferred.
2862 Bugzilla 1078. Patch from John Horne.
2864 PP/06 Stop make process more reliably on build failure.
2865 Bugzilla 1087. Patch from Heiko Schlittermann.
2867 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2868 Bugzilla 1089. Patch from Heiko Schlittermann.
2870 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2871 Bugzilla 1131. Patch from Holger Weiß.
2873 PP/09 Handle IPv6 addresses with SPF.
2874 Bugzilla 860. Patch from Wolfgang Breyha.
2876 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2878 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2881 PP/11 match_* no longer expand right-hand-side by default.
2882 New compile-time build option, EXPAND_LISTMATCH_RHS.
2883 New expansion conditions, "inlist", "inlisti".
2885 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2887 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2889 PP/14 fix log_write() format string regression from TF/03.
2890 Bugzilla 1152. Patch from Dmitry Isaikin.
2896 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2898 PP/02 Harmonised TLS library version reporting; only show if debugging.
2899 Layout now matches that introduced for other libraries in 4.74 PP/03.
2901 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2903 PP/04 New "dns_use_edns0" global option.
2905 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2908 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2909 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2911 TK/01 Updated PolarSSL code to 0.14.2.
2912 Bugzilla 1097. Patch from Andreas Metzler.
2914 PP/07 Catch divide-by-zero in ${eval:...}.
2915 Fixes bugzilla 1102.
2917 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2920 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2921 format-string attack -- SECURITY: remote arbitrary code execution.
2923 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2924 time unintentionally subject to list matching rules, letting the header
2925 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2926 arbitrary string expansion). This allowed for information disclosure.
2928 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2929 INT_MIN/-1 -- value coerced to INT_MAX.
2935 NM/01 Workaround for PCRE version dependency in version reporting
2938 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2939 This fixes portability to compilers other than gcc, notably
2940 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2942 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2943 makefiles for portability to HP-UX and POSIX correctness.
2945 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2946 Also via indented variable definition in the Makefile.
2947 (Debugging by Oliver Heesakkers).
2949 PP/02 Restore caching of spamd results with expanded spamd_address.
2950 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2952 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2953 Improves build reliability. Fix from: Frank Elsner
2955 NM/02 Fix wide character breakage in the rfc2047 coding
2956 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2958 NM/03 Allow underscore in dnslist lookups
2959 Fixes bug 1026. Patch from Graeme Fowler
2961 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2962 Code patches from Adam Ciarcinski of NetBSD.
2964 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2967 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2968 is logged, to avoid truncation. Patch from John Horne.
2970 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2971 Patch from Jakob Hirsch.
2973 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2974 SQL string expansion failure details.
2975 Patch from Andrey Oktyabrski.
2977 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2978 Patch from Simon Arlott.
2980 PP/09 New lookups functionality failed to compile on old gcc which rejects
2981 extern declarations in function scope.
2982 Patch from Oliver Fleischmann
2984 PP/10 Use sig_atomic_t for flags set from signal handlers.
2985 Check getgroups() return and improve debugging.
2986 Fixed developed for diagnosis in bug 927 (which turned out to be
2989 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2990 Patch from Mark Zealey.
2992 PP/12 Bugzilla 1056: Improved spamd server selection.
2993 Patch from Mark Zealey.
2995 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2996 Based on patch from Heiko Schlittermann.
2998 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2999 Patch from Uwe Doering, sign-off by Michael Haardt.
3001 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3002 variable declaration deep within a block. Bug and patch from
3005 PP/15 lookups-Makefile IRIX compatibility coercion.
3007 PP/16 Make DISABLE_DKIM build knob functional.
3009 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3010 Patch by Simon Arlott
3012 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3013 variable argument macros. Our copy now differs from upstream.
3019 TF/01 Failure to get a lock on a hints database can have serious
3020 consequences so log it to the panic log.
3022 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3023 controlled using the smtp_confirmation log selector.
3025 TF/03 Include the error message when we fail to unlink a spool file.
3027 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3028 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3029 for maintaining out-of-tree patches for some time.
3031 PP/01 Bugzilla 139: Documentation and portability issues.
3032 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3033 Handle per-OS dynamic-module compilation flags.
3035 PP/02 Let /dev/null have normal permissions.
3036 The 4.73 fixes were a little too stringent and complained about the
3037 permissions on /dev/null. Exempt it from some checks.
3038 Reported by Andreas M. Kirchwitz.
3040 PP/03 Report version information for many libraries, including
3041 Exim version information for dynamically loaded libraries. Created
3042 version.h, now support a version extension string for distributors
3043 who patch heavily. Dynamic module ABI change.
3045 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3046 privilege escalation vulnerability whereby the Exim run-time user
3047 can cause root to append content of the attacker's choosing to
3050 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3053 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3054 If dropping privileges for untrusted macros, we disabled normal logging
3055 on the basis that it would fail; for the Exim run-time user, this is not
3056 the case, and it resulted in successful deliveries going unlogged.
3057 Fixed. Reported by Andreas Metzler.
3063 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3064 only prepend for the Resent-* case. Fixes regression introduced in
3065 Exim 4.70 by NM/22 for Bugzilla 607.
3067 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3068 increasing numbers of administrators be bitten by this.
3070 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3072 PP/03 Bugzilla 994: added openssl_options main configuration option.
3074 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3076 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3078 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3080 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3081 without a peer certificate, leading to a segfault because of an
3082 assumption that peers always have certificates. Be a little more
3083 paranoid. Problem reported by Martin Tscholak.
3085 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3086 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3087 NB: ClamAV planning to remove STREAM in "middle of 2010".
3088 CL also introduces -bmalware, various -d+acl logging additions and
3089 more caution in buffer sizes.
3091 PP/09 Implemented reverse_ip expansion operator.
3093 PP/10 Bugzilla 937: provide a "debug" ACL control.
3095 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3097 PP/12 Bugzilla 973: Implement --version.
3099 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3101 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3103 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3105 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3106 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3107 ignore trailing whitespace.
3109 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3111 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3114 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3115 Notification from Dr Andrew Aitchison.
3117 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3118 ExtendedDetectionInfo response format.
3119 Notification from John Horne.
3121 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3124 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3125 XSL and documented dependency on system catalogs, with examples of how
3128 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3131 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3132 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3133 configuration file which is writeable by the Exim user or group.
3135 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3136 of configuration files to cover files specified with the -C option if
3137 they are going to be used with root privileges, not just the default
3140 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3141 option (effectively making it always true).
3143 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3144 files to be used while preserving root privileges.
3146 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3147 that rogue child processes cannot use them.
3149 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3150 run-time user, instead of root.
3152 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3153 Exim run-time user without dropping privileges.
3155 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3156 result string, instead of calling string_vformat() twice with the same
3159 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3160 for other users. Others should always drop root privileges if they use
3161 -C on the command line, even for a whitelisted configure file.
3163 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3165 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3171 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3172 $data_path, and $header_path variables; fixed documentation bugs and
3175 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3176 exipick to access non-standard spools, including the "frozen" queue
3179 NM/01 Bugzilla 965: Support mysql stored procedures.
3180 Patch from Alain Williams
3182 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3184 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3185 Patch from Andreas Metzler
3187 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3188 Patch from Kirill Miazine
3190 NM/05 Bugzilla 671: Added umask to procmail example.
3192 JJ/03 installed exipick 20100323.0, fixing doc bug
3194 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3195 directory. Notification and patch from Dan Rosenberg.
3197 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3199 TK/02 Improve log output when DKIM signing operation fails.
3201 MH/01 Treat the transport option dkim_domain as a colon separated
3202 list, not as a single string, and sign the message with each element,
3203 omitting multiple occurences of the same signer.
3205 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3206 Bugzilla 985, 986. Patch by Simon Arlott
3208 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3209 Patch by Simon Arlott
3211 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3212 MBX locking. Notification from Dan Rosenberg.
3218 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3220 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3222 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3224 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3226 NM/04 Bugzilla 915: Flag broken perl installation during build.
3232 TK/01 Added patch by Johannes Berg that expands the main option
3233 "spamd_address" if it starts with a dollar sign.
3235 TK/02 Write list of recipients to X-Envelope-Sender header when building
3236 the mbox-format spool file for content scanning (suggested by Jakob
3239 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3240 (http://www.dcc-servers.net/) support via dccifd. Activated by
3241 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3243 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3244 by Mark Daniel Reidel <mr@df.eu>.
3246 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3247 When building exim an external PCRE library is now needed -
3248 PCRE is a system library on the majority of modern systems.
3249 See entry on PCRE_LIBS in EDITME file.
3251 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3252 conversation. Added nologin parameter to request.
3253 Patch contributed by Kirill Miazine.
3255 TF/01 Do not log submission mode rewrites if they do not change the address.
3257 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3259 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3260 log files in place. Contributed by Roberto Lima.
3262 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3264 TF/03 Bugzilla 615: When checking the local_parts router precondition
3265 after a local_part_suffix or local_part_prefix option, Exim now
3266 does not use the address's named list lookup cache, since this
3267 contains cached lookups for the whole local part.
3269 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3270 Robert Millan. Documentation is in experimental-spec.txt.
3272 TF/04 Bugzilla 668: Fix parallel build (make -j).
3274 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3276 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3277 Patch provided by Jan Srzednicki.
3279 TF/05 Leading white space used to be stripped from $spam_report which
3280 wrecked the formatting. Now it is preserved.
3282 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3283 that they are available at delivery time.
3285 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3287 TF/08 TLS error reporting now respects the incoming_interface and
3288 incoming_port log selectors.
3290 TF/09 Produce a more useful error message if an SMTP transport's hosts
3291 setting expands to an empty string.
3293 NM/06 Bugzilla 744: EXPN did not work under TLS.
3294 Patch provided by Phil Pennock.
3296 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3297 Patch provided by Richard Godbee.
3299 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3300 acl_smtp_notquit, added index entry.
3302 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3303 Patch provided by Eugene Bujak.
3305 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3306 accept(). Patch provided by Maxim Dounin.
3308 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3309 Patch provided by Phil Pennock.
3311 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3313 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3314 Patch provided by Brad "anomie" Jorsch.
3316 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3317 Patch provided by Dean Brooks.
3319 TK/05 Add native DKIM support (does not depend on external libraries).
3321 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3322 Patch provided by Graeme Fowler.
3324 NM/16 Bugzilla 851: Documentation example syntax fix.
3326 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3328 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3331 NM/19 Bugzilla 745: TLS version reporting.
3332 Patch provided by Phil Pennock.
3334 NM/20 Bugzilla 167: bool: condition support.
3335 Patch provided by Phil Pennock.
3337 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3338 clients. Patch provided by Phil Pennock.
3340 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3341 Patch provided by Brad "anomie" Jorsch.
3343 NM/23 Bugzilla 687: Fix misparses in eximstats.
3344 Patch provided by Heiko Schlittermann.
3346 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3347 Patch provided by Heiko Schlittermann.
3349 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3350 plus update to original patch.
3352 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3354 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3355 Patch provided by David Brownlee.
3357 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3359 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3361 NM/30 Bugzilla 888: TLS documentation bugfixes.
3363 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3365 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3366 Unlike the original bugzilla I have changed all shell scripts in src tree.
3368 NM/33 Bugzilla 898: Transport filter timeout fix.
3369 Patch by Todd Rinaldo.
3371 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3372 Patch by Serge Demonchaux.
3374 NM/35 Bugzilla 39: Base64 decode bug fixes.
3375 Patch by Jakob Hirsch.
3377 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3379 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3381 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3383 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3389 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3390 ALT-N's libdkim that I have put here:
3391 http://duncanthrax.net/exim-experimental/
3393 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3394 were called 'true' and it seems that C99 defines that as a reserved
3395 keyword to be used with 'bool' variable types. That means you could
3396 not include C99-style headers which use bools without triggering
3397 build errors in sieve.c.
3399 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3400 as mailq or other aliases. Changed the --help handling significantly
3401 to do whats expected. exim_usage() emits usage/help information.
3403 SC/01 Added the -bylocaldomain option to eximstats.
3405 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3407 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3409 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3415 PH/01 Another patch from the Sieve maintainer.
3417 PH/02 When an IPv6 address is converted to a string for single-key lookup
3418 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3419 dots are used instead of colons so that keys in lsearch files need not
3420 contain colons. This was done some time before quoting was made available
3421 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3422 (notated using the quote facility) so as to distinguish them from IPv4
3423 keys. This meant that lookups for IP addresses in host lists did not work
3424 for iplsearch lookups.
3426 This has been fixed by arranging for IPv6 addresses to be expressed with
3427 colons if the lookup type is iplsearch. This is not incompatible, because
3428 previously such lookups could never work.
3430 The situation is now rather anomalous, since one *can* have colons in
3431 ordinary lsearch keys. However, making the change in all cases is
3432 incompatible and would probably break a number of configurations.
3434 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3437 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3438 conversion specification without a maximum field width, thereby enabling
3439 a rogue spamd server to cause a buffer overflow. While nobody in their
3440 right mind would setup Exim to query an untrusted spamd server, an
3441 attacker that gains access to a server running spamd could potentially
3442 exploit this vulnerability to run arbitrary code as the Exim user.
3444 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3445 $primary_hostname instead of what libspf2 thinks the hosts name is.
3447 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3448 a directory entry by the name of the lookup key. Previously, if a
3449 symlink pointed to a non-existing file or a file in a directory that
3450 Exim lacked permissions to read, a lookup for a key matching that
3451 symlink would fail. Now it is enough that a matching directory entry
3452 exists, symlink or not. (Bugzilla 503.)
3454 PH/03 The body_linecount and body_zerocount variables are now exported in the
3457 PH/04 Added the $dnslist_matched variable.
3459 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3460 This means they are set thereafter only if the connection becomes
3463 PH/06 Added the client_condition to authenticators so that some can be skipped
3464 by clients under certain conditions.
3466 PH/07 The error message for a badly-placed control=no_multiline_responses left
3467 "_responses" off the end of the name.
3469 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3471 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3472 (without spaces) instead of just copying the configuration text.
3474 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3476 PH/11 Added $max_received_linelength.
3478 PH/12 Added +ignore_defer and +include_defer to host lists.
3480 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3481 way in which PCRE > 7.0 is built.
3483 PH/14 Implemented queue_only_load_latch.
3485 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3486 MAIL command. The effect was to mangle the value on 64-bit systems.
3488 PH/16 Another patch from the Sieve maintainer.
3490 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3492 PH/18 If a system quota error occurred while trying to create the file for
3493 a maildir delivery, the message "Mailbox is full" was not appended to the
3494 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3495 only to a quota excession during the actual writing of the file.
3497 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3498 characters?) which causes problems in log lines. The DN values are now
3499 passed through string_printing() before being added to log lines.
3501 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3502 and InterBase are left for another time.)
3504 PH/21 Added message_body_newlines option.
3506 PH/22 Guard against possible overflow in moan_check_errorcopy().
3508 PH/23 POSIX allows open() to be a macro; guard against that.
3510 PH/24 If the recipient of an error message contained an @ in the local part
3511 (suitably quoted, of course), incorrect values were put in $domain and
3512 $local_part during the evaluation of errors_copy.
3518 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3519 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3522 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3523 issue a MAIL command.
3525 PH/02 In an ACL statement such as
3527 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3529 if a client was not listed at all, or was listed with a value other than
3530 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3531 the condition was not true (as it should be), so access was not denied.
3532 The bug was that the ! inversion was incorrectly passed on to the second
3533 item. This has been fixed.
3535 PH/03 Added additional dnslists conditions == and =& which are different from
3536 = and & when the dns lookup returns more than one IP address.
3538 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3539 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3541 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3542 FSYNC, which compiles an option called disable_fsync that allows for
3543 bypassing fsync(). The documentation is heavily laced with warnings.
3545 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3547 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3548 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3549 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3550 including adding "make clean"; (3) Added -fPIC when compiling the test
3551 dynamically loaded module, to get rid of a warning.
3553 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3554 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3555 = 0s. The bug is otherwise harmless.
3557 PH/07 There was a bug in the dovecot authenticator such that the value of
3558 $auth1 could be overwritten, and so not correctly preserved, after a
3559 successful authentication. This usually meant that the value preserved by
3560 the server_setid option was incorrect.
3562 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3564 PH/09 Installed PCRE release 7.0.
3566 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3567 run for batched SMTP input. It is now run at the start of every message
3568 in the batch. While fixing this I discovered that the process information
3569 (output by running exiwhat) was not always getting set for -bs and -bS
3570 input. This is fixed, and it now also says "batched" for BSMTP.
3572 PH/11 Added control=no_pipelining.
3574 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3575 patch, slightly modified), and move the expansion of helo_data till after
3576 the connection is made in the smtp transport (so it can use these
3579 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3581 PH/14 Added log_selector = +pid.
3583 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3585 PH/16 Add ${if forany and ${if forall.
3587 PH/17 Added dsn_from option to vary the From: line in DSNs.
3589 PH/18 Flush SMTP output before performing a callout, unless control =
3590 no_callout_flush is set.
3592 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3593 was true (the default) a successful delivery failed to delete the retry
3594 item, thus causing premature timeout of the address. The bug is now
3597 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3599 PH/21 Long custom messages for fakedefer and fakereject are now split up
3600 into multiline responses in the same way that messages for "deny" and
3601 other ACL rejections are.
3603 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3604 with slight modification.
3606 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3607 draft, changing the syntax and factoring some duplicate code".
3609 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3610 for deliveries of the second and subsequent messages over the same SMTP
3613 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3614 ${reduce, with only minor "tidies".
3616 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3618 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3619 expansion side effects.
3621 PH/27 When a message times out after an over-quota error from an Exim-imposed
3622 quota, the bounce message says "mailbox is full". This message was not
3623 being given when it was a system quota that was exceeded. It now should
3626 MH/03 Made $recipients available in local_scan(). local_scan() already has
3627 better access to the recipient list through recipients_list[], but
3628 $recipients can be useful in postmaster-provided expansion strings.
3630 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3631 in the case of a MAIL command with additional options following the
3632 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3633 were accidentally chopped off.
3635 PH/29 SMTP synchronization checks are implemented when a command is read -
3636 there is a check that no more input is waiting when there shouldn't be
3637 any. However, for some commands, a delay in an ACL can mean that it is
3638 some time before the response is written. In this time, more input might
3639 arrive, invalidly. So now there are extra checks after an ACL has run for
3640 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3641 pipelining has not been advertised.
3643 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3645 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3646 returned in the SMTP response when smtp_return_error_details was false.
3647 This has been fixed.
3649 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3650 instead of the C I/O that was originally supplied, because problems were
3651 reported on Solaris.
3653 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3654 Exim which did not show up earlier: it was assuming that a call to
3655 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3656 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3657 was a macro that became an assignment, so it seemed to work. This has
3658 changed to a proper function call with a void return, hence the compile
3659 error. Exim's code has been fixed.
3661 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3664 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3666 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3668 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3669 because it turns out to be too big or there is a timeout) there is no
3670 "Completed" line in the log. When some messages of this type were
3671 selected by exigrep, they were listed as "not completed". Others were
3672 picked up by some special patterns. I have improved the selection
3673 criteria to be more general.
3675 PH/38 The host_find_failed option in the manualroute router can now be set
3676 to "ignore", to completely ignore a host whose IP address cannot be
3677 found. If all hosts are ignored, the behaviour is controlled by the new
3678 host_all_ignored option.
3680 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3681 homing or because of multiple MX records with /mx) generated more than
3682 one IP address, and the following item turned out to be the local host,
3683 all the secondary addresses of the first item were incorrectly removed
3684 from the list, along with the local host and any following hosts (which
3685 is what is supposed to happen).
3687 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3688 whoever called Exim into the -H file. In the case of the daemon it was
3689 behaving confusingly. When first started, it used values for whoever
3690 started the daemon, but after a SIGHUP it used the Exim user (because it
3691 calls itself on a restart). I have changed the code so that it now always
3694 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3695 message are rejected with the same error (e.g. no authentication or bad
3696 sender address), and a DATA command is nevertheless sent (as can happen
3697 with PIPELINING or a stupid MUA), the error message that was given to the
3698 RCPT commands is included in the rejection of the DATA command. This is
3699 intended to be helpful for MUAs that show only the final error to their
3702 PH/42 Another patch from the Sieve maintainer.
3704 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3705 Eximstats - Fixed some broken HTML links and added missing column headers
3707 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3708 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3710 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3712 PH/43 Yet another patch from the Sieve maintainer.
3714 PH/44 I found a way to check for a TCP/IP connection going away before sending
3715 the response to the final '.' that terminates a message, but only in the
3716 case where the client has not sent further data following the '.'
3717 (unfortunately, this is allowed). However, in many cases there won't be
3718 any further data because there won't be any more messages to send. A call
3719 to select() can be used: if it shows that the input is "ready", there is
3720 either input waiting, or the socket has been closed. An attempt to read
3721 the next input character can distinguish the two cases. Previously, Exim
3722 would have sent an OK response which the client would never have see.
3723 This could lead to message repetition. This fix should cure that, at
3724 least in a lot of common cases.
3726 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3727 advertised in response to EHLO.
