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.
123 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
124 related applications. Previously an "H" was used where available info
125 says that "M" should be, so change to match.
127 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
128 as arguments, so an implementation trying to copy these into a local
129 buffer was taking a taint-enforcement trap. Fix by using dynamically
130 created buffers. Similar fix for radius expansion condition.
132 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
133 reasonable, eg. to count headers. Fix by using dynamically created
134 buffers rather than a local. Do similar fixes for ACL actions "dcc",
135 "log_reject_target", "malware" and "spam"; the arguments are expanded
136 so could be handling tainted values.
138 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
139 broken the (no-op) support for this sendmail command. Restore it
140 to doing nothing, silently, and returning good status.
142 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
143 record path was given (or the default used) without a leading directory
144 path, an error occurred on trying to open it. Use the transport's working
147 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
148 transport. Previously it was the name on the DNS A-record; use instead
149 the head of the CNAME chain leading there (if there is one). This seems
150 to align better with RFC 6125.
152 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
153 smtp_accept_max_per_host allocated resources which were not released
154 when the limit was exceeded. This eventually crashed the daemon. Fix
155 by adding a release action in that path.
157 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
158 expanded; previously using tainted values was rejected. Fix by using
159 dynamically-created buffers.
161 JH/09 Relax restrictions on ACL verify condition needing access to message
162 headers. Previously they were only permitted in data and non-smtp ACLs;
163 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
164 not_blind, header_sender and header_names_ascii verification.
166 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
167 Previously a macro used one argument twice; when called with the
168 argument as an expression having side-effects, incorrect operation
169 resulted. Use an inlineable function.
171 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
172 held open for a verify callout. Previously this wan not accounted for
173 and a corrupt onward SMTP conversation resulted.
175 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
176 Previously they were embedded naively in the constructed address; when
177 needed, strip the quoting and quote the entire local_part.
178 Also make the inbound_srs expansion condition handle quoting.
180 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
181 excluded, not matching the documentation.
183 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
184 was given for the sqlite_dbfile a trap resulted.
186 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
187 "name" argument resulted in a trap. There is no reason to disallow such;
188 this was a coding error.
190 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
191 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
192 spent suspended, ignoring the POSIX definition. Previously we assumed
193 it did and a constant offset from real time could be used as a correction.
194 Change to using the same clock source for the start-of-message and the
195 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
196 exists, just to get a clock slightly more aligned to reality.
198 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
199 RFC says it is optional some validators care. The missing char was not
200 intended but triggered by a line-wrap alignment. Discovery and fix by
201 Guillaume Outters, hacked on by JH.
203 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
204 name being quoted was tainted a trap would be taken. Fix by using
205 dynamically created buffers. The routine could have been called by a
206 rewrite with the "h" flag, by using the "-F" command-line option, or
207 by using a "name=" option on a control=submission ACL modifier.
209 JH/19 SPF: change the Authentication-Results expansion component to give
210 smtp.helo when the sender domain is empty. Previously it gave
213 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
214 not in 127.0.0.0/8 to help in spotting list domains taken over by a
215 domain-parking registrar.
217 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
218 Previously when a whitespace character was specified it was not inserted
219 after removing the newline.
221 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
222 the domain part of the recipient address. This overrides any tls_sni
223 option set, which was previously used.
225 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
228 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
229 is_tainted() had an off-by-one error in the overenthusiastic direction.
230 Find and fix by Gavan. Although NetBSD is not a supported platform for
231 4.94 this bug could affect other platforms.
233 PP/01 Fix default prime selection to be consistent.
234 One path used ike23 still, instead of exim.dev.20160529.3; now both
235 execution flows will use the same DH primes (currently
236 exim.dev.20160529.3).
238 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
239 option in smtp transport, to match the documentation. Previously
240 verification was not being done in some cases where it should have been.
242 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
243 than one server was defined and depending on the platform memory layout
244 details, an internal consistency trap could be hit while walking the list
247 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
248 files. The value is used by the authresults expansion item. Previously
249 if this was used in a router or transport, a crash could result.