3733 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3734 fixed by 4.65/MH/01 (is this a record?) are fixed:
3736 (i) An empty string was always treated as zero by the numeric comparison
3737 operators. This behaviour has been restored.
3739 (ii) It is documented that the numeric comparison operators always treat
3740 their arguments as decimal numbers. This was broken in that numbers
3741 starting with 0 were being interpreted as octal.
3743 While fixing these problems I realized that there was another issue that
3744 hadn't been noticed. Values of message_size_limit (both the global option
3745 and the transport option) were treated as octal if they started with 0.
3746 The documentation was vague. These values are now always treated as
3747 decimal, and I will make that clear in the documentation.
3753 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3754 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3757 MH/01 Don't check that the operands of numeric comparison operators are
3758 integers when their expansion is in "skipping" mode (fixes bug
3759 introduced by 4.64-PH/07).
3761 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3762 child addresses, Exim now panics and dies. Previously, because the count
3763 is held in a short int, deliveries were likely to be lost. As such a
3764 large number of recipients for a single message is ridiculous
3765 (performance will be very, very poor), I have chosen to impose a limit
3766 rather than extend the field.
3772 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3773 leftover -K file (the existence of which was triggered by #402).
3774 While we were at it, introduced process PID as part of the -K
3775 filename. This should rule out race conditions when creating
3778 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3779 processing considerably. Previous code took too long for large mails,
3780 triggering a timeout which in turn triggers #401.
3782 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3783 in the DK code in transports.c. sendfile() is not really portable,
3784 hence the _LINUX specificness.
3786 TF/01 In the add_headers option to the mail command in an Exim filter,
3787 there was a bug that Exim would claim a syntax error in any
3788 header after the first one which had an odd number of characters
3791 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3792 callout verification, Exim cached a "reject" for the entire domain. This
3793 is correct for most verifications, but it is not correct for a recipient
3794 verification with use_sender or use_postmaster set, because in that case
3795 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3796 case of MAIL FROM:<> rejection from other early rejections (e.g.
3797 rejection of HELO). When verifying a recipient using a non-null MAIL
3798 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3799 Whatever the result of the callout, the value of the domain cache is
3800 left unchanged (for any other kind of callout, getting as far as trying
3801 RCPT means that the domain itself is ok).
3803 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3806 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3807 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3808 session, the daemon ignores ECONNECT errors and logs others; it now
3809 ignores EPIPE as well.
3811 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3812 (quoted-printable decoding).
3814 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3815 later the small subsequent patch to fix an introduced bug.
3817 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3819 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3821 PH/08 An error is now given if message_size_limit is specified negative.
3823 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3824 to be given (somewhat) arbitrary names.
3826 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3829 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3830 miscellaneous code fixes
3832 PH/10 Added the log_reject_target ACL modifier to specify where to log
3835 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3836 hostname. This is wrong, because it relates to the incoming message (and
3837 probably the interface on which it is arriving) and not to the outgoing
3838 callout (which could be using a different interface). This has been
3839 changed to use the value of the helo_data option from the smtp transport
3840 instead - this is what is used when a message is actually being sent. If
3841 there is no remote transport (possible with a router that sets up host
3842 addresses), $smtp_active_hostname is used.
3844 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3845 tweaks were necessary in order to get it to work (see also 21 below):
3846 (a) The code assumed that strncpy() returns a negative number on buffer
3847 overflow, which isn't the case. Replaced with Exim's string_format()
3849 (b) There were several signed/unsigned issues. I just did the minimum
3850 hacking in of casts. There is scope for a larger refactoring.
3851 (c) The code used strcasecmp() which is not a standard C function.
3852 Replaced with Exim's strcmpic() function.
3853 (d) The code set only $1; it now sets $auth1 as well.
3854 (e) A simple test gave the error "authentication client didn't specify
3855 service in request". It would seem that Dovecot has changed its
3856 interface. Fortunately there's a specification; I followed it and
3857 changed what the client sends and it appears to be working now.
3859 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3862 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3863 address A is aliased to B and C, where B exists and C does not. Without
3864 -v the output is "A verified" because verification stops after a
3865 successful redirection if more than one address is generated. However,
3866 with -v the child addresses are also verified. Exim was outputting "A
3867 failed to verify" and then showing the successful verification for C,
3868 with its parentage. It now outputs "B failed to verify", showing B's
3869 parentage before showing the successful verification of C.
3871 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3872 look up a TXT record in a specific list after matching in a combined
3875 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3876 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3877 they consult the DNS. I had assumed they would set it the way they
3878 wanted; and indeed my experiments on Linux seem to show that in some
3879 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3880 To be on the safe side, however, I have now made the interface to
3881 host_find_byname() similar to host_find_bydns(), with an argument
3882 containing the DNS resolver options. The host_find_byname() function now
3883 sets these options at its start, just as host_find_bydns() does. The smtp
3884 transport options dns_qualify_single and dns_search_parents are passed to
3885 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3886 of host_find_byname() use the default settings of RES_DEFNAMES
3887 (qualify_single) but not RES_DNSRCH (search_parents).
3889 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3890 spool_read_header() do less string testing, by means of a preliminary
3891 switch on the second character of optional "-foo" lines. (This is
3892 overdue, caused by the large number of possibilities that now exist.
3893 Originally there were few.) While I was there, I also converted the
3894 str(n)cmp tests so they don't re-test the leading "-" and the first
3895 character, in the hope this might squeeze out yet more improvement.
3897 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3898 flag allowing group syntax was set by the header_syntax check but not
3899 turned off, possible causing trouble later; (2) The flag was not being
3900 set at all for the header_verify test, causing "group"-style headers to
3901 be rejected. I have now set it in this case, and also caused header_
3902 verify to ignore an empty address taken from a group. While doing this, I
3903 came across some other cases where the code for allowing group syntax
3904 while scanning a header line wasn't quite right (mostly, not resetting
3905 the flag correctly in the right place). These bugs could have caused
3906 trouble for malformed header lines. I hope it is now all correct.
3908 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3909 with the "reply" argument non-NULL. The code, however (which originally
3910 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3911 but it didn't always do it. This confused somebody who was copying the
3912 code for some other use. I have removed all the tests.
3914 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3915 feature that was used to support insecure browsers during the U.S. crypto
3916 embargo. It requires special client support, and Exim is probably the
3917 only MTA that supported it -- and would never use it because real RSA is
3918 always available. This code has been removed, because it had the bad
3919 effect of slowing Exim down by computing (never used) parameters for the
3920 RSA_EXPORT functionality.
3922 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3923 authenticator to fail if there's a tab character in the incoming data
3924 (there should never be unless someone is messing about, as it's supposed
3925 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3926 if the connection is using TLS or if the remote IP is the same as the
3927 local IP, and the "valid-client-cert option" if a client certificate has
3930 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3931 authenticators. This can be used for authorization after authentication
3932 succeeds. (In the case of plaintext, it servers for both authentication
3935 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3936 if any retry times were supplied.
3938 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3939 connection, where there is no client IP address to check. In this
3940 situation, the verify now always succeeds.
3942 PH/25 Applied John Jetmore's -Mset patch.
3944 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3946 PH/27 In a string expansion for a processed (not raw) header when multiple
3947 headers of the same name were present, leading whitespace was being
3948 removed from all of them, but trailing whitespace was being removed only
3949 from the last one. Now trailing whitespace is removed from each header
3950 before concatenation. Completely empty headers in a concatenation (as
3951 before) are ignored.
3953 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3954 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3956 PH/29 [Removed. This was a change that I later backed out, and forgot to
3957 correct the ChangeLog entry (that I had efficiently created) before
3958 committing the later change.]
3960 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3961 address errors (4xx response to RCPT) over the same connection as other
3962 messages routed to the same hosts. Such deliveries are always "forced",
3963 so retry times are not inspected. This resulted in far too many retries
3964 for the affected addresses. The effect occurred only when there were more
3965 hosts than the hosts_max_try setting in the smtp transport when it had
3966 the 4xx errors. Those hosts that it had tried were not added to the list
3967 of hosts for which the message was waiting, so if all were tried, there
3968 was no problem. Two fixes have been applied:
3970 (i) If there are any address or message errors in an SMTP delivery, none
3971 of the hosts (tried or untried) are now added to the list of hosts
3972 for which the message is waiting, so the message should not be a
3973 candidate for sending over the same connection that was used for a
3974 successful delivery of some other message. This seems entirely
3975 reasonable: after all the message is NOT "waiting for some host".
3976 This is so "obvious" that I'm not sure why it wasn't done
3977 previously. Hope I haven't missed anything, but it can't do any
3978 harm, as the worst effect is to miss an optimization.
3980 (ii) If, despite (i), such a delivery is accidentally attempted, the
3981 routing retry time is respected, so at least it doesn't keep
3982 hammering the server.
3984 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3985 in ${readsocket because some servers need this prod.
3987 PH/32 Added some extra debug output when updating a wait-xxx database.
3989 PH/33 The hint "could be header name not terminated by colon", which has been
3990 given for certain expansion errors for a long time, was not being given
3991 for the ${if def:h_colon_omitted{... case.
3993 PH/34 The spec says: "With one important exception, whenever a domain list is
3994 being scanned, $domain contains the subject domain." There was at least
3995 one case where this was not true.
3997 PH/35 The error "getsockname() failed: connection reset by peer" was being
3998 written to the panic log as well as the main log, but it isn't really
3999 panic-worthy as it just means the connection died rather early on. I have
4000 removed the panic log writing for the ECONNRESET error when getsockname()
4003 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4004 runs only) independently of the message's sender address. This meant
4005 that, if the 4xx error was in fact related to the sender, a different
4006 message to the same recipient with a different sender could confuse
4007 things. In particular, this can happen when sending to a greylisting
4008 server, but other circumstances could also provoke similar problems.
4009 I have changed the default so that the retry time for these errors is now
4010 based a combination of the sender and recipient addresses. This change
4011 can be overridden by setting address_retry_include_sender=false in the
4014 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4015 remote server are returned as part of bounce messages. This was not
4016 happening for LMTP over a pipe (the lmtp transport), but now it is the
4017 same for both kinds of LMTP.
4019 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4020 in header lines that were in fact CNAMEs. This is no longer the case.
4022 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4023 and queue runs started by the daemon processed all messages. This has
4024 been fixed so that -R and -S can now usefully be given with -q<time>.
4026 PH/40 Import PCRE release 6.7 (fixes some bugs).
4028 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4030 PH/42 Give an error if -q is specified more than once.
4032 PH/43 Renamed the variables $interface_address and $interface_port as
4033 $received_ip_address and $received_port, to make it clear that these
4034 values apply to message reception, and not to the outgoing interface when
4035 a message is delivered. (The old names remain recognized, of course.)
4037 PH/44 There was no timeout on the connect() call when using a Unix domain
4038 socket in the ${readsocket expansion. There now is.
4040 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4041 be meaningful with "accept".
4043 SC/01 Eximstats V1.43
4044 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4046 SC/02 Eximstats V1.44
4047 Use a glob alias rather than an array ref in the generated
4048 parser. This improves both readability and performance.
4050 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4051 Collect SpamAssassin and rejection statistics.
4052 Don't display local sender or destination tables unless
4053 there is data to show.
4054 Added average volumes into the top table text output.
4056 SC/04 Eximstats V1.46
4057 Collect data on the number of addresses (recipients)
4058 as well as the number of messages.
4060 SC/05 Eximstats V1.47
4061 Added 'Message too big' to the list of mail rejection
4062 reasons (thanks to Marco Gaiarin).
4064 SC/06 Eximstats V1.48
4065 Mainlog lines which have GMT offsets and are too short to
4066 have a flag are now skipped.
4068 SC/07 Eximstats V1.49 (Alain Williams)
4069 Added the -emptyok flag.
4071 SC/08 Eximstats V1.50
4072 Fixes for obtaining the IP address from reject messages.
4074 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4075 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4076 whitespace changes from 4.64-PH/27
4078 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4081 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4082 are found, allow negative numbers in numeric criteria)
4084 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4086 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4087 to match changes made in 4.64-PH/43
4089 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4091 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4092 "domain" is actually an IP address, and force a failure. This locks out
4093 those revolvers/nameservers that support "A-for-A" lookups, in
4094 contravention of the specifications.
4096 PH/48 When a host name was looked up from an IP address, and the subsequent
4097 forward lookup of the name timed out, the host name was left in
4098 $sender_host_name, contrary to the specification.
4100 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4101 restricted to single-key lookups, Exim was not diagnosing an error if
4102 * or *@ was used with a query-style lookup.
4104 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4106 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4107 long ago, but noone interested enough thought of it. Let's just say that
4108 the "1.1" means that there are some new functions that weren't there at
4109 some point in the past.
4111 PH/51 Error processing for expansion failure of helo_data from an smtp
4112 transport during callout processing was broken.
4114 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4115 tested/used via the -bh/-bhc/-bs options.
4117 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4118 bug, fixed in subsequent PCRE releases).
4120 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4121 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4123 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4129 SC/01 Use a glob alias rather than an array ref in eximstats generated
4130 parser. This improves both readability and performance.
4132 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4133 Don't display local sender or destination tables in eximstats unless
4134 there is data to show.
4135 Added average volumes into the eximstats top table text output.
4137 SC/03 Collect data on the number of addresses (recipients) as well
4138 as the number of messages in eximstats.
4140 TF/01 Correct an error in the documentation for the redirect router. Exim
4141 does (usually) call initgroups() when daemonizing.
4143 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4144 with consistent privilege compared to when running as a daemon.
4146 TF/03 Note in the spec that $authenticated_id is not set for local
4147 submissions from trusted users.
4149 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4150 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4152 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4153 by adding some example configuration directives to the default
4154 configuration file. A little bit of work is required to uncomment the
4155 directives and define how usernames and passwords are checked, but
4156 there is now a framework to start from.
4158 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4159 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4160 without this. I don't know how relevant this is to other LDAP libraries.
4162 PH/02 Add the verb name to the "unknown ACL verb" error.
4164 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4166 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4168 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4169 directory not expanded when it should be if an expanded home directory
4170 was set for the address (which is overridden by the transport).
4172 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4175 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4176 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4177 because it is too late at that time, and has no effect.
4179 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4180 security issue with \' (bugzilla #107). I could not use the
4181 PQescapeStringConn() function, because it needs a PGconn value as one of
4184 PH/08 When testing addresses using -bt, indicate those final addresses that
4185 are duplicates that would not cause an additional delivery. At least one
4186 person was confused, thinking that -bt output corresponded to deliveries.
4187 (Suppressing duplicates isn't a good idea as you lose the information
4188 about possibly different redirections that led to the duplicates.)
4190 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4191 systems where poll() doesn't work, in particular OS X.
4193 PH/10 Added more information to debugging output for retry time not reached.
4195 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4196 operations in malware.c.
4198 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4201 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4202 syslog_duplication set false, log lines that would normally be written
4203 both the the main log and to the reject log were not written to syslog at
4206 PH/14 In the default configuration, change the use of "message" in ACL warn
4207 statements to "add_header".
4209 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4210 not followed by a command (e.g. "seen endif").
4212 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4213 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4216 PH/17 Added extra conditions to the default value of delay_warning_condition
4220 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4221 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4222 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4225 The Auto-Submitted: and various List- headers are standardised, whereas I
4226 don't think Precedence: ever was.
4228 PH/18 Refactored debugging code in route_finduser() to show more information,
4229 in particular, the error code if getpwnam() issues one.
4231 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4232 This is apparently needed in addition to the PH/07 change above to avoid
4233 any possible encoding problems.
4235 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4236 but not after initializing Perl.
4238 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4239 output them only if debugging. By default they are written stderr,
4240 apparently, which is not desirable.
4242 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4245 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4248 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4250 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4251 authenticated or an ident call has been made. Suppress the default
4252 values for $authenticated_id and $authenticated_sender (but permit -oMai
4253 and -oMas) when testing with -bh.
4255 PH/24 Re-jigged the order of the tests in the default configuration so that the
4256 tests for valid domains and recipients precede the DNS black list and CSA
4257 tests, on the grounds that those ones are more expensive.
4259 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4260 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4261 This bug exists in every version of Exim that I still have, right back to
4264 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4265 However, an attempt to turn on case-sensitivity in a regex key by
4266 including (?-i) didn't work because the subject string was already
4267 lowercased, and the effects were non-intuitive. It turns out that a
4268 one-line patch can be used to allow (?-i) to work as expected.
4274 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4275 other effects) broke the use of negated acl sub-conditions.
4277 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4280 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4281 "Deny" causes Exim to reject the incoming connection with a 554 error.
4282 Unfortunately, if there is a major crisis, such as a disk failure,
4283 tcp-wrappers gives "deny", whereas what one would like would be some
4284 kind of temporary error. A kludge has been added to help with this.
4285 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4286 554 error is used if errno is still zero or contains ENOENT (which occurs
4287 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4290 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4292 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4293 errors. Otherwise a message that provokes a temporary error (when other
4294 messages do not) can cause a whole host to time out.
4296 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4297 addresses were routed directly to files or pipes from a redirect router.
4298 File deliveries just didn't batch; pipe deliveries might have suffered
4301 PH/06 A failure to get a lock for a hints database would erroneously always say
4302 "Failed to get write lock", even when it was really a read lock.
4304 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4305 of 0600. This has been changed to use the value of the lockfile_mode
4306 option (which defaults to 0600).
4308 PH/08 Applied small patch from the Sieve maintainer.
4310 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4311 folder from quota calculations, a direct delivery into this folder messed
4312 up the contents of the maildirsize file. This was because the regex was
4313 used only to exclude .Trash (or whatever) when the size of the mailbox
4314 was calculated. There was no check that a delivery was happening into an
4315 excluded directory. This bug has been fixed by ignoring all quota
4316 processing for deliveries into excluded directories.
4318 PH/10 Added the maildirfolder_create_regex option to appendfile.
4324 PH/01 The code for finding all the local interface addresses on a FreeBSD
4325 system running IPv6 was broken. This may well have applied to all BSD
4326 systems, as well as to others that have similar system calls. The broken
4327 code found IPv4 interfaces correctly, but gave incorrect values for the
4328 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4329 that it would not match correctly against @[] and not recognize the IPv6
4332 PH/02 The ipliteral router was not recognizing addresses of the form user@
4333 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4335 PH/03 Added disable_ipv6.
4337 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4338 decoded form, because it is most often used to construct To: headers
4339 lines in autoreplies, and the decoded form may well be syntactically
4340 invalid. However, $reply_address has leading white space removed, and all
4341 newlines turned into spaces so that the autoreply transport does not
4344 PH/05 If group was specified without a user on a router, and no group or user
4345 was specified on a transport, the group from the router was ignored.
4347 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4348 for visible compile-time settings that can be used to change these
4349 numbers, for those that want even more. Backwards compatibility with old
4350 spool files has been maintained. However, going back to a previous Exim
4351 release will lost any variables that are in spool files.
4353 PH/07 Two small changes when running in the test harness: increase delay when
4354 passing a TCP/IP connection to a new process, in case the original
4355 process has to generate a bounce, and remove special handling of
4356 127.0.0.2 (sic), which is no longer necessary.
4358 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4359 be the same on different OS.
4361 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4364 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4365 whether --show-vars was specified or not
4367 JJ/02 exipick: Added support for new ACL variable spool format introduced
4370 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4371 syntactically invalid From: or Reply-to: line, and a filter used this to
4372 generate an autoreply, and therefore failed to obtain an address for the
4373 autoreply, Exim could try to deliver to a non-existent relative file
4374 name, causing unrelated and misleading errors. What now happens is that
4375 it logs this as a hard delivery error, but does not attempt to create a
4378 PH/11 The exinext utility has a -C option for testing purposes, but although
4379 the given file was scanned by exinext itself; it wasn't being passed on
4380 when Exim was called.
4382 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4383 an end-of-file indication when reading a command response.
4385 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4386 compiled. In many other places in Exim, IPv6 addresses are always
4387 recognized, so I have changed this. It also means that IPv4 domain
4388 literals of the form [IPV4:n.n.n.n] are now always recognized.
4390 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4391 used if the router is not running as root, for example, when verifying at
4392 ACL time, or when using -bh. The debugging output from this situation was
4393 non-existent - all you got was a failure to exec. I have made two
4396 (a) Failures to set uid/gid, the current directory, or a process leader
4397 in a subprocess such as that created by queryprogram now generate
4398 suitable debugging output when -d is set.
4400 (b) The queryprogram router detects when it is not running as root,
4401 outputs suitable debugging information if -d is set, and then runs
4402 the subprocess without attempting to change uid/gid.
4404 PH/15 Minor change to Makefile for building test_host (undocumented testing
4407 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4408 additional section of a DNS packet that returns MX or SRV records.
4409 Instead, it always explicitly searches for A/AAAA records. This avoids
4410 major problems that occur when a DNS server includes only records of one
4411 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4412 fixed another bug: if SRV records were looked up and the corresponding
4413 address records were *not* found in the additional section, the port
4414 values from the SRV records were lost.
4416 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4417 using the correct key (the original address) when searching the retry
4418 rules in order to find which one to use for generating the retry hint.
4420 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4421 adding the default one. Similarly, if it contains a Reply-To: header, the
4422 errors_reply_to option, if set, is not used.
4424 PH/19 When calculating a retry time, Exim used to measure the "time since
4425 failure" by looking at the "first failed" field in the retry record. Now
4426 it does not use this if it is later than than the arrival time of the
4427 message. Instead it uses the arrival time. This makes for better
4428 behaviour in cases where some deliveries succeed, thus re-setting the
4429 "first failed" field. An example is a quota failure for a huge message
4430 when small messages continue to be delivered. Without this change, the
4431 "time since failure" will always be short, possible causing more frequent
4432 delivery attempts for the huge message than are intended.
4433 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4435 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4436 $1, $2, $3) because the numerical variables can be reset during some
4437 expansion items (e.g. "match"), thereby losing the authentication data.
4439 PH/21 Make -bV show the size of off_t variables so that the test suite can
4440 decide whether to run tests for quotas > 2G.
4442 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4443 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4444 filecount value is greater than 2G or if a quota value is greater than 2G
4445 on a system where the size of off_t is not greater than 4, a panic error
4448 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4449 never match. The debug and -bh output now contains an explicit error
4450 message indicating a malformed IPv4 address or mask.
4452 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4453 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4454 PH/23 above applies.
4456 PH/25 Do not write to syslog when running in the test harness. The only
4457 occasion when this arises is a failure to open the main or panic logs
4458 (for which there is an explicit test).
4460 PH/26 Added the /no_tell option to "control=freeze".
4462 PH/27 If a host name lookup failed very early in a connection, for example, if
4463 the IP address matched host_lookup and the reverse lookup yielded a name
4464 that did not have a forward lookup, an error message of the form "no IP
4465 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4466 could be logged. Now it outputs the IP address instead of "NULL".
4468 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4469 allows the sender and the authenticated sender to be set when
4470 submitting a message from within Exim. Since child_open_exim() is
4471 documented for local_scan(), the new function should be too.
4473 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4474 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4475 results in an empty string is now treated as unset.
4477 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4479 PH/31 Added sender_verify_fail logging option.
4481 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4482 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4483 tidied the source and removed it altogether.
4485 PH/33 When a queue run was abandoned because the load average was too high, a
4486 log line was always written; now it is written only if the queue_run log
4487 selector is set. In addition, the log line for abandonment now contains
4488 information about the queue run such as the pid. This is always present
4489 in "start" and "stop" lines but was omitted from the "abandon" line.
4491 PH/34 Omit spaces between a header name and the colon in the error message that
4492 is given when verify = headers_syntax fails (if there are lots of them,
4493 the message gets confusing).
4495 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4496 names, as there are now some PTR records that contain slashes. This check
4497 is only to protect against broken name servers that fall over on strange
4498 characters, so the fact that it applies to all lookups doesn't matter.
4500 PH/36 Now that the new test suite is complete, we can remove some of the
4501 special code in Exim that was needed for the old test suite. For example,
4502 sorting DNS records because real resolvers return them in an arbitrary
4503 order. The new test suite's fake resolver always returns records in the
4506 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4507 bounces) except when queue_only is set, to avoid logging races between
4508 the different processes.
4510 PH/38 Panic-die if .include specifies a non-absolute path.
4512 PH/39 A tweak to the "H" retry rule from its user.
4514 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4515 a label. They prevented compilation on older perls.
4517 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4518 a warning to be raised on newish perls.
4520 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4521 on queue. Changes to match documented behaviour of showing count of
4522 messages matching specified criteria.
4524 PH/40 Changed the default ident timeout from 30s to 5s.
4526 PH/41 Added support for the use of login_cap features, on those BSD systems
4527 that have them, for controlling the resources used by pipe deliveries.
4529 PH/42 The content-scanning code uses fopen() to create files in which to put
4530 message data. Previously it was not paying any attention to the mode of
4531 the files. Exim runs with umask(0) because the rest of the code creates
4532 files with open(), and sets the required mode explicitly. Thus, these
4533 files were ending up world-writeable. This was not a big issue, because,
4534 being within the spool directory, they were not world-accessible. I have
4535 created a function called modefopen, which takes an additional mode
4536 argument. It sets umask(777), creates the file, chmods it to the required
4537 mode, then resets the umask. All the relevant calls to fopen() in the
4538 content scanning code have been changed to use this function.
4540 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4541 to 24 hours. This avoids potential overflow problems when processing G
4542 and H retry rules. I suspect nobody ever tinkers with this value.
4544 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4546 PH/45 When the plaintext authenticator is running as a client, the server's
4547 challenges are checked to ensure they are valid base64 strings. By
4548 default, the authentication attempt is cancelled if an invalid string is
4549 received. Setting client_ignore_invalid_base64 true ignores these errors.
4550 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4551 they are received. Thus, the responses can be made to depend on the
4552 challenges. If an invalid string is ignored, an empty string is placed in
4555 PH/46 Messages that are created by the autoreply transport now contains a
4556 References: header, in accordance with RFCs 2822 and 3834.
4558 PH/47 Added authenticated_sender_force to the smtp transport.
4560 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4562 PH/49 Installed latest patch from the Sieve maintainer.
4564 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4565 was also set, the appendfile transport was unnecessarily scanning a
4566 directory of message files (e.g. for maildir delivery) to find the count
4567 of files (along with the size), even though it did not need this
4568 information. It now does the scan only if it needs to find either the
4569 size of the count of files.
4571 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4573 PH/52 Two bugs concerned with error handling when the smtp transport is
4576 (i) Exim was not creating retry information for temporary errors given
4577 for individual recipients after the DATA command when the smtp transport
4578 was used in LMTP mode. This meant that they could be retried too
4579 frequently, and not timed out correctly.
4581 (ii) Exim was setting the flag that allows error details to be returned
4582 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4583 recipients that were returned after the DATA command.
4585 PH/53 This is related to PH/52, but is more general: for any failing address,
4586 when detailed error information was permitted to be returned to the
4587 sender, but the error was temporary, then after the final timeout, only
4588 "retry timeout exceeded" was returned. Now it returns the full error as
4589 well as "retry timeout exceeded".
4591 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4592 do this, and (what is worse) MTAs that accept it.
4594 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4595 will now be deprecated.
4597 PH/56 New os.c-cygwin from the Cygwin maintainer.
4599 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4600 formats (previously only available in exim formats via -bpr, -bpru,
4601 and -bpra. Now also available in native and exiqgrep formats)
4603 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4604 with very large, slow to parse queues
4606 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4608 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4610 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4611 responses to authentication challenges, though it was showing the
4612 challenges; (ii) I've removed the CR characters from the debug output for
4615 PH/58 Allow for the insertion of a newline as well as a space when a string
4616 is turned into more than one encoded-word during RFC 2047 encoding. The
4617 Sieve code now uses this.
4619 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4620 data_4xx, lost_connection, tls_required.
4622 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4623 message was being sent as an SMTP response.
4625 PH/61 Add -l and -k options to exicyclog.
4627 PH/62 When verifying, if an address was redirected to one new address, so that
4628 verification continued, and the new address failed or deferred after
4629 having set something in $address_data, the value of $address_data was not
4630 passed back to the ACL. This was different to the case when no
4631 redirection occurred. The value is now passed back in both cases.
4633 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4634 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4635 use login_cap.h, so on its own it isn't the distinguishing feature. The
4636 new name refers directly to the setclassresources() function.
4638 PH/65 Added configuration files for NetBSD3.
4640 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4642 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4643 is preferred over IPv4.
4645 PH/68 The bounce_return_message and bounce_return_body options were not being
4646 honoured for bounces generated during the reception of non-SMTP messages.
4647 In particular, this applied to messages rejected by the ACL. This bug has
4648 been fixed. However, if bounce_return_message is true and bounce_return_
4649 body is false, the headers that are returned for a non-SMTP message
4650 include only those that have been read before the error was detected.
4651 (In the case of an ACL rejection, they have all been read.)
4653 PH/69 The HTML version of the specification is now built in a directory called
4654 spec_html instead of spec.html, because the latter looks like a path with
4655 a MIME-type, and this confuses some software.
4657 PH/70 Catch two compiler warnings in sieve.c.
4659 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4660 function verify_get_ident() calls ip_connect() to connect a socket, but
4661 if the "connect()" function timed out, ip_connect() used to close the
4662 socket. However, verify_get_ident() also closes the socket later, and in
4663 between Exim writes to the log, which may get opened at this point. When
4664 the socket was closed in ip_connect(), the log could get the same file
4665 descriptor number as the socket. This naturally causes chaos. The fix is
4666 not to close the socket in ip_connect(); the socket should be closed by
4667 the function that creates it. There was only one place in the code where
4668 this was missing, in the iplookup router, which I don't think anybody now
4669 uses, but I've fixed it anyway.
4671 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4672 well as to direct DNS lookups. Otherwise the handling of names in host
4673 lists is inconsistent and therefore confusing.
4679 PH/01 Two changes to the default runtime configuration:
4681 (1) Move the checks for relay_from_hosts and authenticated clients from
4682 after to before the (commented out) DNS black list checks.
4684 (2) Add control=submission to the relay_from_hosts and authenticated
4685 clients checks, on the grounds that messages accepted by these
4686 statements are most likely to be submissions.
4688 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4690 (1) Generate an error if the third argument for the ${prvs expansion is
4693 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4696 (3) Reset the variables that are obtained from the first argument of
4697 ${prvscheck and used in the second argument before leaving the code,
4698 because their memory is reclaimed, so using them afterwards may do
4701 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4702 one (it's much easier than Tom thought :-).
4704 (5) Because of (4), we can now allow for the use of $prvscheck_result
4705 inside the third argument.
4707 PH/03 For some reason, the default setting of PATH when running a command from
4708 a pipe transport was just "/usr/bin". I have changed it to
4711 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4712 anything to be listed in the output from -bV.
4714 PH/05 When a filter generated an autoreply, the entire To: header line was
4715 quoted in the delivery log line, like this:
4717 => >A.N.Other <ano@some.domain> <original@ddress> ...
4719 This has been changed so that it extracts the operative address. There
4720 may be more than one such address. If so, they are comma-separated, like
4723 => >ano@some.domain,ona@other.domain <original@ddress> ...
4725 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4726 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4727 not being looked up in the rDNS to get a host name, Exim was showing the
4728 IP address twice in Received: lines, even though the IP addresses were
4729 identical. For example:
4731 Received: from [1.2.3.4] (helo=[1.2.3.4])
4733 However, if the real host name was known, it was omitting the HELO data
4734 if it matched the actual IP address. This has been tidied up so that it
4735 doesn't show the same IP address twice.
4737 PH/07 When both +timestamp and +memory debugging was on, the value given by
4738 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4739 called by the debug printing, thereby overwriting the timestamp buffer.
4740 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4742 PH/08 When the original message was included in an autoreply transport, it
4743 always said "this is a copy of the message, including all the headers",
4744 even if body_only or headers_only was set. It now gives an appropriate
4747 PH/09 Applied a patch from the Sieve maintainer which:
4749 o fixes some comments
4750 o adds the (disabled) notify extension core
4751 o adds some debug output for the result of if/elsif tests
4752 o points to the current vacation draft in the documentation
4753 and documents the missing references header update
4757 o fixes a bug in processing the envelope test (when testing
4758 multiple envelope elements, the last element determined the
4761 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4762 Electronic Mail") by including:
4764 Auto-submitted: auto-generated
4766 in the messages that it generates (bounce messages and others, such as
4767 warnings). In the case of bounce messages for non-SMTP messages, there was
4768 also a typo: it was using "Auto_submitted" (underscore instead of
4769 hyphen). Since every message generated by Exim is necessarily in response
4770 to another message, thes have all been changed to:
4772 Auto-Submitted: auto-replied
4774 in accordance with these statements in the RFC:
4776 The auto-replied keyword:
4778 - SHOULD be used on messages sent in direct response to another
4779 message by an automatic process,
4781 - MUST NOT be used on manually-generated messages,
4783 - MAY be used on Delivery Status Notifications (DSNs) and Message
4784 Disposition Notifications (MDNs),
4786 - MUST NOT be used on messages generated by automatic or periodic
4787 processes, except for messages which are automatic responses to
4790 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4791 to the default Received: header definition.
4793 PH/12 Added log selector acl_warn_skipped (default on).
4795 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4796 variables because (a) they are in the wrong storage pool and (b) even if
4797 they were copied, it wouldn't work properly because of the caching.
4799 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4800 checking when decoding. Apparently there are clients that generate
4801 overlong encoded strings. Why am I not surprised?
4803 PH/15 If the first argument of "${if match_address" was not empty, but did not
4804 contain an "@" character, Exim crashed. Now it writes a panic log message
4805 and treats the condition as false.
4807 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4809 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4810 extension "envelope-auth". The code is finished and in agreement with
4811 other implementations, but there is no documentation so far and in fact,
4812 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4813 not changing the active code.
4815 Print executed "if" and "elsif" statements when debugging is used. This
4816 helps a great deal to understand what a filter does.
4818 Document more things not specified clearly in RFC3028. I had all this
4819 sorted out, when out of a sudden new issues came to my mind. Oops."
4821 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4824 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4825 canonical form (as documented). However, after a host name lookup from
4826 the IP address, check_host() was doing a simple string comparison with
4827 addresses acquired from the DNS when checking that the found name did
4828 have the original IP as one of its addresses. Since any found IPv6
4829 addresses are likely to be in abbreviated form, the comparison could
4830 fail. Luckily, there already exists a function for doing the comparison
4831 by converting both addresses to binary, so now that is used instead of
4832 the text comparison.
4834 PH/20 There was another similar case to PH/19, when a complete host name was
4835 given in a host list; looking up its IP address could give an abbreviated
4836 form, whereas the current host's name might or might not be abbreviated.
4837 The same fix has been applied.
4843 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4844 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4847 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4848 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4850 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4852 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4853 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4854 Also, for "Auto-Submitted", treat anything other than "no" as
4855 non-personal, in accordance with RFC 3834. (Previously it treated
4856 anything starting "auto-" as non-personal.)
4858 TF/01 The control=submission/name=... option had a problem with syntax
4859 errors if the name included a slash character. The /name= option
4860 now slurps the rest of the string, so it can include any characters
4861 but it must come last in the list of options (after /sender_retain
4864 PH/05 Some modifications to the interface to the fake nameserver for the new
4872 TK/01 Added the "success_on_redirect" address verification option. See
4873 NewStuff for rationale and an example.
4875 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4877 PH/02 Patch to exigrep to allow it to work on syslog lines.
4879 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4880 fread() to skip over the body file's header line, because in Cygwin the
4881 header line is locked and is inaccessible.
4883 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4884 co-exist for some time) to make it clear that it is the Exim ID that is
4885 referenced, not the Message-ID: header line.
4887 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4888 string_format() function, because snprintf() does not exist on all
4891 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4892 ${stat: expansion item.
4894 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4895 protocol synchronization error", to keep the pedants happy.
4897 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4898 well as for IRIX systems, when gcc is being used. See the host.c source
4901 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4903 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4906 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4907 transport and to the smtp transport in LMTP mode.
4909 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4911 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4912 run a filter in a subprocess. This could lead to confusion in subsequent
4913 lookups in the parent process. There should also be a search_tidyup() at
4914 the end of the subprocess.
4916 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4917 only if the host matched helo_try_verify_hosts, which caused the
4918 verification to occur when the EHLO/HELO command was issued. The ACL just
4919 tested the remembered result. Now, if a previous verification attempt has
4920 not happened, "verify = helo" does it there and then.
4922 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4924 TK/03 Fix log output including CR from clamd.
4926 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4927 exist provoked a memory error which could cause a segfault.
4929 PH/15 Installed PCRE 6.2
4931 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4933 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4934 of the problem. Specifically, suggested +O2 rather than +O1 for the
4937 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4939 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4940 record, the port was not correctly logged when the outgoing_port log
4941 selector was set (it logged the transort's default port).
4943 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4944 fallback_hosts, and "hosts" in the smtp transport.
4946 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4947 host errors such as "Connection refused".
4949 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4950 authentication with radiusclient 0.4.9:
4952 - Error returned from rc_read_config was caught wrongly
4953 - Username/password not passed on to radius server due to wrong length.
4955 The presumption is that some radiusclient API changes for 4.51/PH/17
4956 were not taken care of correctly. The code is still untested by me (my
4957 Linux distribution still has 0.3.2 of radiusclient), but it was
4958 contributed by a Radius user.
4960 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4961 expanding the "port" option of the smtp transport.
4963 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4964 while reading a MIME header. Thanks to Tom Hughes for a patch.
4966 PH/24 Include config.h inside local_scan.h so that configuration settings are
4969 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4970 that in an ACL for RCPT (for example), you can examine exactly what was
4973 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4974 commands, but it was not correctly comparing the address with the actual
4975 client host address. Thus, it would show the EHLO address in Received:
4976 header lines when this was not necessary.
4978 PH/27 Added the % operator to ${eval:}.
4980 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4981 it should be ignoring failures (because with -C, for example, it has lost
4982 privilege). It wasn't ignoring creation failures other than "already
4985 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4988 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4989 because an input line was too long, either on its own, or by virtue of
4990 too many continuations, the temporary file was not being removed, and the
4991 return code was incorrect.
4993 PH/31 Missing "BOOL" in function definition in filtertest.c.
4995 PH/32 Applied Sieve patches from the maintainer.
4997 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4999 PH/33 Added "verify = not_blind".
5001 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5002 Local/Makefile (with some defaults set). These are used in built scripts
5003 such as exicyclog, but they have never been used in the exim_install
5004 script (though there are many overriding facilities there). I have
5005 arranged that the exim_install script now takes note of these two
5008 PH/35 Installed configuration files for Dragonfly.
5010 PH/36 When a locally submitted message by a trusted user did not contain a
5011 From: header, and the sender address was obtained from -f or from an SMTP
5012 MAIL command, and the trusted user did not use -F to supply a sender
5013 name, $originator_name was incorrectly used when constructing a From:
5014 header. Furthermore, $originator_name was used for submission mode
5015 messages from external hosts without From: headers in a similar way,
5016 which is clearly wrong.
5018 PH/37 Added control=suppress_local_fixups.
5020 PH/38 When log_selector = +received_sender was set, and the addition of the
5021 sender made the log line's construction buffer exactly full, or one byte
5022 less than full, an overflow happened when the terminating "\n" was
5025 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5026 when the result of a list match is failure because a DNS lookup failed.
5028 PH/40 RM_COMMAND is now used in the building process.
5030 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5031 the "build-* directories that it finds.
5033 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5034 address in a domain literal was a prefix of an interface address.
5036 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5037 when verifying a sender address, unless rewrite_headers is false.
5039 PH/44 Wrote a long comment about why errors_to addresses are verified as
5040 recipients, not senders.
5042 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5043 the ratelimit ACL was added.
5045 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5047 PH/46 Added extra information about PostgreSQL errors to the error string.
5049 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5050 avoiding the need to install special zones in a real server. This is
5051 backwards compatible; if it can't find the fake resolver, it drops back.
5052 Thus, both old and new test suites can be run.
5054 TF/02 Added util/ratelimit.pl
5056 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5057 clock is set back in time.
5059 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5060 Candler <B.Candler@pobox.com>.
5062 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5063 OK for addresses that are the result of redirections.
5065 PH/48 A number of further additions for the benefit of the new test suite,
5066 including a fake gethostbyname() that interfaces to the fake DNS resolver
5069 TF/06 The fix for widen_domains has also been applied to qualify_single and
5070 search_parents which are the other dnslookup options that can cause
5073 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5076 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5078 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5079 service" if it encountered an MX record with an empty target hostname.
5080 The message is now "an MX or SRV record indicated no SMTP service".
5082 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5083 if the DNS of the sending site is misconfigured. This is quite a
5084 common situation. This change restores the behaviour of treating a
5085 helo verification defer as a failure.
5087 PH/51 If self=fail was set on a router, the bounce message did not include the
5088 actual error message.
5094 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5096 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5097 command itself ended in error, the underlying message about the transport
5098 filter timeout was being overwritten with the pipe command error. Now the
5099 underlying error message should be appended to the second error message.
5101 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5103 PH/02 Reduce the amount of output that "make" produces by default. Full output
5104 can still be requested.
5106 PH/03 The warning log line about a condition test deferring for a "warn" verb
5107 was being output only once per connection, rather than after each
5108 occurrence (because it was using the same function as for successful
5109 "warn" verbs). This seems wrong, so I have changed it.
5111 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5112 it should not have, which might have caused a crash in the right
5113 circumstances, but probably never did.
5115 PH/04 Installed a modified version of Tony Finch's patch to make submission
5116 mode fix the return path as well as the Sender: header line, and to
5117 add a /name= option so that you can make the user's friendly name appear
5120 TF/03 Added the control = fakedefer ACL modifier.
5122 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5123 Mark Lowes for thorough testing.
5125 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5127 TK/03 Merged latest SRS patch from Miles Wilton.
5129 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5130 with the definition in sysexits.h (which is #included earlier).
5131 Fortunately, Exim does not actually use EX_OK. The code used to try to
5132 preserve the sysexits.h value, by assuming that macro definitions were
5133 scanned for macro replacements. I have been disabused of this notion,
5134 so now the code just undefines EX_OK before #including unistd.h.