251 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
252 have a sufficient incidence of EINTR returns from select that an
253 interaction with other operations done by the main daemon loop exposed
254 a bug in the error-handling. This was benign apart from the log
257 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
258 for consistency with delivered "=>" and failed "**" lines. While we're
259 there, handle PRX and TFO.
261 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
262 applied. This resulted, if any header-line rewrite rules were configured,
263 in a panic-log triggerable by sending a message with a long address in
264 a header. Fix by increasing the arbitrary limit to larger than a single
265 (dewrapped) 5322 header line maximum size.
267 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
268 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
269 RFC 6648 which deprecates X- options in protocols as a general practice.
270 Changeover between the implementations is handled by the mechanisms
273 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
274 a remote which uses callout/hold. Previously the local was queued.
276 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
279 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
280 log line. The asterisk is now consistently appended to the remote IP
281 (and port, if given), and will also be provided on defer and fail log
282 lines. Previously it could be placed on the local IP if that was being
283 logged, and was only provided on delivery lines.
285 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
287 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
288 authenticator client_send option. Previously the next char, after a pair
289 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
290 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
291 documentation. There is still no way to get a leading ^ immediately
292 after a NUL (ie. for the password of a PLAIN method authenticator.
294 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
295 files. For bad sizes read, delete the record and whine to paniclog.
297 JH/38 When logging an AUTH failure, as server, do not include sensitive
298 information. Previously, the credentials would be included if given
299 as part of the AUTH command line and an ACL denied authentication.
301 JH/39 Bug 2691: fix $local_part_data. When the matching list element
302 referred to a file, bad data was returned. This likely also affected
305 JH/40 The gsasl authenticator now supports caching of the salted password
306 generated by the client-side implementation. This required the addition
307 of a new variable: $auth4.
309 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
310 left undeleted; the attempt to re-create it then failed - resulting in
311 the usual "SIGHUP tp have daemon reload configuration" to not work.
312 This affected any platform not supporting "abstract" Unix-domain
313 sockets (i.e. not Linux).
315 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
316 recipients" response to RCPT in a later response, with a 250. The
317 previous coding assumed this would not happen, and under PIPELINING
318 would result in both lost and duplicate recipients for a message.
320 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
321 Previously the weighting was incorrectly applied. Similar fix for socks
322 proxies. Found and fixed by Heiko Schlichting.
324 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
325 not handle sub-lists included using the +namedlist syntax. While
326 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
327 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
330 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
331 Previously the permanent pool was used, so the sore could not be freed.
332 This meant a connection with many messages would use continually-growing
335 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
336 per-pool so as not to waste too much space. Previously a constant size
337 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
338 DOS attacks harder. The cost is wasted memory use in the larger blocks.
340 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
341 same space repeatedly, and a smaller process footprint.
343 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
344 for GnuTLS. Previously the name started "%s" which, while not a bug,
345 looked as if if might be one.
347 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
348 "continued connection") make the $proxy_* variables available. Previously
349 the information was not passed across the exec() call for subsequent
350 transport executions. This also mean that the log lines for the
351 messages can show the proxy information.
353 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
354 receive time. With modern systems the difference is significant.
355 The historical behaviour can be restored by disabling (a new) log_selector
356 "queue_time_exclusive".
358 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
359 content) were specifically tested for. Now, also cover expansions
360 resulting in ACL names and inline ACL content.
362 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
363 making it unusable in complex expressions.
365 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
366 fail with a taint-check on the spoolfile name, and leave the message
369 HS/01 Enforce absolute PID file path name.
371 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
373 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
375 PP/02 Bug 2643: Correct TLS DH constants.
376 A missing NUL termination in our code-generation tool had led to some
377 incorrect Diffie-Hellman constants in the Exim source.
378 Reported by kylon94, code-gen tool fix by Simon Arlott.
380 PP/03 Impose security length checks on various command-line options.
381 Fixes CVE-2020-SPRSS reported by Qualys.
383 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
384 better. Reported by Qualys.
386 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
387 providing a particularly obnoxious sender full name.