5136 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5137 in the smtp transport. When a block could not be written in a single
5138 write() function, the timeout was being re-applied to each part-write.
5139 This seems wrong - if the receiver was accepting one byte at a time it
5140 would take for ever. The timeout is now adjusted when this happens. It
5141 doesn't have to be particularly precise.
5143 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5144 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5146 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5147 without a domain if the check to <postmaster@domain> fails.
5149 SC/01 Eximstats: added -xls and the ability to specify output files
5150 (patch written by Frank Heydlauf).
5152 SC/02 Eximstats: use FileHandles for outputting results.
5154 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5156 SC/04 Eximstats: fixed display of large numbers with -nvr option
5158 SC/05 Eximstats: fixed merging of reports with empty tables.
5160 SC/06 Eximstats: added the -include_original_destination flag
5162 SC/07 Eximstats: removed tabs and trailing whitespace.
5164 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5166 TK/06 MBOX spool code: Add real "From " MBOX separator line
5167 so the .eml file is really in mbox format (even though
5168 most programs do not really care). Patch from Alex Miller.
5170 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5171 The latter is generated from $received_to and is only set if the
5172 message has one envelope recipient. SA can use these headers,
5173 obviously out-of-the-box. Patch from Alex Miller.
5175 PH/08 The ${def test on a variable was returning false if the variable's
5176 value was "0", contrary to what the specification has always said!
5177 The result should be true unless the variable is empty.
5179 PH/09 The syntax error of a character other than { following "${if
5180 def:variable_name" (after optional whitespace) was not being diagnosed.
5181 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5182 accidental colon was present, for example, could give incorrect results.
5184 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5185 result is used (not including appendfile, where other changes are about
5188 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5189 This involved changing a lot of size variables from int to off_t. It
5190 should work with maildirs and everything.
5192 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5193 spamd dying while we are connected to it.
5195 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5198 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5199 function for BDB 4.3.
5201 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5203 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5204 the sources for PCRE have been re-arranged and more files are now
5207 PH/15 The code I had for printing potentially long long variables in PH/11
5208 above was not the best (it lost precision). The length of off_t variables
5209 is now inspected at build time, and an appropriate printing format (%ld
5210 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5211 to be "long long int" or "long int". This is needed for the internal
5212 formatting function string_vformat().
5214 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5215 the configuration file to be ":syslog", then the script "guesses" where
5216 the logs files are, rather than using the compiled in default. In our
5217 case the guess is not the same as the compiled default, so the script
5218 suddenly stopped working when I started to use syslog. The patch checks
5219 to see if log_file_path is "". If so, it attempts to read it from exim
5220 with no configuration file to get the compiled in version, before it
5221 falls back to the previous guessing code."
5223 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5224 implementing BATV in an Exim configuration. See NewStuff for the gory
5227 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5228 Makefile that are specific to HP-UX.
5230 PH/18 If the "use_postmaster" option was set for a recipient callout together
5231 with the "random" option, the postmaster address was used as the MAIL
5232 FROM address for the random test, but not for the subsequent recipient
5233 test. It is now used for both.
5235 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5236 patch removes a few documentation additions to RFC 3028, because the
5237 latest draft now contains them. It adds the new en;ascii-case comparator
5238 and a new error check for 8bit text in MIME parts. Comparator and
5239 require names are now matched exactly. I enabled the subaddress
5240 extension, but it is not well tested yet (read: it works for me)."
5242 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5243 rework some of the code of TK/09 above to avoid the hardwired use of
5244 "%lld" and "long long". Replaced the call to snprintf() with a call to
5247 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5248 records point to non-existent hosts", "retry timeout exceeded", and
5249 "retry time not reached for any host after a long failure period".
5251 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5252 experimental DomainKeys support:
5254 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5255 (2) On an error such as an illegally used "control", the wrong name for
5256 the control was given.
5258 These problems did NOT occur unless DomainKeys support was compiled.
5260 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5262 PH/24 Added ${if match_ip condition.
5264 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5265 databases so that it will be absolutely obvious if a crash occurs in the
5266 DB library. This is a regular occurrence (often caused by mis-matched
5269 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5270 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5271 picked up on a user's system that detects such things. There doesn't seem
5272 to be a gcc warning option for this - only an attribute that has to be
5273 put on the function's prototype. It seems that in Fedora Core 4 they have
5274 set this on a number of new functions. No doubt there will be more in due
5277 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5278 specify a transport. However, if an address that was verified by such a
5279 router was the subject of a callout, Exim crashed because it tried to
5280 read the rcpt_include_affixes from the non-existent transport. Now it
5281 just assumes that the setting of that option is false. This bug was
5282 introduced by 4.51/PH/31.
5284 PH/28 Changed -d+all to exclude +memory, because that information is very
5285 rarely of interest, but it makes the output a lot bigger. People tend to
5286 do -d+all out of habit.
5288 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5289 code in os-type was giving problems when libc.so lives in lib64, like on
5292 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5293 aren't the modern standard, and it seems that some systems' include files
5294 don't always have them. Exim was already checking for some of the newer
5295 ones like T_AAAA, and defining it itself. I've added checks for all the
5296 record types that Exim uses.
5298 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5299 not automatically generating a new one, as it is supposed to. This
5300 prevented TLS from working. If the file did exist, but contained invalid
5301 data, a new version was generated, as expected. It was only the case of a
5302 non-existent file that was broken.
5304 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5305 with a change in libdomainkeys > 0.64.
5307 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5308 from DNS. If the selector record carries the flag, it now has
5309 precedence over the domain-wide flag.
5311 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5313 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5314 the use of a port name that isn't defined in /etc/services) occurred, the
5315 message was deferred as in a normal delivery, and thus remained on the
5316 spool, instead of being failed because of the mua_wrapper setting. This
5317 is now fixed, and I tidied up some of the mua_wrapper messages at the
5320 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5321 the messages in a hash of arrays rather than using individual hashes.
5322 This is a bit cleaner and results in dramatic memory savings, albeit
5323 at a slight CPU cost.
5325 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5326 as requested by Marc Sherman.
5328 SC/10 Eximstats: added histograms for user specified patterns as requested
5331 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5333 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5334 fopen() in the content-scanning modules that did not already have it.
5340 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5341 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5343 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5345 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5347 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5348 to test Sieve filters that use "vacation".
5350 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5351 that changes the way the GnuTLS parameters are stored in the cache file.
5352 The new format can be generated externally. For backward compatibility,
5353 if the data in the cache doesn't make sense, Exim assumes it has read an
5354 old-format file, and it generates new data and writes a new file. This
5355 means that you can't go back to an older release without removing the
5358 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5359 work if there are any delivery delays because "one_time" forces the
5360 parent to be marked "delivered", so its unseen clone is never tried
5361 again. For this reason, Exim now forbids the simultaneous setting of
5364 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5365 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5366 entry if you want to know the details. The fix, however, neglected to
5367 consider the case when local delivery batching is involved. The test for
5368 "previously delivered" was not happening when checking to see if an
5369 address could be batched with a previous (undelivered) one; under
5370 certain circumstances this could lead to multiple deliveries to the same
5373 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5374 in its include files, and this causes problems building Exim.
5376 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5377 header_syntax) but Exim was just ignoring anything given after a slash.
5378 In particular, this caused confusion with an attempt to use "verify =
5379 reverse_host_lookup/defer_ok". An error is now given when options are
5380 supplied for verify items that do not have them. (Maybe reverse_host_
5381 lookup should have a defer_ok option, but that's a different point.)
5383 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5384 defined by RFC 821) to 2048, because there were problems with some AUTH
5385 commands, and RFC 1869 says the size should be increased for extended
5386 SMTP commands that take arguments.
5388 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5391 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5392 "unknown" error; now it says that the functionality isn't in the binary.
5394 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5395 an address' error message when a string expansion fails (syntax or
5396 whatever). Otherwise the password may appear in the log. Following change
5397 PH/42 below, there is no longer a chance of it appearing in a bounce
5400 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5402 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5403 of its addresses was ever tried. (Bugzilla bug #2.)
5405 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5406 the result incorrectly in the debug output. (It correctly added a newline
5407 to what was transported.)
5409 TF/01 Added $received_time.
5411 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5412 commented out examples of how to interface to a virus scanner and to
5413 SpamAssassin. Also added commented examples of av_scanner and
5414 spamd_address settings.
5416 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5417 and controls are allowed in which ACLs. There were a couple of minor
5418 errors. Some of the entries in the conditions table (which is a table of
5419 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5420 negation of where the condition IS allowed.
5422 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5424 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5425 header file does not have a version number, so I've had to invent a new
5426 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5427 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5428 radiusclient), but it was contributed by a Radius user.
5430 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5431 files or directories, for OpenSSL.
5433 PH/19 When an Exim process that is running as root has to create an Exim log
5434 file, it does so in a subprocess that runs as exim:exim so as to get the
5435 ownership right at creation (otherwise, other Exim processes might see
5436 the file with the wrong ownership). There was no test for failure of this
5437 fork() call, which would lead to the process getting stuck as it waited
5438 for a non-existent subprocess. Forks do occasionally fail when resources
5439 run out. I reviewed all the other calls to fork(); they all seem to check
5442 PH/20 When checking for unexpected SMTP input at connect time (before writing
5443 the banner), Exim was not dealing correctly with a non-positive return
5444 from the read() function. If the client had disconnected by this time,
5445 the result was a log entry for a synchronization error with an empty
5446 string after "input=" when read() returned zero. If read() returned -1
5447 (an event I could not check), uninitialized data bytes were printed.
5448 There were reports of junk text (parts of files, etc) appearing after
5451 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5453 PH/22 Added support for macro redefinition, and (re)definition in between
5454 driver and ACL definitions.
5456 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5457 forgetting to use the resulting value; it was using the unexpanded value.
5459 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5460 hadn't been configured. The fix is from Juergen Kreileder, who
5461 understands it better than I do:
5463 "Here's what I see happening with three configured cyrus_sasl
5464 authenticators configured (plain, login, cram-md5):
5466 On startup auth_cyrus_sasl_init() gets called for each of these.
5467 This means three calls to sasl_listmech() without a specified mech_list.
5468 => SASL tests which mechs of all available mechs actually work
5469 => three warnings about OTP not working
5470 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5472 With the patch, sasl_listmech() also gets called three times. But now
5473 SASL's mech_list option is set to the server_mech specified in the the
5474 authenticator. Or in other words, the answer from sasl_listmech()
5475 gets limited to just the mech you're testing for (which is different
5477 => the return list contains just 'plain' or 'login', 'cram-md5' or
5478 nothing depending on the value of ob->server_mech.
5480 I've just tested the patch: Authentication still works fine,
5481 unavailable mechs specified in the exim configuration are still
5482 caught, and the auth.log warnings about OTP are gone."
5484 PH/25 When debugging is enabled, the contents of the command line are added
5485 to the debugging output, even when log_selector=+arguments is not
5488 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5489 answer is "GNU", and only if the return is "GNU/something" is the answer
5492 PH/27 $acl_verify_message is now set immediately after the failure of a
5493 verification in an ACL, and so is available in subsequent modifiers. In
5494 particular, the message can be preserved by coding like this:
5496 warn !verify = sender
5497 set acl_m0 = $acl_verify_message
5499 Previously, $acl_verify_message was set only while expanding "message"
5500 and "log_message" when a very denied access.
5502 PH/28 Modified OS/os.c-Linux with
5504 -#ifndef OS_LOAD_AVERAGE
5505 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5507 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5508 nomenclature these days.)
5510 PH/29 Installed patch from the Sieve maintainer that adds the options
5511 sieve_useraddress and sieve_subaddress to the redirect router.
5513 PH/30 In these circumstances:
5514 . Two addresses routed to the same list of hosts;
5515 . First host does not offer TLS;
5516 . First host accepts first address;
5517 . First host gives temporary error to second address;
5518 . Second host offers TLS and a TLS session is established;
5519 . Second host accepts second address.
5520 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5521 and peerdn, if requested) that were in fact used only for the second
5524 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5525 attention to any local part prefix or suffix that was matched by the
5526 router that accepted the address. It now behaves in the same way as it
5527 does for delivery: the affixes are removed from the local part unless
5528 rcpt_include_affixes is set on the transport.
5530 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5531 timeout during the DATA phase of an incoming message.
5533 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5534 applied a patch sanctioned by the Sieve maintainer.
5536 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5537 the uid or gid is negative. A case of a negative gid caused this to be
5538 noticed. The fix allows for either to be negative.
5540 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5541 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5544 PH/36 The change PH/12 above was broken. Fixed it.
5546 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5547 the grounds that routing the same address twice would always produce the
5548 same answer. This might have been true once, but it is certainly no
5549 longer true now. Routing a child address may depend on the previous
5550 routing that produced that child. Some complicated redirection strategies
5551 went wrong when messages had multiple recipients, and made Exim's
5552 behaviour dependent on the order in which the addresses were given.
5554 I have moved the duplicate checking until after the routing is complete.
5555 Exim scans the addresses that are assigned to local and remote
5556 transports, and removes any duplicates. This means that more work will be
5557 done, as duplicates will always all be routed, but duplicates are
5558 presumably rare, so I don't expect this is of any significance.
5560 For deliveries to pipes, files, and autoreplies, the duplicate checking
5561 still happens during the routing process, since they are not going to be
5564 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5565 It corrects a timeout issue with spamd. This is Ian's comment: "The
5566 background is that sometimes spamd either never reads data from a
5567 connection it has accepted, or it never writes response data. The exiscan
5568 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5569 blindly assumes that writes won't block so it may never time out."
5571 PH/39 Allow G after quota size as well as K and M.
5573 PH/40 The value set for $authenticated_id in an authenticator may not contain
5574 binary zeroes or newlines because the value is written to log lines and
5575 to spool files. There was no check on this. Now the value is run through
5576 the string_printing() function so that such characters are converted to
5577 printable escape sequences.
5579 PH/41 $message_linecount is a new variable that contains the total number of
5580 lines in the message. Compare $body_linecount, which is the count for the
5583 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5584 bounce and delay warning messages, except in certain special cases, which
5587 (a) An SMTP error message from a remote host;
5588 (b) A message specified in a :fail: redirection;
5589 (c) A message specified in a "fail" command in a system filter;
5590 (d) A message specified in a FAIL return from the queryprogram router;
5591 (e) A message specified by the cannot_route_message router option.
5593 In these cases only, Exim does include the error details in bounce and
5594 warning messages. There are also a few cases where bland messages such
5595 as "unrouteable address" or "local delivery error" are given.
5597 PH/43 $value is now also set for the "else" part of a ${run expansion.
5599 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5600 being worked on, but at least Exim now implements the latest version to
5603 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5604 process to complete was treated as a delivery failure, a timeout while
5605 writing the message to the pipe was logged, but erroneously treated as a
5606 successful delivery. Such timeouts include transport filter timeouts. For
5607 consistency with the overall process timeout, these timeouts are now
5608 treated as errors, giving rise to delivery failures by default. However,
5609 there is now a new Boolean option for the pipe transport called
5610 timeout_defer, which, if set TRUE, converts the failures into defers for
5611 both kinds of timeout. A transport filter timeout is now identified in
5614 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5615 systems where "make" and "gmake" are different, calling "gmake" at top
5616 level broke things. I've arranged for the value of $(MAKE) to be passed
5617 from the Makefile to this script so that it can call the same version of
5621 A note about Exim versions 4.44 and 4.50
5622 ----------------------------------------
5624 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5625 changes of various kinds. As a consequence, a big documentation update was
5626 needed. This delayed the release for rather longer than seemed good, especially
5627 in the light of a couple of (minor) security issues. Therefore, the changes
5628 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5629 release. So 4.44 and 4.50 are in effect two different branches that both start
5632 I have left the 4.50 change log unchanged; it contains all the changes since
5633 4.43. The change log for 4.44 is below; many of its items are identical to
5634 those for 4.50. This seems to be the most sensible way to preserve the
5635 historical information.
5641 1. Minor wording change to the doc/README.SIEVE file.
5643 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5644 computation of the current number of files was incorrect.
5646 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5647 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5650 4. Give more explanation in the error message when the command for a transport
5651 filter fails to execute.
5653 5. There are several places where Exim runs a non-Exim command in a
5654 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5655 was being done only for the command run by the queryprogram router. It is
5656 now done for all such subprocesses. The other cases are: ${run, transport
5657 filters, and the commands run by the lmtp and pipe transports.
5659 6. Added CONFIGURE_GROUP build-time option.
5661 7. Some older OS have a limit of 256 on the maximum number of file
5662 descriptors. Exim was using setrlimit() to set 1000 as a large value
5663 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5664 systems. I've change it so that if it can't get 1000, it tries for 256.
5666 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5667 was an oversight, and furthermore, ever since the addition of extra
5668 controls (e.g. 4.43/32), the checks on when to allow different forms of
5669 "control" were broken. There should now be diagnostics for all cases when a
5670 control that does not make sense is encountered.
5672 9. Added the /retain_sender option to "control=submission".
5674 10. $recipients is now available in the predata ACL (oversight).
5676 11. Tidy the search cache before the fork to do a delivery from a message
5677 received from the command line. Otherwise the child will trigger a lookup
5678 failure and thereby defer the delivery if it tries to use (for example) a
5679 cached ldap connection that the parent has called unbind on.
5681 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5682 of $address_data from the recipient verification was clobbered by the
5683 sender verification.
5685 13. The value of address_data from a sender verification is now available in
5686 $sender_address_data in subsequent conditions in the ACL statement.
5688 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5690 15. Added a new option "connect=<time>" to callout options, to set a different
5693 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5694 was its contents. (It was OK if the option was not defined at all.)
5696 17. A "Completed" log line is now written for messages that are removed from
5697 the spool by the -Mrm option.
5699 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5700 information about exactly what failed.
5702 19. Added -dd to debug only the daemon process.
5704 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5705 handles timeouts, both on the server side and network timeouts. Renamed the
5706 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5708 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5709 It is now set to "smtps".
5711 22. $host_address is now set to the target address during the checking of
5712 ignore_target_hosts.
5714 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5715 being passed; this would have caused $sender_host_name to have been used if
5716 matching the list had actually called for a host name (not very likely,
5717 since this list is usually IP addresses). A host name is now passed as
5720 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5721 code that specifies a non-restarting handler (typically sigaction() in
5722 modern systems) in an attempt to fix a rare and obscure crash bug.
5724 25. Narrowed the window for a race in the daemon that could cause it to ignore
5725 SIGCHLD signals. This is not a major problem, because they are used only to
5726 wake it up if nothing else does.
5728 26. A malformed maildirsize file could cause Exim to calculate negative values
5729 for the mailbox size or file count. Odd effects could occur as a result.
5730 The maildirsize information is now recalculated if the size or filecount
5733 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5734 support for a long time. Removed HAVE_SYS_VFS_H.
5736 28. Installed the latest version of exipick from John Jetmore.
5738 29. In an address list, if the pattern was not a regular expression, an empty
5739 subject address (from a bounce message) matched only if the pattern was an
5740 empty string. Non-empty patterns were not even tested. This was the wrong
5741 because it is perfectly reasonable to use an empty address as part of a
5742 database query. An empty address is now tested by patterns that are
5743 lookups. However, all the other forms of pattern expect the subject to
5744 contain a local part and a domain, and therefore, for them, an empty
5745 address still always fails if the pattern is not itself empty.
5747 30. Exim went into a mad DNS loop when attempting to do a callout where the
5748 host was specified on an smtp transport, and looking it up yielded more
5749 than one IP address.
5751 31. Re-factored the code for checking spool and log partition space into a
5752 function that finds that data and another that does the check. The former
5753 is then used to implement four new variables: $spool_space, $log_space,
5754 $spool_inodes, and $log_inodes.
5756 32. The RFC2047 encoding function was originally intended for short strings
5757 such as real names; it was not keeping to the 75-character limit for
5758 encoded words that the RFC imposes. It now respects the limit, and
5759 generates multiple encoded words if necessary. To be on the safe side, I
5760 have increased the buffer size for the ${rfc2047: expansion operator from
5763 33. It is now permitted to omit both strings after an "if" condition; if the
5764 condition is true, the result is "true". As before, when the second string
5765 is omitted, a false condition yields an empty string. This makes it less
5766 cumbersome to write custom ACL and router conditions.
5768 34. Failure to deliver a bounce message always caused it to be frozen, even if
5769 there was an errors_to setting on the router. The errors_to setting is now
5772 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5773 canonical form (fully expanded) before being placed in
5774 $sender_host_address.
5776 36. The table in the code that translates DNS record types into text (T_A to
5777 "A" for instance) was missing entries for NS and CNAME. It is just possible
5778 that this could have caused confusion if both these types were looked up
5779 for the same domain, because the text type is used as part of Exim's
5780 per-process caching. But the chance of anyone hitting this buglet seems
5783 37. The dnsdb lookup has been extended in a number of ways.
5785 (1) There is a new type, "zns", which walks up the domain tree until it
5786 finds some nameserver records. It should be used with care.
5788 (2) There is a new type, "mxh", which is like "mx" except that it returns
5789 just the host names, not the priorities.
5791 (3) It is now possible to give a list of domains (or IP addresses) to be
5792 looked up. The behaviour when one of the lookups defers can be
5793 controlled by a keyword.
5795 (4) It is now possible to specify the separator character for use when
5796 multiple records are returned.
5798 38. The dnslists ACL condition has been extended: it is now possible to supply
5799 a list of IP addresses and/or domains to be looked up in a particular DNS
5802 39. Added log_selector=+queue_time_overall.
5804 40. When running the queue in the test harness, wait just a tad after forking a
5805 delivery process, to get repeatability of debugging output.
5807 41. Include certificate and key file names in error message when GnuTLS fails
5808 to set them up, because the GnuTLS error message doesn't include the name
5809 of the failing file when there is a problem reading it.
5811 42. Allow both -bf and -bF in the same test run.
5813 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5815 44. The "Exiscan patch" is now merged into the mainline Exim source.
5817 45. Sometimes the final signoff response after QUIT could fail to get
5818 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5819 < 0 before doing a fflush(). This bug looks as though it goes back to the
5820 introduction of TLS in release 3.20, but "sometimes" must have been rare
5821 because the tests only now provoked it.
5823 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5824 (this can affect the format of dates).
5826 47. exim_tidydb, when checking for the continued existence of a message for
5827 which it has found a message-specific retry record, was not finding
5828 messages that were in split spool directories. Consequently, it was
5829 deleting retry records that should have stayed in existence.
5831 48. Steve fixed some bugs in eximstats.
5833 49. The SPA authentication driver was not abandoning authentication and moving
5834 on to the next authenticator when an expansion was forced to fail,
5835 contradicting the general specification for all authenticators. Instead it
5836 was generating a temporary error. It now behaves as specified.
5838 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5839 (the order specifies the preference for clients). The order is now AES256,
5840 AES128, 3DES, ARCFOUR128.
5842 51. Small patch to Sieve code - explicitly set From: when generating an
5845 52. Exim crashed if a remote delivery caused a very long error message to be
5846 recorded - for instance if somebody sent an entire SpamAssassin report back
5847 as a large number of 550 error lines. This bug was coincidentally fixed by
5848 increasing the size of one of Exim's internal buffers (big_buffer) that
5849 happened as part of the Exiscan merge. However, to be on the safe side, I
5850 have made the code more robust (and fixed the comments that describe what
5853 53. Now that there can be additional text after "Completed" in log lines (if
5854 the queue_time_overall log selector is set), a one-byte patch to exigrep
5855 was needed to allow it to recognize "Completed" as not the last thing in
5858 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5859 patch that reportedly fixes this has been added. I am not expert enough to
5860 create a test for it. This is what the patch creator wrote:
5862 "I found a little strange behaviour of ldap code when working with
5863 Windows 2003 AD Domain, where users was placed in more than one
5864 Organization Units. When I tried to give exim partial DN, the exit code
5865 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5866 But simultaneously result of request was absolutely normal ldap result,
5867 so I produce this patch..."
5869 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5870 REFERENCE, so I have modified the code to exclude the patch when that macro
5873 55. Some experimental protocols are using DNS PTR records for new purposes. The
5874 keys for these records are domain names, not reversed IP addresses. The
5875 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5876 leaves it alone. Component reversal etc. now happens only for IP addresses.
5879 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5881 57. Double the size of the debug message buffer (to 2048) so that more of very
5882 long debug lines gets shown.
5884 58. The exicyclog utility now does better if the number of log files to keep
5885 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5887 59. Two changes related to the smtp_active_hostname option:
5889 (1) $smtp_active_hostname is now available as a variable.
5890 (2) The default for smtp_banner uses $smtp_active_hostname instead
5891 of $primary_hostname.
5893 60. The host_aton() function is supposed to be passed a string that is known
5894 to be a valid IP address. However, in the case of IPv6 addresses, it was
5895 not checking this. This is a hostage to fortune. Exim now panics and dies
5896 if the condition is not met. A case was found where this could be provoked
5897 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5898 components; fortuitously, this particular loophole had already been fixed
5899 by change 4.50/55 above.
5901 If there are any other similar loopholes, the new check in host_aton()
5902 itself should stop them being exploited. The report I received stated that
5903 data on the command line could provoke the exploit when Exim was running as
5904 exim, but did not say which command line option was involved. All I could
5905 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5906 running as the user.
5909 61. There was a buffer overflow vulnerability in the SPA authentication code
5910 (which came originally from the Samba project). I have added a test to the
5911 spa_base64_to_bits() function which I hope fixes it.
5914 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5915 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5917 63. The daemon start-up calls getloadavg() while still root for those OS that
5918 need the first call to be done as root, but it missed one case: when
5919 deliver_queue_load_max is set with deliver_drop_privilege. This is
5920 necessary for the benefit of the queue runner, because there is no re-exec
5921 when deliver_drop_privilege is set.
5923 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5924 This has been fixed.
5926 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5927 was in use, was not putting the data itself into the right store pool;
5928 consequently, it could be overwritten for a subsequent message in the same
5929 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5932 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5934 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5935 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5936 the function were treating the return as a boolean value, which happened to
5937 work because 0=false and not-0=true, but is not correct code.
5939 68. The host_aton() function was not handling scoped IPv6 addresses (those
5940 with, for example, "%eth0" on the end) correctly.
5942 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5943 negated items (that is, ~something) in unsigned ints. Some compilers
5944 apparently mutter when there is no cast.
5946 70. If an address verification called from an ACL failed, and did not produce a
5947 user-specific message (i.e. there was only a "system" message), nothing was
5948 put in $acl_verify_message. In this situation, it now puts the system
5951 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5952 session; change 4.31/43 added the unwanted input to the log line - except
5953 that it did not do this in the start of session case. It now does.
5955 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5956 This is wrong and can cause the other end to generate a synchronization
5957 error if it is another Exim or anything else that does the synchronization
5958 check. A QUIT command is no longer sent after a timeout.
5960 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5961 during host lookups.
5963 74. The defer_ok option of callout verification was not working if it was used
5964 when verifying addresses in header lines, that is, for this case:
5966 verify = header_sender/callout=defer_ok
5968 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5969 those file descriptors could be used for SMTP connections. If anything
5970 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5971 could be sent to the SMTP client, causing chaos. The daemon now opens
5972 stdin, stdout, and stderr to /dev/null when it puts itself into the
5975 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5976 log by default. The user can override this with suitable Perl magic.
5978 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5979 the log message when discard triggers, was not working for the DATA ACL or
5980 for the non-SMTP ACL.
5982 78. Error message wording change in sieve.c.
5984 79. If smtp_accept_max_per_host was set, the number of connections could be
5985 restricted to fewer than expected, because the daemon was trying to set up
5986 a new connection before checking whether the processes handling previous
5987 connections had finished. The check for completed processes is now done
5988 earlier. On busy systems, this bug wouldn't be noticed because something
5989 else would have woken the daemon, and it would have reaped the completed
5992 80. If a message was submitted locally by a user whose login name contained one
5993 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5994 It caused a spool format error. I have fixed the spool reading code. A
5995 related problem was that the "from" clause in the Received: line became
5996 illegal because of the space(s). It is now covered by ${quote_local_part.
5998 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6001 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6003 83. Patch from Sieve maintainer to fix unterminated string problem in
6004 "vacation" handling.
6006 84. Some minor changes to the Linux configuration files to help with other
6007 OS variants using glibc.
6009 85. One more patch for Sieve to update vacation handling to latest spec.
6012 ----------------------------------------------------
6013 See the note above about the 4.44 and 4.50 releases.
6014 ----------------------------------------------------
6020 1. Change 4.43/35 introduced a bug that caused file counts to be
6021 incorrectly computed when quota_filecount was set in an appendfile
6024 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6025 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6028 3. Give more explanation in the error message when the command for a transport
6029 filter fails to execute.
6031 4. There are several places where Exim runs a non-Exim command in a
6032 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6033 was being done only for the command run by the queryprogram router. It is
6034 now done for all such subprocesses. The other cases are: ${run, transport
6035 filters, and the commands run by the lmtp and pipe transports.
6037 5. Some older OS have a limit of 256 on the maximum number of file
6038 descriptors. Exim was using setrlimit() to set 1000 as a large value
6039 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6040 systems. I've change it so that if it can't get 1000, it tries for 256.
6042 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6043 was an oversight, and furthermore, ever since the addition of extra
6044 controls (e.g. 4.43/32), the checks on when to allow different forms of
6045 "control" were broken. There should now be diagnostics for all cases when a
6046 control that does not make sense is encountered.
6048 7. $recipients is now available in the predata ACL (oversight).
6050 8. Tidy the search cache before the fork to do a delivery from a message
6051 received from the command line. Otherwise the child will trigger a lookup
6052 failure and thereby defer the delivery if it tries to use (for example) a
6053 cached ldap connection that the parent has called unbind on.
6055 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6056 of $address_data from the recipient verification was clobbered by the
6057 sender verification.
6059 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6060 was its contents. (It was OK if the option was not defined at all.)
6062 11. A "Completed" log line is now written for messages that are removed from
6063 the spool by the -Mrm option.
6065 12. $host_address is now set to the target address during the checking of
6066 ignore_target_hosts.
6068 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6069 being passed; this would have caused $sender_host_name to have been used if
6070 matching the list had actually called for a host name (not very likely,
6071 since this list is usually IP addresses). A host name is now passed as
6074 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6075 code that specifies a non-restarting handler (typically sigaction() in
6076 modern systems) in an attempt to fix a rare and obscure crash bug.
6078 15. Narrowed the window for a race in the daemon that could cause it to ignore
6079 SIGCHLD signals. This is not a major problem, because they are used only to
6080 wake it up if nothing else does.
6082 16. A malformed maildirsize file could cause Exim to calculate negative values
6083 for the mailbox size or file count. Odd effects could occur as a result.
6084 The maildirsize information is now recalculated if the size or filecount
6087 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6088 support for a long time. Removed HAVE_SYS_VFS_H.
6090 18. Updated exipick to current release from John Jetmore.
6092 19. Allow an empty sender to be matched against a lookup in an address list.
6093 Previously the only cases considered were a regular expression, or an
6096 20. Exim went into a mad DNS lookup loop when doing a callout where the
6097 host was specified on the transport, if the DNS lookup yielded more than
6100 21. The RFC2047 encoding function was originally intended for short strings
6101 such as real names; it was not keeping to the 75-character limit for
6102 encoded words that the RFC imposes. It now respects the limit, and
6103 generates multiple encoded words if necessary. To be on the safe side, I
6104 have increased the buffer size for the ${rfc2047: expansion operator from
6107 22. Failure to deliver a bounce message always caused it to be frozen, even if
6108 there was an errors_to setting on the router. The errors_to setting is now
6111 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6112 canonical form (fully expanded) before being placed in
6113 $sender_host_address.
6115 24. Updated eximstats to version 1.33
6117 25. Include certificate and key file names in error message when GnuTLS fails
6118 to set them up, because the GnuTLS error message doesn't include the name
6119 of the failing file when there is a problem reading it.
6121 26. Expand error message when OpenSSL has problems setting up cert/key files.
6124 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6125 (this can affect the format of dates).
6127 28. exim_tidydb, when checking for the continued existence of a message for
6128 which it has found a message-specific retry record, was not finding
6129 messages that were in split spool directories. Consequently, it was
6130 deleting retry records that should have stayed in existence.
6132 29. eximstats updated to version 1.35
6133 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6134 1.35 - bugfix such that pie charts by volume are generated correctly
6136 30. The SPA authentication driver was not abandoning authentication and moving
6137 on to the next authenticator when an expansion was forced to fail,
6138 contradicting the general specification for all authenticators. Instead it
6139 was generating a temporary error. It now behaves as specified.
6141 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6142 (the order specifies the preference for clients). The order is now AES256,
6143 AES128, 3DES, ARCFOUR128.
6145 31. Small patch to Sieve code - explicitly set From: when generating an
6148 32. Exim crashed if a remote delivery caused a very long error message to be
6149 recorded - for instance if somebody sent an entire SpamAssassin report back
6150 as a large number of 550 error lines. This bug was coincidentally fixed by
6151 increasing the size of one of Exim's internal buffers (big_buffer) that
6152 happened as part of the Exiscan merge. However, to be on the safe side, I
6153 have made the code more robust (and fixed the comments that describe what
6156 33. Some experimental protocols are using DNS PTR records for new purposes. The
6157 keys for these records are domain names, not reversed IP addresses. The
6158 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6159 leaves it alone. Component reversal etc. now happens only for IP addresses.
6162 34. The host_aton() function is supposed to be passed a string that is known
6163 to be a valid IP address. However, in the case of IPv6 addresses, it was
6164 not checking this. This is a hostage to fortune. Exim now panics and dies
6165 if the condition is not met. A case was found where this could be provoked
6166 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6167 components; fortuitously, this particular loophole had already been fixed
6168 by change 4.50/55 or 4.44/33 above.
6170 If there are any other similar loopholes, the new check in host_aton()
6171 itself should stop them being exploited. The report I received stated that
6172 data on the command line could provoke the exploit when Exim was running as
6173 exim, but did not say which command line option was involved. All I could
6174 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6175 running as the user.
6178 35. There was a buffer overflow vulnerability in the SPA authentication code
6179 (which came originally from the Samba project). I have added a test to the
6180 spa_base64_to_bits() function which I hope fixes it.
6183 36. The daemon start-up calls getloadavg() while still root for those OS that
6184 need the first call to be done as root, but it missed one case: when
6185 deliver_queue_load_max is set with deliver_drop_privilege. This is
6186 necessary for the benefit of the queue runner, because there is no re-exec
6187 when deliver_drop_privilege is set.
6189 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6190 was in use, was not putting the data itself into the right store pool;
6191 consequently, it could be overwritten for a subsequent message in the same
6192 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6195 38. Sometimes the final signoff response after QUIT could fail to get
6196 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6197 < 0 before doing a fflush(). This bug looks as though it goes back to the
6198 introduction of TLS in release 3.20, but "sometimes" must have been rare
6199 because the tests only now provoked it.
6205 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6206 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6207 Now it returns an integer. A number of calls were still expecting a T/F
6208 return. Fortuitously, in all cases, the tests worked in OK situations,
6209 which is the norm. However, things would have gone wrong on any write
6210 failures on the smtp file descriptor. This function is used when sending
6211 messages over SMTP and also when doing verify callouts.
6213 2. When Exim is called to do synchronous delivery of a locally submitted
6214 message (the -odf or -odi options), it no longer closes stderr before doing
6217 3. Implemented the mua_wrapper option.
6219 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6221 5. Implemented the functions header_remove(), header_testname(),
6222 header_add_at_position(), and receive_remove_recipient(), and exported them
6225 6. If an ACL "warn" statement specified the addition of headers, Exim already
6226 inserted X-ACL-Warn: at the start if there was no header name. However, it
6227 was not making this test for the second and subsequent header lines if
6228 there were newlines in the string. This meant that an invalid header could
6229 be inserted if Exim was badly configured.
6231 7. Allow an ACL "warn" statement to add header lines at the start or after all
6232 the Received: headers, as well as at the end.
6234 8. Added the rcpt_4xx retry error code.
6236 9. Added postmaster_mailfrom=xxx to callout verification option.
6238 10. Added mailfrom=xxxx to the callout verification option, for verify=
6241 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6242 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6244 12. Inserted some casts to stop certain compilers warning when using pointer
6245 differences as field lengths or precisions in printf-type calls (mostly
6246 affecting debugging statements).
6248 13. Added optional readline() support for -be (dynamically loaded).
6250 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6251 same clock tick as a message's arrival, so that its received time was the
6252 same as the "first fail" time on the retry record, and that message
6253 remained on the queue past the ultimate address timeout, every queue runner
6254 would try a delivery (because it was past the ultimate address timeout) but
6255 after another failure, the ultimate address timeout, which should have then
6256 bounced the address, did not kick in. This was a "< instead of <=" error;
6257 in most cases the first failure would have been in the next clock tick
6258 after the received time, and all would be well.
6260 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6261 being recognized when the domain list was tested by the match_domain
6262 condition in an expansion string.
6264 16. Added the ${str2b64: operator.
6266 17. Exim was always calling setrlimit() to set a large limit for the number of
6267 processes, without checking whether the existing limit was already
6268 adequate. (It did check for the limit on file descriptors.) Furthermore,
6269 errors from getrlimit() and setrlimit() were being ignored. Now they are
6270 logged to the main and panic logs, but Exim does carry on, to try to do its
6271 job under whatever limits there are.
6273 18. Imported PCRE 5.0.
6275 19. Trivial typo in log message " temporarily refused connection" (the leading
6278 20. If the log selector return_path_on_delivery was set and an address was
6279 redirected to /dev/null, the delivery process crashed because it assumed
6280 that a return path would always be set for a "successful" delivery. In this
6281 case, the whole delivery is bypassed as an optimization, and therefore no
6284 21. Internal re-arrangement: the function for sending a challenge and reading
6285 a response while authentication was assuming a zero-terminated challenge
6286 string. It's now changed to take a pointer and a length, to allow for
6287 binary data in such strings.
6289 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6291 23. Exim was not respecting finduser_retries when seeking the login of the
6292 uid under which it was called; it was always trying 10 times. (The default
6293 setting of finduser_retries is zero.) Also, it was sleeping after the final
6294 failure, which is pointless.
6296 24. Implemented tls_on_connect_ports.
6298 25. Implemented acl_smtp_predata.
6300 26. If the domain in control=submission is set empty, Exim assumes that the
6301 authenticated id is a complete email address when it generates From: or
6302 Sender: header lines.
6304 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6305 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6306 chown and chgrp in /bin and hostname in /usr/bin.
6308 28. Exim was keeping the "process log" file open after each use, just as it
6309 does for the main log. This opens the possibility of it remaining open for
6310 long periods when the USR1 signal hits a daemon. Occasional processlog
6311 errors were reported, that could have been caused by this. Anyway, it seems
6312 much more sensible not to leave this file open at all, so that is what now
6315 29. The long-running daemon process does not normally write to the log once it
6316 has entered its main loop, and it closes the log before doing so. This is
6317 so that log files can straightforwardly be renamed and moved. However,
6318 there are a couple of unusual error situations where the daemon does write
6319 log entries, and I had neglected to close the log afterwards.
6321 30. The text of an SMTP error response that was received during a remote
6322 delivery was being truncated at 512 bytes. This is too short for some of
6323 the long messages that one sometimes sees. I've increased the limit to
6326 31. It is now possible to make retry rules that apply only when a message has a
6327 specific sender, in particular, an empty sender.
6329 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6330 it possible to be selective about when SMTP synchronization is enforced.
6332 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6334 32. Implemented hosts_connection_nolog.
6336 33. Added an ACL for QUIT.
6338 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6341 35. Added mailbox_size and mailbox_filecount to appendfile.
6343 36. Added control = no_multiline_responses to ACLs.
6345 37. There was a bug in the logic of the code that waits for the clock to tick
6346 in the case where the clock went backwards by a substantial amount such
6347 that the microsecond fraction of "now" was more than the microsecond
6348 fraction of "then" (but the whole seconds number was less).
6350 38. Added support for the libradius Radius client library this is found on
6351 FreeBSD (previously only the radiusclient library was supported).
6357 1. When certain lookups returned multiple values in the form name=value, the
6358 quoting of the values was not always being done properly. Specifically:
6359 (a) If the value started with a double quote, but contained no whitespace,
6361 (b) If the value contained whitespace other than a space character (i.e.
6362 tabs or newlines or carriage returns) it was not quoted.
6363 This fix has been applied to the mysql and pgsql lookups by writing a
6364 separate quoting function and calling it from the lookup code. The fix
6365 should probably also be applied to nisplus, ibase and oracle lookups, but
6366 since I cannot test any of those, I have not disturbed their existing code.
6368 2. A hit in the callout cache for a specific address caused a log line with no
6369 reason for rejecting RCPT. Now it says "Previous (cached) callout
6370 verification failure".
6372 3. There was an off-by-one bug in the queryprogram router. An over-long
6373 return line was truncated at 256 instead of 255 characters, thereby
6374 overflowing its buffer with the terminating zero. As well as fixing this, I
6375 have increased the buffer size to 1024 (and made a note to document this).
6377 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6378 when Exim is waiting for an SMTP response from a remote server, Exim
6379 restarts its select() call on the socket, thereby resetting its timeout.
6380 This is not a problem when such interrupts are rare. Somebody set up a cron
6381 job to run exiwhat every 2 minutes, which is less than the normal select()
6382 timeout (5 or 10 minutes). This meant that the select() timeout never
6383 kicked in because it was always reset. I have fixed this by comparing the
6384 time when an interrupt arrives with the time at the start of the first call
6385 to select(). If more time than the timeout has elapsed, the interrupt is
6386 treated as a timeout.
6388 5. Some internal re-factoring in preparation for the addition of Sieve
6389 extensions (by MH). In particular, the "personal" test is moved to a
6390 separate function, and given an option for scanning Cc: and Bcc: (which is
6391 not set for Exim filters).