390 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
392 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
395 PP/08 Change default for recipients_max from unlimited to 50,000.
397 PP/09 Fix security issue with too many recipients on a message (to remove a
398 known security problem if someone does set recipients_max to unlimited,
399 or if local additions add to the recipient list).
400 Fixes CVE-2020-RCPTL reported by Qualys.
402 PP/10 Fix security issue in SMTP verb option parsing
403 Fixes CVE-2020-EXOPT reported by Qualys.
405 PP/11 Fix security issue in BDAT state confusion.
406 Ensure we reset known-good where we know we need to not be reading BDAT
407 data, as a general case fix, and move the places where we switch to BDAT
408 mode until after various protocol state checks.
409 Fixes CVE-2020-BDATA reported by Qualys.
411 HS/03 Die on "/../" in msglog file names
413 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
414 the Exim runtime user are allowed to create files.
416 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
419 QS/03 When reading the output from interpreted forward files we do not
420 pass the pipe between the parent and the interpreting process to
421 executed child processes (if any).
423 QS/04 Always die if requested from internal logging, even is logging is
426 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
427 the API; compilation noo longer completes with DMARC support included.
428 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
429 been reported on other platforms.
431 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
433 JH/56 Make the majority of info read from config files readonly, for defence-in-
434 depth against exploits. Suggestion by Qualys.
435 Not supported on Solaris 10.
437 JH/57 Fix control=fakreject for a custom message containing tainted data.
438 Previously this resulted in a log complaint, due to a re-expansion present
439 since fakereject was originally introduced.
441 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
442 modifier was given, a loop resulted.
444 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
445 pipelined SMTP command followed the BDAT LAST then it would be
446 incorrectly treated as part of the message body, causing a verification
449 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
450 require looser validation rules than those for 821-level addresses,
451 which only permit IP addresses.
457 JH/01 Avoid costly startup code when not strictly needed. This reduces time
458 for some exim process initialisations. It does mean that the logging
459 of TLS configuration problems is only done for the daemon startup.
461 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
463 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
464 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
467 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
468 DKIM signing is being done. Previously a transport_filter would always
469 disable CHUNKING, falling back to traditional DATA.
471 JH/05 Regard command-line recipients as tainted.
473 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
475 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
476 PAM library frees one of the arguments given to it, despite the
477 documentation. Therefore a plain malloc must be used.
479 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
480 on-stack buffers were used, resulting in a taint trap when DSN information
481 copied from a received message was written into the buffer.
483 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
484 the ordering of its ARC headers. This caused a crash.
486 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
487 a new record was being constructed with information from the peer, a trap
490 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
491 installation would get error messages from DMARC verify, when it hit the
492 nonexistent file indicated by the default. Distros wanting DMARC enabled
493 should both provide the file and set the option.
494 Also enforce no DMARC verification for command-line sourced messages.
496 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
497 could, depending on the platform, hang at the STARTTLS response.
499 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
500 message on a connection. Previously if one message had ARC headers and
501 the following one did not, a crash could result when adding an
502 Authentication-Results: header.
504 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
505 Exim main code and Exim-related utities. The introduction of taint
506 tracking also did many adjustments to string handling. Since then, eximon
507 frequently terminated with an assert failure.
509 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
510 check for 452 responses. This slightly helps the inefficieny of doing
511 a large alias-expansion into a recipient-limited target. The max_rcpt
512 transport option still applies (and at the current default, will override
513 the new feature). The check is done for either cause of synch, and forces
514 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
515 connection. The new facility is not tunable at this time.
517 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
518 library live data was being used, so the results became garbage. Make
519 copies while it is still usable.
521 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
522 on delivery deferred (==) and failed (**) lines (if a delivery was
523 attemtped). Previously it was only on completion (=>) lines.
525 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
526 for the expansion of the server_scram_iter and server_scram_salt options.
528 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
529 are now specifically given a NO_DATA response without hitting the system
530 resolver. The library goes on to do the now-standard TXT lookup.