6393 6. When Exim created an email address using the login of the caller as the
6394 local part (e.g. when creating a From: or Sender: header line), it was not
6395 quoting the local part when it contained special characters such as @.
6397 7. Installed new OpenBSD configuration files.
6399 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6400 try to make them clearer.
6402 9. Callout options, other than the timeout value, were being ignored when
6403 verifying sender addresses in header lines. For example, when using
6405 verify = header_sender/callout=no_cache
6407 the cache was (incorrectly) being used.
6409 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6410 only the Cygwin environment.
6412 11. When return_path_on_delivery was set as a log selector, if different remote
6413 addresses in the same message used different return paths and parallel
6414 remote delivery occurred, the wrong values would sometimes be logged.
6415 (Whenever a remote delivery process finished, the return path value from
6416 the most recently started remote delivery process was logged.)
6418 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6419 lines when AUTH and/or TLS are in use. This is the "received protocol"
6420 field. Exim used to use "asmtp" for authenticated SMTP, without any
6421 indication (in the protocol name) for TLS use. Now it follows the RFC and
6422 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6423 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6424 names appear in log lines as well as in Received: header lines.
6426 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6427 extensions, and comparison tests, and to fix some bugs.
6429 14. Changes to the "personal" filter test:
6431 (1) The test was buggy in that it was just doing the equivalent of
6432 "contains" tests on header lines. For example, if a user's address was
6433 anne@some.where, the "personal" test would incorrectly be true for
6435 To: susanne@some.where
6437 This test is now done by extracting each address from the header in turn,
6438 and checking the entire address. Other tests that are part of "personal"
6439 are now done using regular expressions (for example, to check local parts
6440 of addresses in From: header lines).
6442 (2) The list of non-personal local parts in From: addresses has been
6443 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6444 taken from the Sieve specification recommendations.
6446 (3) If the message contains any header line starting with "List-" it is
6447 treated as non-personal.
6449 (4) The test for "circular" in the Subject: header line has been removed
6450 because it now seems ill-conceived.
6452 15. Minor typos in src/EDITME comments corrected.
6454 16. Installed latest exipick from John Jetmore.
6456 17. If headers_add on a router specified a text string that was too long for
6457 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6458 of string_sprintf() is now avoided.
6460 18. $message_body_size was not set (it was always zero) when running the DATA
6461 ACL and the local_scan() function.
6463 19. For the "mail" command in an Exim filter, no default was being set for
6464 the once_repeat time, causing a random time value to be used if "once" was
6465 specified. (If the value happened to be <= 0, no repeat happened.) The
6466 default is now 0s, meaning "never repeat". The "vacation" command was OK
6467 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6468 (I found it when inspecting the code).
6470 20. There is now an overall timeout for performing a callout verification. It
6471 defaults to 4 times the callout timeout, which applies to individual SMTP
6472 commands during the callout. The overall timeout applies when there is more
6473 than one host that can be tried. The timeout is checked before trying the
6474 next host. This prevents very long delays if there are a large number of
6475 hosts and all are timing out (e.g. when the network connections are timing
6476 out). The value of the overall timeout can be changed by specifying an
6477 additional sub-option for "callout", called "maxwait". For example:
6479 verify = sender/callout=5s,maxwait=20s
6481 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6482 to the end before writing, but this should make it even safer).
6484 22. Exim was forgetting that it had advertised PIPELINING for the second and
6485 subsequent messages on an SMTP connection. It was also not resetting its
6486 memory on STARTTLS and an internal HELO.
6488 23. When Exim logs an SMTP synchronization error within a session, it now
6489 records whether PIPELINING has been advertised or not.
6491 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6492 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6493 rather than long int.
6495 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6497 26. Added the never_mail option to autoreply.
6503 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6504 crash if the getsockname() call failed; this can happen if a connection is
6505 closed very soon after it is established. The problem was simply in the
6506 order in which certain operations were done, causing Exim to try to write
6507 to the SMTP stream before it had set up the file descriptor. The bug has
6508 been fixed by making things happen in the correct order.
6514 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6515 before the connection was closed, thus losing the rejection response.
6517 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6518 some early Solaris releases, but causes trouble in current releases where
6519 socklen_t is defined.
6521 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6524 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6527 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6528 that releasing the top of it at the end releases what was used for sub-
6529 expansions (unless the block got too big). However, discard this block if
6530 the first thing is a variable or header, so that we can use its block when
6531 it is dynamic (useful for very large $message_headers, for example).
6533 6. Lookups now cache *every* query, not just the most recent. A new, separate
6534 store pool is used for this. It can be recovered when all lookup caches are
6535 flushed. Lookups now release memory at the end of their result strings.
6536 This has involved some general refactoring of the lookup sources.
6538 7. Some code has been added to the store_xxx() functions to reduce the amount
6539 of flapping under certain conditions.
6541 8. log_incoming_interface used to affect only the <= reception log lines. Now
6542 it causes the local interface and port to be added to several more SMTP log
6543 lines, for example "SMTP connection from", and rejection lines.
6545 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6547 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6549 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6550 could be overwritten at the end of the current message (or the start of a
6551 new message if it was set in a HELO ACL). The value is now preserved for
6552 the duration of the SMTP connection.
6554 12. If a transport had a headers_rewrite setting, and a matching header line
6555 contained an unqualified address, that address was qualified, even if it
6556 did not match any rewriting rules. The underlying bug was that the values
6557 of the flags that permit the existence of unqualified sender and recipient
6558 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6559 non-local messages, and by -bnq for local messages) were not being
6560 preserved with the message after it was received.
6562 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6563 "next input=" as part of the text comprising the host identity instead of
6564 the correct text. The code was using the same buffer for two different
6565 strings. However, depending on which order the printing function evaluated
6566 its arguments, the bug did not always show up. Under Linux, for example, my
6567 test suite worked just fine.
6569 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6570 doesn't work with some older versions of Perl. It has been changed to "my",
6571 which in any case is probably the better facility to use.
6573 15. A really picky compiler found some instances of statements for creating
6574 error messages that either had too many or two few arguments for the format
6577 16. The size of the buffer for calls to the DNS resolver has been increased
6578 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6579 for addresses that have a lot of PTR records. This alleviates a problem; it
6580 does not fully solve it.
6582 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6583 buffer now truncates the list and logs the incident, which is the same
6584 action as happens when Exim is looking up a host name and its aliases.
6585 Previously in this situation something unpredictable would happen;
6586 sometimes it was "internal error: store_reset failed".
6588 18. If a server dropped the connection unexpectedly when an Exim client was
6589 using GnuTLS and trying to read a response, the client delivery process
6590 crashed while trying to generate an error log message.
6592 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6593 string, for example:
6595 warn message = H1: something\nH2: something
6597 the text was added as a single header line from Exim's point of view
6598 though it ended up OK in the delivered message. However, searching for the
6599 second and subsequent header lines using $h_h2: did not work. This has been
6600 fixed. Similarly, if a system filter added multiple headers in this way,
6601 the routers could not see them.
6603 20. Expanded the error message when iplsearch is called with an invalid key to
6604 suggest using net-iplsearch in a host list.
6606 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6607 ACLs are no longer actually imposed (and a message to that effect is
6610 22. If a "gecos" field in a passwd entry contained escaped characters, in
6611 particular, if it contained a \" sequence, Exim got it wrong when building
6612 a From: or a Sender: header from that name. A second bug also caused
6613 incorrect handling when an unquoted " was present following a character
6614 that needed quoting.
6616 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6617 was not being matched caselessly.
6619 24. Arranged for all hyphens in the exim.8 source to be escaped with
6622 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6623 properly. Recipient callout cache records were still being keyed to include
6624 the sender, even when use_sender was set false. This led to far more
6625 callouts that were necessary. The sender is no longer included in the key
6626 when use_sender is false.
6628 26. Added "control = submission" modifier to ACLs.
6630 27. Added the ${base62d: operator to decode base 62 numbers.
6632 28. dnsdb lookups can now access SRV records.
6634 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6635 the configuration file.
6637 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6638 (-v) mode. This makes the output for a verbose queue run more intelligible.
6640 31. Added a use_postmaster feature to recipient callouts.
6642 32. Added the $body_zerocount variable, containing the number of binary zero
6643 bytes in the message body.
6645 33. The time of last modification of the "new" subdirectory is now used as the
6646 "mailbox time last read" when there is a quota error for a maildir
6649 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6651 35. Added +ignore_unknown as a special item in host lists.
6653 36. Code for decoding IPv6 addresses in host lists is now included, even if
6654 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6655 address was recognized as an IP address, but was then not correctly decoded
6656 into binary, causing unexpected and incorrect effects when compared with
6663 1. Very minor rewording of debugging text in manualroute to say "list of
6664 hosts" instead of "hostlist".
6666 2. If verify=header_syntax was set, and a header line with an unqualified
6667 address (no domain) and a large number of spaces between the end of the
6668 name and the colon was received, the reception process suffered a buffer
6669 overflow, and (when I tested it) crashed. This was caused by some obsolete
6670 code that should have been removed. The fix is to remove it!
6672 3. When running in the test harness, delay a bit after writing a bounce
6673 message to get a bit more predictability in the log output.
6675 4. Added a call to search_tidyup() just before forking a reception process. In
6676 theory, someone could use a lookup in the expansion of smtp_accept_max_
6677 per_host which, without the tidyup, could leave open a database connection.
6679 5. Added the variables $recipient_data and $sender_data which get set from a
6680 lookup success in an ACL "recipients" or "senders" condition, or a router
6681 "senders" option, similar to $domain_data and $local_part_data.
6683 6. Moved the writing of debug_print from before to after the "senders" test
6686 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6687 problems for message scanning, either using a data ACL, or using
6688 local_scan() because the Received: header was not generated till after they
6689 were called (in order to set the time as the time of reception completion).
6690 I have revised the way this works. The header is now generated after the
6691 body is received, but before the ACL or local_scan() are called. After they
6692 are run, the timestamp in the header is updated.
6698 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6699 before starting a queue runner without re-exec. This happened only when
6700 deliver_drop_privilege was set or when the Exim user was set to root. The
6701 effect of the bug was that timeouts during subsequent deliveries caused
6702 crashes instead of being properly handled. The handler is now left at its
6703 default (and expected) setting.
6705 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6706 message, again when deliver_drop_privilege is set or Exim is run as root.
6707 The bug described in (1) was not present in this case, but the tidying up
6708 of the other signals was missing. I have made the two cases consistent.
6710 3. The ignore_target_hosts setting on a manualroute router was being ignored
6711 for hosts that were looked up using the /MX notation.
6713 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6716 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6717 operated on the sender address. After changing the $sender_address to <>
6718 for the sender address verify, Exim was re-instated it as the original
6719 (before rewriting) address, but remembering that it had rewritten it, so it
6720 wasn't rewriting it again. This bug also had the effect of breaking the
6721 sender address verification caching when the sender address was rewritten.
6723 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6724 This has been changed so that if the ip literal address matches
6725 ignore_target_hosts, the router declines.
6727 7. Added expansion conditions match_domain, match_address, and match_local_
6728 part (NOT match_host).
6730 8. The placeholder for the Received: header didn't have a length field set.
6732 9. Added code to Exim itself and to exim_lock to test for a specific race
6733 condition that could lead to file corruption when using MBX delivery. The
6734 issue is with the lockfile that is created in /tmp. If this file is removed
6735 after a process has opened it but before that process has acquired a lock,
6736 there is the potential for a second process to recreate the file and also
6737 acquire a lock. This could lead to two Exim processes writing to the file
6738 at the same time. The added code performs the same test as UW imapd; it
6739 checks after acquiring the lock that its file descriptor still refers to
6740 the same named file.
6742 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6743 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6746 11. Added the smtp_active_hostname option. If used, this will typically be made
6747 to depend on the incoming interface address. Because $interface_address is
6748 not set up until the daemon has forked a reception process, error responses
6749 that can happen earlier (such as "too many connections") no longer contain
6752 12. If an expansion in a condition on a "warn" statement fails because a lookup
6753 defers, the "warn" statement is abandoned, and the next ACL statement is
6754 processed. Previously this caused the whole ACL to be aborted.
6756 13. Added the iplsearch lookup type.
6758 14. Added ident_timeout as a log selector.
6760 15. Added tls_certificate_verified as a log selector.
6762 16. Added a global option tls_require_ciphers (compare the smtp transport
6763 option of the same name). This controls incoming TLS connections.
6765 17. I finally figured out how to make tls_require_ciphers do a similar thing
6766 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6767 before starting the TLS session.
6769 18. Tabs are now shown as \t in -bP output.
6771 19. If the log selector return_path_on_delivery was set, Exim crashed when
6772 bouncing a message because it had too many Received: header lines.
6774 20. If two routers both had headers_remove settings, and the first one included
6775 a superfluous trailing colon, the final name in the first list and the
6776 first name in the second list were incorrectly joined into one item (with a
6777 colon in the middle).
6783 1. Added -C and -D options to the exinext utility, mainly to make it easier
6784 to include in the automated testing, but these could be helpful when
6785 multiple configurations are in use.
6787 2. The exinext utility was not formatting the output nicely when there was
6788 an alternate port involved in the retry record key, nor when there was a
6789 message id as well (for retries that were specific to a specific message
6790 and a specific host). It was also confused by IPv6 addresses, because of
6791 the additional colons they contain. I have fixed the IPv4 problem, and
6792 patched it up to do a reasonable job for IPv6.
6794 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6795 delivery, the log line now contains "pipelined" if PIPELINING was used.
6797 4. An SMTP transport process used to panic and die if the bind() call to set
6798 an explicit outgoing interface failed. This has been changed; it is now
6799 treated in the same way as a connect() failure.
6801 5. A reference to $sender_host_name in the part of a conditional expansion
6802 that was being skipped was still causing a DNS lookup. This no longer
6805 6. The def: expansion condition was not recognizing references to header lines
6806 that used bh_ and bheader_.
6808 7. Added the _cache feature to named lists.
6810 8. The code for checking quota_filecount in the appendfile transport was
6811 allowing one more file than it should have been.
6813 9. For compatibility with Sendmail, the command line option
6821 and sets the incoming protocol and host name (for trusted callers). The
6822 host name and its colon can be omitted when only the protocol is to be set.
6823 Note the Exim already has two private options, -pd and -ps, that refer to
6824 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6825 "s", but I don't think that's a major issue.
6827 10. A number of refactoring changes to the code, none of which should affect
6830 (a) The number of logging options was getting close to filling up the
6831 32-bit word that was used as a bit map. I have split them into two classes:
6832 those that are passed in the argument to log_write(), and those that are
6833 only ever tested independently outside of that function. These are now in
6834 separate 32-bit words, so there is plenty of room for expansion again.
6835 There is no change in the user interface or the logging behaviour.
6837 (b) When building, for example, log lines, the code previously used a
6838 macro that called string_cat() twice, in order to add two strings. This is
6839 not really sufficiently general. Furthermore, there was one instance where
6840 it was actually wrong because one of the argument was used twice, and in
6841 one call a function was used. (As it happened, calling the function twice
6842 did not affect the overall behaviour.) The macro has been replaced by a
6843 function that can join an arbitrary number of extra strings onto a growing
6846 (c) The code for expansion conditions now uses a table and a binary chop
6847 instead of a serial search (which was left over from when there were very
6848 few conditions). Also, it now recognizes conditions like "pam" even when
6849 the relevant support is not compiled in: a suitably worded error message is
6850 given if an attempt is made to use such a condition.
6852 11. Added ${time_interval:xxxxx}.
6854 12. A bug was causing one of the ddress fields not to be passed back correctly
6855 from remote delivery subprocesses. The field in question was not being
6856 subsequently used, so this caused to problems in practice.
6858 13. Added new log selectors queue_time and deliver_time.
6860 14. Might have fixed a bug in maildirsizefile handling that threw up
6861 "unexpected character" debug warnings, and recalculated the data
6862 unnecessarily. In any case, I expanded the warning message to give more
6865 15. Added the message "Restricted characters in address" to the statements in
6866 the default ACL that block characters like @ and % in local parts.
6868 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6869 Three changes have been made:
6871 (a) There was a serious bug; a negative response to MAIL caused the whole
6872 recipient domain to be cached as invalid, thereby blocking all messages
6873 to all local parts at the same domain, from all senders. This bug has
6874 been fixed. The domain is no longer cached after a negative response to
6875 MAIL if the sender used is not empty.
6877 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6880 (c) A new callout option, "use_sender" has been added for people who want
6881 the modified behaviour.
6887 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6890 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6891 indeed breaks things for older releases.
6893 3. Added additional logging to the case where there is a problem reading data
6894 from a filter that is running in a subprocess using a pipe, in order to
6895 try to track down a specific problem.
6897 4. Testing facility fudge: when running in the test harness and attempting
6898 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6899 getting "No route to host". Convert this to a timeout.
6901 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6904 6. Some OS don't have socklen_t but use size_t instead. This affects the
6905 fifth argument of getsockopt() amongst other things. This is now
6906 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6907 can be set for individual OS. I have set it for SunOS5, OSF1, and
6908 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6909 some earlier ones do not.
6911 7. Change 4.30/15 was not doing the test caselessly.
6913 8. The standard form for an IPv6 address literal was being rejected by address
6914 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6915 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6916 this, as well as the form without the "IPv6" on the front (but only when
6917 address literals are enabled, of course).
6919 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6921 10. Exim crashed if a message with an empty sender address specified by -f
6922 encountered a router with an errors_to setting. This could be provoked only
6923 by a command such as
6927 where an empty string was supplied; "<>" did not hit this bug.
6929 11. Installed PCRE release 4.5.
6931 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6932 remained set. It is now erased.
6934 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6935 times from message ids (which are base 36 rather than the normal 62).
6937 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6938 were being counted as actual protocol errors, and logged if the log
6939 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6940 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6941 and DATA following a set of rejected RCPTs do not count as protocol errors.
6942 In other words, Exim assumes they were pipelined, though this may not
6943 actually be the case. Of course, in all cases the client gets an
6944 appropriate error code.
6946 15. If a lookup fails in an ACL condition, a message about the failure may
6947 be available; it is used if testing the ACL cannot continue, because most
6948 such messages specify what the cause of the deferral is. However, some
6949 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6950 that caused an old message to be retained and used if a later statement
6951 caused a defer, replacing the real cause of the deferral.
6953 16. If an IP address had so many PTR records that the DNS lookup buffer
6954 was not large enough to hold them, Exim could crash while trying to process
6955 the truncated data. It now detects and logs this case.
6957 17. Further to 4.21/58, another change has been made: if (and only if) the
6958 first line of a message (the first header line) ends with CRLF, a bare LF
6959 in a subsequent header line has a space inserted after it, so as not to
6960 terminate the header.
6962 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6963 unnecessarily, used atoi() instead of strtol(), and didn't check the
6964 termination when getting file sizes from file names by regex.
6966 19. Completely re-implemented the support for maildirsize files, in the light
6967 of a number of problems with the previous contributed implementation
6968 (4.30/29). In particular:
6970 . If the quota is zero, the maildirsize file is maintained, but no quota is
6973 . If the maildir directory does not exist, it is created before any attempt
6974 to write a maildirsize file.
6976 . The quota value in the file is just a cache; if the quota is changed in
6977 the transport, the new value overrides.
6979 . A regular expression is available for excluding directories from the
6982 20. The autoreply transport checks the characters in options that define the
6983 message's headers; it allows continued headers, but it was checking with
6984 isspace() after an embedded newline instead of explicitly looking for a
6987 21. If all the "regular" hosts to which an address was routed had passed their
6988 expiry times, and had not reached their retry times, the address was
6989 bounced, even if fallback hosts were defined. Now Exim should go on to try
6992 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6993 equivalent code in the SMTP transport. Some hosts send humungous responses
6994 to HELO/EHLO, more than 1024 it seems.
6996 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6997 gives compiler warnings in some environments. I've now done it "properly",
7000 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7001 (because of problems with the built-in one) was declared to return uschar *
7002 instead of char *, causing compiler failure.
7004 25. Fixed a file descriptor leak when processing alias/forward files.
7006 26. Fixed a minor format string issue in dbfn.c.
7008 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7010 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7011 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7014 29. When a sender address is verified, it is cached, to save repeating the test
7015 when there is more than one recipient in a message. However, when the
7016 verification involves a callout, it is possible for different callout
7017 options to be set for different recipients. It is too complicated to keep
7018 track of this in the cache, so now Exim always runs a verification when a
7019 callout is required, relying on the callout cache for the optimization.
7020 The overhead is duplication of the address routing, but this should not be
7023 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7024 to be verified with callout=postmaster, and the main callout worked but the
7025 postmaster check failed, the verification correctly failed. However, if a
7026 subsequent RCPT command asked for sender verification *without* the
7027 postmaster check, incorrect caching caused this verification also to fail,
7030 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7031 it was not caching the DNS options (qualify_single, search_parents) that
7032 were used when the lookup failed. A subsequent lookup with different
7033 options therefore always gave the same answer, though there were cases
7034 where it should not have. (Example: a "domains = !$mx_any" option on a
7035 dnslookup router: the "domains" option is always processed without any
7036 widening, but the router might have qualify_single set.) Now Exim uses the
7037 cached value only when the same options are set.
7039 32. Added John Jetmore's "exipick" utility to the distribution.
7041 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7042 than a timeout (e.g. a certificate is required, and is not provided), an
7043 Exim server now closes the connection immediately. Previously it waited for
7044 the client to close - but if the client is SSL, it seems that they each
7045 wait for each other, leading to a delay before one of them times out.
7047 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7048 maintained 0.8.x compatibility because I don't think many are using it, and
7049 it is clearly obsolete.
7051 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7054 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7055 client certificate was expired. A simple patch fixes this, though I don't
7056 understand the full logic of why the verify callback is called multiple
7059 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7060 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7061 which causes problems with some clients (such as the Certicom SSL Plus
7062 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7063 disables the coutermeasure allowing Eudora to connect."
7065 38. Exim was not checking that a write() to a log file succeeded. This could
7066 lead to Bad Things if a log got too big, in particular if it hit a file
7067 size limit. Exim now panics and dies if it cannot write to a log file, just
7068 as it does if it cannot open a log file.
7070 39. Modified OS/Makefile-Linux so that it now contains
7072 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7074 The two -D definitions ensure that Exim is compiled with large file
7075 support, which makes it possible to handle log files that are bigger than
7078 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7079 instance) a domain was checked against a named list that involved a lookup,
7080 causing $domain_data to be set, then another domain was checked against the
7081 same list, then the first domain was re-checked, the value of $domain_data
7082 after the final check could be wrong. In particular, if the second check
7083 failed, it could be set empty. This bug probably also applied to
7086 41. The strip_trailing_dot option was not being applied to the address given
7087 with the -f command-line option.
7089 42. The code for reading a message's header from the spool was incrementing
7090 $received_count, but never initializing it. This meant that the value was
7091 incorrect (doubled) while delivering a message in the same process in which
7092 it was received. In the most common configuration of Exim, this never
7093 happens - a fresh exec is done - but it can happen when
7094 deliver_drop_privilege is set.
7096 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7097 it now includes up to 150 characters of the unexpected data in the log
7100 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7101 and building data strings. The size of both of these buffers was 10 000
7102 bytes - far larger than anybody would *ever* want, thought I. Needless to
7103 say, somebody hit the limit. I have increased the maximum line length to
7104 20 000 and the maximum data length of concatenated lines to 100 000. I have
7105 also fixed two bugs, because there was no checking on these buffers. Tsk,
7106 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7107 buffer is too small.
7109 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7110 lsearch lookups. Now it does.
7112 46. When parsing a route_list item in a manualroute router, a fixed-length
7113 buffer was used for the list of hosts. I made this 1024 bytes long,
7114 thinking that nobody would ever have a list of hosts that long. Wrong.
7115 Somebody had a whole pile of complicated expansion conditions, and the
7116 string was silently truncated, leading to an expansion error. It turns out
7117 that it is easier to change to an unlimited length (owing to other changes
7118 that have happened since this code was originally written) than to build
7119 structure for giving a limitation error. The length of the item that
7120 expands into the list of hosts is now unlimited.
7122 47. The lsearch lookup could not handle data where the length of text line was
7123 more than 4095 characters. Such lines were truncated, leading to shortened
7124 data being returned. It should now handle lines of any length.
7126 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7127 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7130 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7131 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7132 no longer happens in comment lines. A list of replacements is now placed
7133 at the head of all of the source files, except those whose only change is
7134 to replace PERL_COMMAND in the very first #! line.
7136 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7137 messages on the queue, with a bottom-up merge sort, using code contributed
7138 by Michael Haardt. This should make operations like -bp somewhat faster on
7139 large queues. It won't affect queue runners, except when queue_run_in_order
7142 51. Installed eximstats 1.31 in the distribution.
7144 52. Added support for SRV lookups to the dnslookup router.
7146 53. If an ACL referred to $message_body or $message_body_end, the value was not
7147 reset for any messages that followed in the same SMTP session.
7149 54. The store-handling optimization for building very long strings was not
7150 differentiating between the different store pools. I don't think this
7151 actually made any difference in practice, but I've tidied it.
7153 55. While running the routers to verify a sender address, $sender_address
7154 was still set to the sender address. This is wrong, because when routing to
7155 send a bounce to the sender, it would be empty. Therefore, I have changed
7156 it so that, while verifying a sender address, $sender_address is set to <>.
7157 (There is no change to what happens when verifying a recipient address.)
7159 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7160 target A or AAAA records (if not already returned) without resetting the
7161 qualify_single or search_parents options of the DNS resolver. These are
7162 inappropriate in this case because the targets of MX and SRV records must
7163 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7164 target that, when qualified, matched something in the local domain. These
7165 two options are now turned off when doing these lookups.
7167 57. It seems that at least some releases of Reiserfs (which does not have the
7168 concept of a fixed number of inodes) returns zero and not -1 for the
7169 number of available inodes. This interacted badly with check_spool_inodes,
7170 which assumed that -1 was the "no such thing" setting. What I have done is
7171 to check that the total number of inodes is greater than zero before doing
7172 the test of how many are available.
7174 58. When a "warn" ACL statement has a log_message modifier, the message is
7175 remembered, and not repeated. This is to avoid a lot of repetition when a
7176 message has many recipients that cause the same warning to be written.
7177 However, Exim was preserving the list of already written lines for an
7178 entire SMTP session, which doesn't seem right. The memory is now reset if a
7179 new message is started.
7181 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7182 debugging output unless log_rewrite was also set.
7184 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7185 of (int)(handle) when we know that handle contains (void *)(-1).
7187 61. The Exim daemon panic-logs an error return when it closes the incoming
7188 connection. However "connection reset by peer" seems to be common, and
7189 isn't really an error worthy of noting specially, so that particular error
7192 62. When Exim is trying to find all the local interfaces, it used to panic and
7193 die if the ioctl to get the interface flags failed. However, it seems that
7194 on at least one OS (Solaris 9) it is possible to have an interface that is
7195 included in the list of interfaces, but for which you get a failure error
7196 for this call. This happens when the interface is not "plumbed" into a
7197 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7198 failure of the "get flags" call assumes that the interface is down.
7200 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7201 makes life easier for people who are doing arithmetic on fields extracted
7202 from dates, where you often get leading zeros that should not be
7203 interpreted as octal.
7205 64. Added qualify_domain to the redirect router, to override the global
7208 65. If a pathologically long header line contained very many addresses (the
7209 report of this problem mentioned 10 000) and each of them was rewritten,
7210 Exim could use up a very large amount of memory. (It kept on making new
7211 copies of the header line as it rewrote, and never released the old ones.)
7212 At the expense of a bit more processing, the header rewriting function has
7213 been changed so that it no longer eats memory in this way.
7215 66. The generation of the Received: header has been moved from the time that a
7216 message starts to be received, to the time that it finishes. The timestamp
7217 in the Received: header should now be very close to that of the <= log
7218 line. There are two side-effects of this change:
7220 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7221 logged header lines no longer include the local Received: line, because
7222 it has not yet been created. The same applies to a copy of the message
7223 that is returned to a non-SMTP sender when a message is rejected.
7225 (b) When a filter file is tested using -bf, no additional Received: header
7226 is added to the test message. After some thought, I decided that this
7229 This change does not affect the value of $received_for. It is still set
7230 after address rewriting, but before local_scan() is called.
7232 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7234 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7235 gave an unhelpful panic error message, and a defer error. I have managed to
7236 change this behaviour so that it now rejects any supplied certificate,
7237 which seems right, as the list of acceptable certificates is empty.
7239 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7240 gave an unhelpful defer error. I have not managed to make this reject any
7241 supplied certificates, but the error message it gives is "no certificate
7242 supplied", which is not helpful.
7244 70. exigrep's output now also includes lines that are not associated with any
7245 message, but which match the given pattern. Implemented by a patch from
7246 Martin Sluka, which also tidied up the Perl a bit.
7248 71. Recipient callout verification, like sender verification, was using <> in
7249 the MAIL FROM command. This isn't really the right thing, since the actual
7250 sender may affect whether the remote host accepts the recipient or not. I
7251 have changed it to use the actual sender in the callout; this means that
7252 the cache record is now keyed on a recipient/sender pair, not just the
7253 recipient address. There doesn't seem to be a real danger of callout loops,
7254 since a callout by the remote host to check the sender would use <>.
7255 [SEE ABOVE: changed after hitting problems.]
7257 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7258 temporary errors. However, in the case of such a code being given after
7259 the end of a data transmission (i.e. after ".") Exim was failing to write
7260 a retry record for the message. (Yes, there was some broken host that was
7261 actually sending 8xx at this point.)
7263 73. An unknown lookup type in a host list could cause Exim to panic-die when
7264 the list was checked. (An example that provoked this was putting <; in the
7265 middle of a list instead of at the start.) If this happened during a DATA
7266 ACL check, a -D file could be left lying around. This kind of configuration
7267 error no longer causes Exim to die; instead it causes a defer error. The
7268 incident is still logged to the main and panic logs.
7270 74. Buglet left over from Exim 3 conversion. The message "too many messages
7271 in one connection" was written to the rejectlog but not the mainlog, except
7272 when address rewriting (yes!) was being logged.
7274 75. Added write_rejectlog option.
7276 76. When a system filter was run not as root (that is, when system_filter_user
7277 was set), the values of the $n variables were not being returned to the
7278 main process; thus, they were not subsequently available in the $sn
7281 77. Added +return_path_on_delivery log selector.
7283 78. A connection timeout was being treated differently from recipients deferred
7284 when testing hosts_max_try with a message that was older than the host's
7285 retry timeout. (The host should not be counted, thus allowing all hosts to
7286 be tried at least once before bouncing.) This may have been the cause of an
7287 occasionally reported bug whereby a message would remain on the queue
7288 longer than the retry timeout, but would be bounced if a delivery was
7289 forced. I say "may" because I never totally pinned down the problem;
7290 setting up timeout/retry tests is difficult. See also the next item.
7292 79. The ultimate address timeout was not being applied to errors that involved
7293 a combination of host plus message (for example, a timeout on a MAIL
7294 command). When an address resolved to a number of possible hosts, and they
7295 were not all tried for each delivery (e.g. because of hosts_max_try), a
7296 message could remain on the queue longer than the retry timeout.
7298 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7301 81. Fixed an obscure SMTP outgoing bug which required at least the following
7302 conditions: (a) there was another message waiting for the same server;
7303 (b) the server returned 5xx to all RCPT commands in the first message so
7304 that the message was not completed; (c) the server dropped the connection
7305 or gave a negative response to the RSET that Exim sends to abort the
7306 transaction. The observed case was a dropped connection after DATA that had
7307 been sent in pipelining mode. That is, the server had advertised PIPELINING
7308 but was not implementing it correctly. The effect of the bug was incorrect
7309 behaviour, such as trying another host, and this could lead to a crash.
7315 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7316 and daemon.c were passed as pointers to ints; they should have been
7317 pointers to socklen_t variables (which are typically unsigned ints).
7319 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7322 3. Fixed a really odd bug that affected only the testing scheme; patching a
7323 certain fixed string in the binary changed the value of another string that
7324 happened to be identical to the end of the original first string.
7326 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7327 name", it returns that address as the IP address. On some operating
7328 systems (e.g. Solaris), it also passes back the IP address string as the
7329 "host name". However, on others (e.g. Linux), it passes back an empty
7330 string. Exim wasn't checking for this, and was changing the host name to an
7331 empty string, assuming it had been canonicalized.
7333 5. Although rare, it is permitted to have more than one PTR record for a given
7334 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7335 all the names associated with an address, because they do in Solaris.
7336 However, it seems that they do not in Linux for data that comes from the
7337 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7338 I found this out when I moved to a new Linux workstation and tried to run
7339 the Exim test suite.
7341 To get round this problem I have changed the code so that it now does its
7342 own call to the DNS to look up PTR records when searching for a host name.
7343 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7344 addresses that are only in /etc/hosts are still found.
7346 This behaviour is, however, controlled by an option called host_lookup_
7347 order, which defaults to "bydns:byaddr". If people want to use the other
7348 order, or indeed, just use one or the other means of lookup, they can
7349 specify it in this variable.
7351 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7352 some operating systems, this comes back from gethostbyaddr() as an empty
7353 string, and this is what Exim used to test for. However, it seems that in
7354 other systems, "." is yielded. Exim now tests for this case too.
7356 7. The values of check_spool_space and check_log_space are now held internally
7357 as a number of kilobytes instead of an absolute number of bytes. If a
7358 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7359 kilobyte. This means that much larger values can be stored.
7361 8. Exim monitor: an attempt to get the action menu when not actually pointing
7362 at a message produces an empty menu entitled "No message selected". This
7363 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7364 no entries in it ("Shell widget menu has zero width and/or height"). So I
7365 have added a single, blank menu entry in this case.
7367 9. Added ${quote_local_part.
7369 10. MIME decoding is now applied to the contents of Subject: header lines when
7372 11. Now that a reference to $sender_host_address automatically causes a reverse
7373 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7374 host lookup before query-style lookups in lists that might use this
7375 variable. This has therefore been abolished, and the "net-" prefix is no
7376 longer necessary for query-style lookups.
7378 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7379 have been a typo for LFLAGS, so it has been changed.
7381 13. The install script calls Exim with "-C /dev/null" in order to find the
7382 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7383 to be output. However, since Exim outputs its version number before the
7384 error, it didn't break the script. It just looked ugly. I fixed this by
7385 always allowing "-C /dev/null" if the caller is root.
7387 14. Ignore overlarge ACL variable number when reading spool file - insurance
7388 against a later release with more variables having written the file.
7390 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7391 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7392 this, as well as the form without the "IPv6" on the front.
7394 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7395 OS/Makefile-Darwin file.
7397 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7398 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7400 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7401 variable. However, some OS use ints for inodes. I've added cast to long int
7402 to get rid of the compiler warning.
7404 19. I had forgotten to lock out "/../" in configuration file names when
7405 ALT_CONFIG_PREFIX was set.
7407 20. Routers used for verification do not need to specify transports. However,
7408 if such a router generated a host list, and callout was configured, Exim
7409 crashed, because it could not find a port number from the (non-existent)
7410 transport. It now assumes port 25 in this circumstance.
7412 21. Added the -t option to exigrep.
7414 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7415 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7416 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7417 them is set, LOOKUP_LSEARCH is forced.
7419 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7421 24. Added sender and host information to the "rejected by local_scan()" log
7422 line; previously there was no indication of these.
7424 25. Added .include_if_exists.
7426 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7427 Linux. It turns out that not all file systems support this. Apparently some
7428 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7429 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7430 not supported on the file descriptor, is now ignored when Exim is trying to
7431 sync a directory. This applies only to Linux.
7433 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7435 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7436 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7439 29. Added support for maildirsize files from supplied patch (modified a bit).
7441 30. The use of :fail: followed by an empty string could lead Exim to respond to
7442 sender verification failures with (e.g.):
7444 550 Verification failed for <xxx>
7445 550 Sender verify failed
7447 where the first response line was missing the '-' that indicates it is not
7448 the final line of the response.
7450 31. The loop for finding the name of the user that called Exim had a hardwired
7451 limit of 10; it now uses the value of finduser_retries, which is used for
7452 all other user lookups.
7454 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7457 33. Exim was neglecting to zero errno before one call of strtol() when
7458 expanding a string and expecting an integer value. On some systems this
7459 resulted in spurious "integer overflow" errors. Also, it was casting the
7460 result into an int without checking.
7462 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7463 did not work. The code looks as if it has *never* worked, though it appears
7464 to have been documented since at least release 1.62. I have made it work.
7466 35. The "timeout_DNS" error in retry rules, also documented since at least
7467 1.62, also never worked. As it isn't clear exactly what this means, and
7468 clearly it isn't a major issue, I have abolished the feature by treating it
7469 as "timeout", and writing a warning to the main and panic logs.
7471 36. The display of retry rules for -brt wasn't always showing the error code
7474 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7475 timeout_connect_A, timeout_connect_MX.
7477 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7478 to the empty sender.
7480 39. The daemon was not analysing the content of -oX till after it had closed
7481 stderr and disconnected from the controlling terminal. This meant that any
7482 syntax errors were only noted on the panic log, and the return code from
7483 the command was 0. By re-arranging the code a little, I've made the
7484 decoding happen first, so such errors now appear on stderr, and the return
7485 code is 1. However, the actual setting up of the sockets still happens in
7486 the disconnected process, so errors there are still only recorded on the
7489 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7490 connections (as happens on some IP stacks) was logged at start up time as
7491 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7492 it from "IPv6 and IPv4", which means that two separate sockets are being
7495 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7496 says whether AF_INET or AF_INET6 was passed as an argument.
7498 42. Exiwhat output was messed up when time zones were included in log
7501 43. Exiwhat now gives more information about the daemon's listening ports,
7502 and whether -tls-on-connect was used.
7504 44. The "port" option of the smtp transport is now expanded.
7506 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7507 silently ignored. Now an error message is written to the main and panic
7510 46. There's a new ACL modifier called "logwrite" which writes to a log file
7511 as soon as it is encountered.
7513 47. Added $local_user_uid and $local_user_gid at routing time.
7515 48. Exim crashed when trying to verify a sender address that was being
7518 49. Exim was recognizing only a space character after ".include". It now also
7519 recognizes a tab character.
7521 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7522 extracting the relevant information from the specification. The man page no
7523 longer contains scrambled data for the -d option, and I've added a section
7524 at the front about calling Exim under different names.
7526 51. Added "extra_headers" argument to the "mail" command in filter files.
7528 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7531 53. Installed eximstats 1.29.
7533 54. Added transport_filter_timeout as a generic transport option.
7535 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7536 Cc: header lines. This was required by RFC 822, but it not required by RFC
7539 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7540 incoming messages that did not have them. Now it does so only if the
7541 message originates locally, that is, if there is no associated remote host
7542 address. When Resent- header lines are present, this applies to the Resent-
7543 lines rather than the non-Resent- lines.
7545 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7546 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7548 58. Messages for configuration errors now include the name of the main
7549 configuration file - useful now that there may be more than one file in a
7550 list (.included file names were always shown).
7552 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7553 for those rare installations that do not start the daemon as root or run it
7554 setuid root. I've cut out the call to initgroups() if the daemon is not
7557 60. The Exim user and group can now be bound into the binary as text strings
7558 that are looked up at the start of Exim's processing.
7560 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7562 62. Added $mailstore_basename variable.
7564 63. Installed patch to sieve.c from Michael Haardt.
7566 64. When Exim failed to open the panic log after failing to open the main log,
7567 the original message it was trying to log was written to stderr and debug
7568 output, but if they were not available (the usual case in production), it
7569 was lost. Now it is written to syslog before the two lines that record the
7570 failures to open the logs.
7572 65. Users' Exim filters run in subprocesses under the user's uid. It is
7573 possible for a "deliver" command or an alias in a "personal" command to
7574 provoke an address rewrite. If logging of address rewriting is configured,
7575 this fails because the process is not running as root or exim. There may be
7576 a better way of dealing with this, but for the moment (because 4.30 needs
7577 to be released), I have disabled address rewrite logging when running a
7578 filter in a non-root, non-exim process.
7584 1. The buildconfig auxiliary program wasn't quoting the value set for
7585 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7586 not defined. This bug was masked in 4.22 by the effect that was fixed in
7589 2. Some messages that were rejected after a message id was allocated were
7590 shown as "incomplete" by exigrep. It no longer does this for messages that
7591 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7593 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7594 have allow_domain_literals set, the ID did not get logged in the <= line.
7595 Domain literals are now always recognized in Message-ID: header lines.
7597 4. The first argument for a ${extract expansion item is the key name or field
7598 number. Leading and trailing spaces in this item were not being ignored,
7599 causing some misleading effects.
7601 5. When deliver_drop_privilege was set, single queue runner processes started
7602 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7603 same command in the process it spins off) were not dropping privilege.
7605 6. When the daemon running as "exim" started a queue runner, it always
7606 re-executed Exim in the spun-off process. This is a waste of effort when
7607 deliver_drop_privilege is set. The new process now just calls the
7608 queue-runner function directly.
7614 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7617 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7618 ignored. Though the use of -oP was forcing the writing of a pid file, it
7619 was always written to the default place.
7621 3. If the message "no IP address found for host xxxx" is generated during
7622 incoming verification, it is now followed by identification of the incoming
7623 connection (so you can more easily find what provoked it).
7625 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7627 5. Added some features to "harden" Exim a bit more against certain attacks:
7629 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7630 be put in Local/Makefile. This is like the never_users runtime option,
7631 but it cannot be overridden. The default setting is "root".
7633 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7634 prefix string with which any file named in a -C command line option
7637 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7638 is retained for -C and -D only if the caller of Exim is root. Without
7639 it, the exim user may also use -C and -D and retain privilege.
7641 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7642 command line option is disabled.
7644 6. Macro names set by the -D option must start with an upper case letter, just
7645 like macro names defined in the configuration file.
7647 7. Added "dereference=" facility to LDAP.
7649 8. Two instances of the typo "uknown" in the source files are fixed.
7651 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7652 the Configure-Makefile script screwed up while processing it.
7654 10. Incorporated PCRE 4.4.