531 Use of dnsdb lookups is not affected.
533 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
534 only retrieve the errormessage once. Previously two calls to dlerror()
535 were used, and the second one (for mainlog/paniclog) retrieved null
538 JH/20 Taint checking: disallow use of tainted data for
539 - the appendfile transport file and directory options
540 - the pipe transport command
541 - the autoreply transport file, log and once options
542 - file names used by the redirect router (including filter files)
544 - paths used by single-key lookups
545 Previously this was permitted.
547 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
548 adjusted the size of a major service buffer; this failed because the
549 buffer was in use at the time. Change to a compile-time increase in the
550 buffer size, when this authenticator is compiled into exim.
552 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
553 previous fast-mode was untenable in the face of glibs using mmap to
554 support larger malloc requests.
556 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
557 New values supported, if defined on system where compiled:
558 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
559 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
561 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
562 running a limited number of proceses in parallel, a benefit is gained. The
563 amount varies with the platform hardware and load. The use of the option
564 queue_run_in_order means we cannot do this, as ordering becomes
567 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
568 had introduced a string-copy (for ensuring NUL-termination) which was not
569 appropriate for that case, which can include embedded NUL bytes in the
570 block of data. Investigation showed the copy to actually be needless, the
571 data being length-specified.
573 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
574 done during a receiving connection, and both used TLS, global info was
575 used rather than per-connection info for tracking the state of data
576 queued for transmission. This could result in a connection hang.
578 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
579 Previously, when delivering serveral messages down a single connection
580 only the first would provide a SIZE. This was due to the size information
581 not being properly tracked.
583 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
584 TAI (at 37 seconds currently), pretend to be in UTC for time-related
585 expansion and logging. Previously, spurious values such as a future
586 minute could be seen.
588 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
589 it could crash from a null-deref. This could also affect the
590 ${addresses: } operator and ${readsock } item.
592 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
593 message following a mime one, the variable was not reset.
595 JH/30 When an pipelined-connect fails at the first response, assume incorrect
596 cached capability (perhaps the peer reneged?) and immediately retry in
599 JH/31 Fix spurious detection of timeout while writing to transport filter.
601 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
602 an attempt to copy the string was made before checking it.
604 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
605 taint of the lookup key was maintained; we now regard the presence in the
606 filesystem as sufficient validation.
608 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
609 argument is supplied.
611 JH/35 The dsearch lookup now requires that the directory is an absolute path.
612 Previously this was not checked, and nonempty relative paths made an
613 access under Exim's current working directory.
615 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
616 Previously no event was raised.
618 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
619 parameter supplied by the sender MAIL FROM command. Previously it was
620 ignored, and only the check_spool_space option value for the required
623 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
624 the size of the signing public-key. Previously it was instead giving
625 the size of the signature hash.
627 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
628 the default. See the (new) dkim_verify_min_keysizes option.
630 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
631 and an ACL use a lookup for checking either the local_part or domain,
632 stale data could be accessed. Ensure that variable references are
633 dropped between messages.
635 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
636 by the client was not checked as pointing within response data before
637 being used. A malicious client could thus cause an out-of-bounds read and
638 possibly gain authentication. Fix by adding the check.
640 JH/42 Internationalisation: change the default for downconversion in the smtp
641 transport to be "if needed". Previously it was "as previously set" for
642 the message, which usually meant "if needed" for message-submission but
643 "no" for everything else. However, MTAs have been seen using SMTPUTF8
644 even when the envelope addresses did not need it, resulting in forwarding
645 failures to non-supporting MTAs. A downconvert in such cases will be
646 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
647 transport. The change does mean that addresses needing conversion will
648 be converted when previously a delivery failure would occur.
650 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
651 response was received it would be used unchecked in a fail-DSN, violating
652 standards on line-length limits. Truncate if needed.
654 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
655 doesn't work. (Additionally add a new main config option to configure the
662 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
663 side, to decode a TLS 1.3 packet capture.
665 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
666 Previously the default library behaviour applied, sending two, each in
669 JH/03 Debug output for ACL now gives the config file name and line number for
672 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
674 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
676 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
677 buffer overrun for (non-chunking) other transports.
679 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
680 TLS1.3, means that a server rejecting a client certificate is not visible
681 to the client until the first read of encrypted data (typically the
682 response to EHLO). Add detection for that case and treat it as a failed
683 TLS connection attempt, so that the normal retry-in-clear can work (if
684 suitably configured).
686 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
687 and/or domain. Found and fixed by Jason Betts.
689 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
690 configuration). If a CNAME target was not a wellformed name pattern, a
693 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
694 the OS reports them interleaved with other addresses.
696 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
697 used both for input and for a verify callout, both encrypted, SMTP
698 responses being sent by the server could be lost. This resulted in
699 dropped connections and sometimes bounces generated by a peer sending
702 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
703 string. Previously it was possible to cause undefined behaviour in a
704 library routine (usually a crash). Found by "zerons".
706 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
709 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
710 API was removed, so update to use the newer ones.
712 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
713 any timeout set, is taking a long time. Previously we would hang on to a
714 rotated logfile "forever" if the input was arriving with long gaps
715 (a previous attempt to fix addressed lack, for a long time, of initial
718 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
719 shared (NFS) environment. The length of the tempfile name is now
720 4 + 16 ("hdr.$message_exim_id") which might break on file
721 systems which restrict the file name length to lower values.
724 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
725 shared (NFS) environment.
727 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
728 did for all versions <4.90). Notably -M, -m, --invert, -I may be
731 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
732 on some platforms for bit 31.
734 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
735 to changes apparently associated with TLS1.3 handling some of the APIs
736 previously used were either nonfunctional or inappropriate. Strings
737 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
738 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
739 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
740 This affects log line X= elements, the $tls_{in,out}_cipher variables,
741 and the use of specific cipher names in the encrypted= ACL condition.
743 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
745 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
746 verification result was not updated unless hosts_require_ocsp applied.
748 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
749 queue_list_requires_admin set to false, non-admin users were denied the
752 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
753 directory-of-certs mode. Previously they were advertised despite the
756 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
757 A single TCP connection by a client will now hold a TLS connection open
758 for multiple message deliveries, by default. Previously the default was to
761 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
762 default. If built with the facility, DANE will be used. The facility
763 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
765 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
766 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
767 must be defined and you must still, unless you define DISABLE_TLS, manage
768 the the include-dir and library-file requirements that go with that
769 choice. Non-TLS builds are still supported.
771 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
774 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
775 default. If the platform supports and has the facility enabled, it will
776 be requested on all coneections.
778 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
779 controlled by the build-time option SUPPORT_PIPE_CONNECT.
781 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
783 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
784 success-DSN messages. Previously the From: header was always the default
785 one for these; the option was ignored.
787 JH/28 Fix the timeout on smtp response to apply to the whole response.
788 Previously it was reset for every read, so a teergrubing peer sending
789 single bytes within the time limit could extend the connection for a
790 long time. Credit to Qualsys Security Advisory Team for the discovery.
792 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
793 delivery address, which leaked information of the results of local
794 forwarding. Change to the original envelope recipient address, per
797 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
798 requested. Previously not bounce was generated and a log entry of
799 error ignored was made.
801 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
803 JH/32 Introduce a general tainting mechanism for values read from the input
804 channel, and values derived from them. Refuse to expand any tainted
805 values, to catch one form of exploit.
807 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
808 was unused and the unexpanded text used for the test. Found and
809 fixed by Ruben Jenster.
811 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
812 an attempt to use a TLS library read routine dereffed a nul pointer,
815 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
816 them in our smtp response.
818 JH/36 Have the general router option retry_use_local_part default to true when
819 any of the restrictive preconditions are set (to anything). Previously it
820 was only for check_local user. The change removes one item of manual
821 configuration which is required for proper retries when a remote router
822 handles a subset of addresses for a domain.
824 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
825 link count into consideration.
827 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
828 caused the extension of big_buffer, the following lines were ignored.
830 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
831 accordance with RFC 2308. Previously there was no expiry, so a longlived
832 receive process (eg. due to ACL delays) versus a short SOA value could
835 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
837 JH/39 Promote DMARC support to mainline.
839 JH/40 Bug 2452: Add a References: header to DSNs.
841 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
842 parameters. The relevant library call is documented as "Deprecated: This
843 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
844 3.6.0, DH parameters are negotiated following RFC7919."
846 HS/06 Change the default of dnssec_request_domains to "*"
848 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
849 carried on and emitted a BDAT command, even when PIPELINING was not
852 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
853 buffer was used for the filename, resulting in a trap when tainted
854 arguments (eg. $domain) were used.
856 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
857 recommended to avoid a possible server-load attack. The feature can be
858 re-enabled via the openssl_options main cofiguration option.
860 JH/45 local_scan API: documented the current smtp_printf() call. This changed
861 for version 4.90 - adding a "more data" boolean to the arguments.
862 Bumped the ABI version number also, this having been missed previously;
863 release versions 4.90 to 4.92.3 inclusive were effectively broken in
864 respect of usage of smtp_printf() by either local_scan code or libraries
865 accessed via the ${dlfunc } expansion item. Both will need coding
866 adjustment for any calls to smtp_printf() to match the new function
867 signature; a FALSE value for the new argument is always safe.
869 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
870 the file-offset (which the Linux syscall does, and exim expects); this
871 resulted in an indefinite loop.
873 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
874 to do ARC verification. The Authentication-Results: header line added
875 by the configuration then had no ARC item.
877 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
878 an incorrect mode was used for reading the result, resulting in it being
885 JH/01 Remove code calling the customisable local_scan function, unless a new
886 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
888 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
889 non-signal-safe functions being used.
891 JH/03 Bug 2269: When presented with a received message having a stupidly large
892 number of DKIM-Signature headers, disable DKIM verification to avoid
893 a resource-consumption attack. The limit is set at twenty.
895 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
896 report of oldest_pass in ${authres } in consequence, and separate out
897 some descriptions of reasons for verification fail.
899 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
900 files in the spool were present and unlocked. A queue-runner could spot
901 them, resulting in a duplicate delivery. Fix that by doing the unlock
902 after the unlink. Investigation by Tim Stewart. Take the opportunity to
903 add more error-checking on spoolfile handling while that code is being
906 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
907 No known attacks, no CVE, this is defensive hardening.
909 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
910 a queue-runner could start a delivery while other operations were ongoing.
911 Cutthrough delivery was a common victim, resulting in duplicate delivery.
912 Found and investigated by Tim Stewart. Fix by using the open message data
913 file handle rather than opening another, and not locally closing it (which
914 releases a lock) for that case, while creating the temporary .eml format
915 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
917 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
918 $sender_verify_failure/$recipient_verify_failure to "random".
920 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
923 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
924 Previously this would segfault.
926 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
929 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
930 like zero, since the resolver should be doing this for us, But we need one
931 as a CNAME but no MX presence gets the CNAME returned; we need to check
932 that doesn't point to an MX to declare it "no MX returned" rather than
933 "error, loop". A new main option is added so the older capability of
934 following some limited number of chain links is maintained.
936 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
938 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
939 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
940 now-deprecated earlier definitions used only specified the range up to TLS
941 1.2 (in the older-version library docs).
943 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
945 JH/15 Rework TLS client-side context management. Stop using a global, and
946 explicitly pass a context around. This enables future use of TLS for
947 connections to service-daemons (eg. malware scanning) while a client smtp
948 connection is using TLS; with cutthrough connections this is quite likely.
950 JH/16 Fix ARC verification to do AS checks in reverse order.
952 JH/17 Support a "tls" option on the ${readsocket } expansion item.
954 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
955 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
956 Previously the "utf8" would be re-prepended for every additional message.
958 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
959 Previously thery were accepted, resulting in issues when attempting to
960 forward messages to a non-supporting MTA.
962 PP/02 Let -n work with printing macros too, not just options.
964 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
965 one parent address was copied, and bogus data was used at delivery-logging
966 time. Either a crash (after delivery) or bogus log data could result.
967 Discovery and analysis by Tim Stewart.
969 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
970 Previously if the string ended mid-character, we did not insert the
971 promised '?' replacement.
973 PP/04 Documentation: current string operators work on bytes, not codepoints.
975 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
976 should pack well giving a smaller memory footprint so better caching and
977 therefore performance. Group the declarations where this can't be done so
978 that the byte-sized flag variables are not interspersed among pointer
979 variables, giving a better chance of good packing by the compiler.
981 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
982 non-null, to avoid issues with sites running BATV. Previously reports were
983 sent with an empty envelope sender so looked like bounces.
985 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
986 The ignore_error flag wasn't being returned from the filter subprocess so
987 was not set for later routers. Investigation and fix by Matthias Kurz.
989 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
990 and a msg:complete for the whole, when a message is manually removed using
991 -Mrm. Developement by Matthias Kurz, hacked on by JH.
993 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
994 a "Gnu special" function, asprintf() in the DB utility binary builds; I
995 hope that is portable enough.
997 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
998 requiring a known-CA anchor certificate; make it now rely entirely on the
999 TLSA as an anchor. Checking the name on the leaf cert against the name
1000 on the A-record for the host is still done for TA (but not for EE mode).
1002 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1003 included in delivery lines for non-proxied connections, when compiled with
1004 SUPPORT_SOCKS and running with proxy logging enabled.
1006 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1007 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1008 move the existing event to fire before the normal logging of message
1009 failure so that custom logging is bracketed by normal logging.
1011 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1012 msg:fail:internal event. Developement by Matthias Kurz.
1014 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1015 far too small for todays use of crypto signatures stored there. Go all
1016 the way to the max DNS message size of 64kB, even though this might be
1017 overmuch for IOT constrained device use.
1019 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1020 copy a string over itself. The library routine is documented as not
1021 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1023 JH/32 For main options check_spool_space and check_inode_space, where the
1024 platform supports 64b integers, support more than the previous 2^31 kB
1025 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1026 the previous G, M, k.
1028 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1029 $authenticated_fail_id variable on authentication failure. Previously
1032 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1033 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1034 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1035 GNUTLS_SEC_PARAM_MEDIUM.
1037 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1038 side. Previously we would continue as if no SNI had been received.
1040 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1041 "<" character, which should be a list-separator specification, we walked
1042 off past the nul-terimation.
1044 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1045 causes) even when the retry time is not yet met. Previously they were
1046 not, meaning that when (say) an account was over-quota and temp-rejecting,
1047 and multiple senders' messages were queued, only one sender would get
1048 notified on each configured delay_warning cycle.
1050 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1052 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1053 cipher-suites, an error can be left on the stack even for a succeeding
1054 accept; this results in impossible error messages when a later operation
1057 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1058 return error codes indicating retry. Under TLS1.3 this becomes required.
1060 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1061 it only wrote the new authenticators, resulting in a lack of tracking of
1062 peer changes of ESMTP extensions until the next cache flush.
1064 JH/41 Fix the loop reading a message header line to check for integer overflow,
1065 and more-often against header_maxsize. Previously a crafted message could
1066 induce a crash of the recive process; now the message is cleanly rejected.
1068 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1069 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1075 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1076 When redis_servers is set to a list of > 1 element, and the Redis servers
1077 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1078 case of MOVED into a DEFER case instead, thus moving the query onto the
1079 next server in the list. For a cluster of N elements, all N servers must
1080 be defined in redis_servers.
1082 GF/02 Catch and remove uninitialized value warning in exiqsumm
1083 Check for existence of @ARGV before looking at $ARGV[0]
1085 JH/01 Replace the store_release() internal interface with store_newblock(),
1086 which internalises the check required to safely use the old one, plus
1087 the allocate and data copy operations duplicated in both (!) of the
1088 extant use locations.
1090 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1091 modifier. This matches the restriction on the commandline.
1093 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1094 Previously only the last row was returned.
1096 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1097 we assumed that tags in the header were well-formed, and parsed the
1098 