7656 11. The SMTP synchronization check was not operating right at the start of an
7657 SMTP session. For example, it could not catch a HELO sent before the client
7658 waited for the greeting. There is now a check for outstanding input at the
7659 point when the greeting is written. Because of the duplex, asynchronous
7660 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7661 way, but not yet received, when the check is performed.
7663 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7664 on TCP/IP sockets, because this apparently causes some broken clients to
7667 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7668 unchanged) from the Cygwin maintainer.
7670 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7671 was supported instead of testing for mbx. Effectively a typo.
7673 15. The spa authenticator server code was not checking that the input it
7674 received was valid base64.
7676 16. The debug output line for the "set" modifier in ACLs was not showing the
7677 name of the variable that was being set.
7679 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7681 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7682 lookup on its own. Something else was needed to trigger the lookup. For
7683 example, a match in host_lookup or the need for a host name in a host list.
7684 Now, if $sender_host_name is referenced and the host name has not yet been
7685 looked up, a lookup is performed. If the lookup fails, the variable remains
7686 empty, and $host_lookup_failed is set to "1".
7688 19. Added "eqi" as a case-independent comparison operator.
7690 20. The saslauthd authentication condition could segfault if neither service
7691 nor realm was specified.
7693 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7694 error message that was logged was misleading, and incoming SMTP
7695 connections were dropped. The message is now more accurate, and temporary
7696 errors are given to SMTP connections.
7698 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7699 (except RSET) with a 421 error, until QUIT is received. However, it was
7700 failing to send a response to QUIT.
7702 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7703 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7704 ACL is now run after the helo_try_verify_hosts code.
7706 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7707 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7708 case-independent, so other case variants are also recognized). Apparently
7709 some systems use these upper case variants.
7711 25. If more than two messages were waiting for the same host, and a transport
7712 filter was specified for the transport, Exim sent two messages over the
7713 same TCP/IP connection, and then failed with "socket operation on non-
7714 socket" when it tried to send the third.
7716 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7718 27. The extern definition of crypt16() in expand.c was not being excluded when
7719 the OS had its own crypt16() function.
7721 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7722 as a preferred synonym for return_size_limit, both as an option and as an
7725 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7727 30. Changed the default configuration ACL to relax the local part checking rule
7728 for addresses that are not in any local domains. For these addresses,
7729 slashes and pipe symbols are allowed within local parts, but the sequence
7730 /../ is explicitly forbidden.
7732 31. SPA server authentication was not clearing the challenge buffer before
7735 32. log_message in a "warn" ACL statement was writing to the reject log as
7736 well as to the main log, which contradicts the documentation and doesn't
7737 seem right (because no rejection is happening). So I have stopped it.
7739 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7740 However, I am unable to do any testing of this.
7742 34. Fixed an infelicity in the appendfile transport. When checking directories
7743 for a mailbox, to see if any needed to be created, it was accidentally
7744 using path names with one or more superfluous leading slashes; tracing
7745 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7747 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7748 contents are added to the log line that is written for every discarded
7749 recipient. (Previously a log_message setting was ignored.)
7751 36. The ${quote: operator now quotes the string if it is empty.
7753 37. The install script runs exim in order to find its version number. If for
7754 some reason other than non-existence or emptiness, which it checks, it
7755 could not run './exim', it was installing it with an empty version number,
7756 i.e. as "exim-". This error state is now caught, and the installation is
7759 38. An argument was missing from the function that creates an error message
7760 when Exim fails to connect to the socket for saslauthd authentication.
7761 This could cause Exim to crash, or give a corrupted message.
7763 39. Added isip, isip4, and isip6 to ${if conditions.
7765 40. The ACL variables $acl_xx are now saved with the message, and can be
7766 accessed later in routers, transports, and filters.
7768 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7769 strings in the file are not string-expanded.
7771 42. If a MAIL command specified a SIZE value that was too large to fit into an
7772 int variable, the check against message_size_limit failed. Such values are
7773 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7774 day this will have to be increased, but I don't think I want to be around
7775 when emails are that large.
7782 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7783 iconv() is not standard in FreeBSD.
7785 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7786 IPv6 enabled. The observed symptom was a segmentation fault on return from
7787 the function os_common_find_running_interfaces() in src/os.c.
7789 3. In the check_special_case() function in daemon.c I had used "errno" as an
7790 argument name, which causes warnings on some systems. This was basically a
7791 typo, since it was named "eno" in the comments!
7793 4. The code that waits for the clock to tick (at a resolution of some fraction
7794 of a second) so as to ensure message-id uniqueness was always waiting for
7795 at least one whole tick, when it could have waited for less. [This is
7796 almost certainly not relevant at current processor speeds, where it is
7797 unlikely to ever wait at all. But we try to future-proof.]
7799 5. The function that sleeps for a time interval that includes fractions of a
7800 second contained a race. It did not block SIGALRM between setting the
7801 timer, and suspending (a couple of lines later). If the interval was short
7802 and the sigsuspend() was delayed until after it had expired, the suspension
7803 never ended. On busy systems this could lead to processes getting stuck for
7806 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7807 process, before it forks any delivery processes. The open lookup caching
7808 mechanism meant that the open file or database connection was passed into
7809 the delivery process. The problem was that delivery processes always tidy
7810 up cached lookup data. This could cause a problem for the next delivery
7811 process started by the queue runner, because the external queue runner
7812 process does not know about the closure. So the next delivery process
7813 still has data in the lookup cache. In the case of a file lookup, there was
7814 no problem because closing a file descriptor in a subprocess doesn't affect
7815 the parent. However, if the lookup was caching a connection to a database,
7816 the connection was closed, and the second delivery process was likely to
7817 see errors such as "PGSQL: query failed: server closed the connection
7818 unexpectedly". The problem has been fixed by closing all cached lookups
7819 in a queue runner before running a delivery process.
7821 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7822 seem to have the "const" qualifier which it has on other OS. I've
7825 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7826 *of the same type* (client or server) with the same public name that an
7827 error should be diagnosed.
7829 9. When Exim looked up a host name for an IP address, but failed to find the
7830 original IP address when looking up the host name (a safety check), it
7831 output the message "<ip address> does not match any IP for NULL", which was
7832 confusing, to say the least. The bug was that the host name should have
7833 appeared instead of "NULL".
7835 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7836 user that is built into the binary, and the -C or -D options is used, root
7837 privilege is dropped before the configuration file is read. In addition,
7838 logging is switched to stderr instead of the normal log files. If the
7839 configuration then re-defines the Exim user, the unprivileged environment
7840 is probably not what is expected, so Exim logs a panic warning message (but
7843 However, if deliver_drop_privilege is set, the unprivileged state may well
7844 be exactly what is intended, so the warning has been cut out in that case,
7845 and Exim is allowed to try to write to its normal log files.
7851 1. smtp_return_error_details was not giving details for temporary sender
7852 or receiver verification errors.
7854 2. Diagnose a configuration error if two authenticators have the same public
7857 3. Exim used not to create the message log file for a message until the first
7858 delivery attempt. This could be confusing when incoming messages were held
7859 for policy or load reasons. The message log file is now created at the time
7860 the message is received, and an initial "Received" line is written to it.
7862 4. The automatically generated man page for command line options had a minor
7863 bug that caused no ill effects; however, a more serious problem was that
7864 the procedure for building the man page automatically didn't always
7865 operate. Consequently, release 4.20 contains an out-of-date version. This
7866 shouldn't happen again.
7868 5. When building Exim with embedded Perl support, the script that builds the
7869 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7870 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7872 6. The freeze_tell option was not being used for messages that were frozen on
7873 arrival, either by an ACL or by local_scan().
7875 7. Added the smtp_incomplete_transaction log selector.
7877 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7878 was accepting AUTH without a new EHLO.
7880 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7881 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7884 10. Logging of TCP/IP connections (when configured) now happens in the main
7885 daemon process instead of the child process, so that the TCP/IP connection
7886 count is more accurate (but it can never be perfect).
7888 11. The use of "drop" in a nested ACL was not being handled correctly in the
7889 outer ACL. Now, if condition failure induced by the nested "drop" causes
7890 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7891 or "require"), the connection is dropped.
7893 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7894 that yield "discard" can now be used with an "accept" or a "discard" verb,
7895 but an error is generated for any others (because I can't see a useful way
7896 to define what should happen).
7898 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7899 are now processed in the same way as lines in the Exim configuration file.
7900 In particular, continuation lines are supported.
7902 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7904 15. Added -ti meaning -t -i.
7906 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7907 domains, and warn by logging if others are found.
7909 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7910 returned from GIFCONF when Exim is trying to find the list of interfaces on
7911 a host. The code in os.c has been modified to copy each ifreq to an aligned
7912 structure in all cases.
7914 Also, in some cases, the returned ifreq's were being copied to a 'struct
7915 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7916 means the last couple of bytes of an IPv6 address could be chopped if the
7917 ifreq contained only a normal sockaddr (14 bytes storage).
7919 18. Named domain lists were not supported in the hosts_treat_as_local option.
7920 An entry such as +xxxx was not recognized, and was treated as a literal
7923 19. Ensure that header lines added by a DATA ACL are included in the reject log
7924 if the ACL subsequently rejects the message.
7926 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7927 MD5 (which is deprecated).
7929 21. When testing a filter file using -bf, Exim was writing a message when it
7930 took the sender from a "From " line in the message, but it was not doing so
7931 when it took $return_path from a Return-Path: header line. It now does.
7933 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7934 with a valid header line field name (a series of printing characters
7935 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7937 23. Changed "disc" in the source to "disk" to conform to the documentation and
7938 the book and for uniformity.
7940 24. Ignore Sendmail's -Ooption=value command line item.
7942 25. When execve() failed while trying to run a command in a pipe transport,
7943 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7944 could be confused with a return value of 69 from the command itself. This
7945 has been changed to 127, the value the shell returns if it is asked to run
7946 a non-existent command. The wording for the related log line suggests a
7947 non-existent command as the problem.
7949 26. If received_header_text expands to an empty string, do not add a Received:
7950 header line to the message. (Well, it adds a token one on the spool, but
7951 marks it "old" so that it doesn't get used or transmitted.)
7953 27. Installed eximstats 1.28 (addition of -nt option).
7955 28. There was no check for failure on the call to getsockname() in the daemon
7956 code. This can fail if there is a shortage of resources on the system, with
7957 ENOMEM, for example. A temporary error is now given on failure.
7959 29. Contrary to the C standard, it seems that in some environments, the
7960 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7961 program. Exim now does this explicitly; it affects the formatting of
7962 timestamps using strftime().
7964 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7965 complaints. I've now initialized all the variables, to avoid this.
7967 32. Header lines added by a system filter were not being "seen" during
7968 transport-time rewrites.
7970 33. The info_callback() function passed to OpenSSL is set up with type void
7971 (*)(SSL *, int, int), as described somewhere. However, when calling the
7972 function (actually a macro) that sets it up, the type void(*)() is
7973 expected. I've put in a cast to prevent warnings from picky compilers.
7975 34. If a DNS black list lookup found a CNAME record, but there were no A
7976 records associated with the domain it pointed at, Exim crashed.
7978 35. If a DNS black list lookup returned more than one A record, Exim ignored
7979 all but the first. It now scans all returned addresses if a particular IP
7980 value is being sought. In this situation, the contents of the
7981 $dnslist_value variable are a list of all the addresses, separated by a
7984 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7985 to accept a host name and all its aliases if the forward lookup for any of
7986 them yielded the IP address of the incoming connection. Now it accepts only
7987 those names whose forward lookup yields the correct IP address. Any other
7988 names are discarded. This closes a loophole whereby a rogue DNS
7989 administrator could create reverse DNS records to break through a
7990 wildcarded host restriction in an ACL.
7992 37. If a user filter or a system filter that ran in a subprocess used any of
7993 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7994 the wrong values were passed to the pipe command ($thisaddress had the
7995 value of $0, $0 had the value of $1, etc). This bug was introduced by
7996 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7998 38. Improved the line breaking for long SMTP error messages from ACLs.
7999 Previously, if there was no break point between 40 and 75 characters, Exim
8000 left the rest of the message alone. Two changes have been made: (a) I've
8001 reduced the minimum length to 35 characters; (b) if it can't find a break
8002 point between 35 and 75 characters, it looks ahead and uses the first one
8003 that it finds. This may give the occasional overlong line, but at least the
8004 remaining text gets split now.
8006 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8007 file descriptors might be low, and that setting 1000 would always raise it.
8008 It turns out that in some environments, the limit is already over 1000 and
8009 that lowering it causes trouble. So now Exim takes care not to decrease it.
8011 40. When delivering a message, the value of $return_path is set to $sender_
8012 address at the start of routing (routers may change the value). By an
8013 oversight, this default was not being set up when an address was tested by
8014 -bt or -bv, which affected the outcome if any router or filter referred to
8017 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8018 log only when "message" or "log_message" are set. However, if one of the
8019 conditions was an address verification, or a call to a nested ACL, the
8020 messages generated by the underlying test were being passed through. This
8021 no longer happens. The underlying message is available in $acl_verify_
8022 message for both "message" and "log_message" expansions, so it can be
8023 passed through if needed.
8025 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8026 new expansion $bh_ to give the encoded byte string without charset
8027 translation. Translation happens only if iconv() is available; HAVE_ICONV
8028 indicates this at build time. HEADERS_CHARSET gives the charset to
8029 translate to; headers_charset can change it in the configuration, and
8030 "headers charset" can change it in an individual filter file.
8032 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8033 that creates RFC 2047 encoded "words" labels them as that charset instead
8034 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8035 expansion operator; (ii) when Exim creates a From: line for a local
8036 message; (iii) when a header line is rewritten to include a "phrase" part.
8038 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8039 buggy, causing it to skip the first lines of messages whose message ID
8040 ended in 'D'. This would not have bitten before Exim release 4.14, because
8041 message IDs were unlikely to end in 'D' before then. The effect was to have
8042 incorrect size information for certain domains.
8044 45. #include "config.h" was missing at the start of the crypt16.c module. This
8045 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8048 46. If there was a timeout during a "random" callout check, Exim treated it as
8049 a failure of the random address, and carried on sending RSET and the real
8050 address. If the delay was just some slowness somewhere, the response to the
8051 original RCPT would be taken as a response to RSET and so on, causing
8052 mayhem of various kinds.
8054 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8055 when I implemented it. It didn't allow for the fact that some option values
8056 may legitimately be negative (e.g. size_addition), and it didn't even do
8057 the right test for positive values.
8059 48. Domain names in DNS records are case-independent. Exim always looks them up
8060 in lower case. Some resolvers return domain names in exactly the case they
8061 appear in the zone file, that is, they may contain uppercase letters. Not
8062 all resolvers do this - some return always lower case. Exim was treating a
8063 change of case by a resolver as a change of domain, similar to a widening
8064 of a domain abbreviation. This triggered its re-routing code and so it was
8065 trying to route what was effectively the same domain again. This normally
8066 caused routing to fail (because the router wouldn't handle the domain
8067 twice). Now Exim checks for this case specially, and just changes the
8068 casing of the domain that it ultimately uses when it transmits the message
8071 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8074 50. If a filter generated a file delivery with a non-absolute name (possible if
8075 no home directory exists for the router), the forbid_file option was not
8078 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8079 the existing equality matching.
8081 52. Exim was using ints instead of ino_t variables in some places where it was
8082 dealing with inode numbers.
8084 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8085 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8086 called TMPDIR, and if it finds it is different, it changes its value.
8088 54. The smtp_printf() function is now made available to local_scan() so
8089 additional output lines can be written before returning. There is also an
8090 smtp_fflush() function to enable the detection of a dropped connection.
8091 The variables smtp_input and smtp_batched_input are exported to
8094 55. Changed the default runtime configuration: the message "Unknown user"
8095 has been removed from the ACL, and instead placed on the localuser router,
8096 using the cannot_route_message feature. This means that any verification
8097 failures that generate their own messages won't get overridden. Similarly,
8098 the "Unrouteable address" message that was in the ACL for unverifiable
8099 relay addresses has also been removed.
8101 56. Added hosts_avoid_esmtp to the smtp transport.
8103 57. The exicyclog script was not checking for the esoteric option
8104 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8105 will work only if exicyclog is run under the appropriate euid.
8107 58. Following a discussion on the list, the rules by which Exim recognises line
8108 endings on incoming messages have been changed. The -dropcr and drop_cr
8109 options are now no-ops, retained only for backwards compatibility. The
8110 following line terminators are recognized: LF CRLF CR. However, special
8111 processing applies to CR:
8113 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8114 nor a local message in the state where . is a terminator.
8116 (ii) If a bare CR is encountered in a header line, an extra space is added
8117 after the line terminator so as not to end the header. The reasoning
8118 behind this is that bare CRs in header lines are most likely either
8119 to be mistakes, or people trying to play silly games.
8121 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8122 was being incorrectly given as 18 bytes larger than it should have been.
8123 This is a VOB (very old bug).
8125 60. This may never have affected anything current, but just in case it has:
8126 When the local host is found other than at the start of a list of hosts,
8127 the local host, those with the same MX, and any that follow, are discarded.
8128 When the list in question was part of a longer list of hosts, the following
8129 hosts (not currently being processed) were also being discarded. This no
8130 longer happens. I'm not sure if this situation could ever has previously
8133 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8136 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8137 header. This is something that is recommended in a new Internet Draft, and
8138 is something that is documented as being done by Sendmail. There are two
8139 possible values. For messages generated by the autoreply transport, Exim
8142 Auto-Submitted: auto-replied
8144 whereas for all other generated messages (e.g. bounces) it adds
8146 Auto-Submitted: auto-generated
8148 63. The "personal" condition in filters now includes a test for the
8149 Auto-Submitted: header. If it contains the string "auto-" the message it
8150 not considered personal.
8152 64. Added rcpt_include_affixes as a generic transport option.
8154 65. Added queue_only_override (default true).
8156 66. Added the syslog_duplication option.
8158 67. If what should have been the first header line of a message consisted of
8159 a space followed by a colon, Exim was mis-interpreting it as a header line.
8160 It isn't of course - it is syntactically invalid and should therefore be
8161 treated as the start of the message body. The misbehaviour could have
8162 caused a number of strange effects, including loss of data in subsequent
8163 header lines, and spool format errors.
8165 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8166 client host had authenticated. This control can now be exercised by an ACL
8167 for more flexibility.
8169 69. By default, callouts do not happen when testing with -bh. There is now a
8170 variant, -bhc, which does actually run the callout code, including
8171 consulting and updating the callout cache.
8173 70. Added support for saslauthd authentication, courtesy of Alexander
8176 71. If statvfs() failed on the spool or log directories while checking their
8177 size for availability, Exim confusingly gave the error "space shortage".
8178 Furthermore, in debugging mode it crashed with a floating point exception.
8179 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8180 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8181 serious problem, Exim now writes to the main and panic logs when this
8182 happens, with details of the failure. It then refuses to accept the
8183 incoming message, giving the message "spool directory problem" or "log
8184 directory problem" with a 421 code for SMTP messages.
8186 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8187 0, 1, and 2 exist, because some OS complain for execs without them (see
8188 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8189 descriptors. However, the code omitted to check that the open succeeded,
8190 causing mysterious errors if for some reason the permissions on /dev/null
8191 got screwed. Now Exim writes a message to the main and panic logs, and
8192 bombs out if it can't open /dev/null.
8194 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8195 interact so that it is all more flexible. It is supposed to remain
8196 backwards compatible. Also added extra_local_interfaces.
8198 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8199 to bomb out with an assertion failure - to the client this appears as a
8200 connection drop. This problem occurs in the part of the code that was taken
8201 from the Samba project. Fortunately, the assertion is in a very simple
8202 function, so I have fixed this by reproducing the function inline in the
8203 one place where it is called, and arranging for authentication to fail
8204 instead of killing the process with assert().
8206 75. The SPA client code was not working when the server requested OEM rather
8207 than Unicode encoding.
8209 76. Added code to make require_files with a specific uid setting more usable in
8210 the case where statting the file as root fails - usually a non-root-mounted
8211 NFS file system. When this happens and the failure is EACCES, Exim now
8212 forks a subprocess and does the per-uid checking as the relevant uid.
8214 77. Added process_log_path.
8216 78. If log_file_path was not explicitly set, a setting of check_log_space or
8217 check_log_inodes was ignored.
8219 79. If a space check for the spool or log partitions fails, the incident is now
8220 logged. Of course, in the latter case the data may get lost...
8222 80. Added the %p formatting code to string_format() so that it can be used to
8223 print addresses in debug_print(). Adjusted all the address printing in the
8224 debugging in store.c to use %p rather than %d.
8226 81. There was a concern that a line of code in smtp_in.c could overflow a
8227 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8228 initially expressed, the concern was not well-founded, because trailing
8229 spaces are removed early. However, if the trailing spaces were followed by
8230 a NULL, they did not get removed, so the overflow was possible. Two fixes
8233 (a) I re-wrote the offending code in a cleaner fashion.
8234 (b) If an incoming SMTP command contains a NULL character, it is rejected
8237 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8238 runs initgroups(), so that if the Exim user is in any additional groups,
8239 they will be used during message reception.
8245 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob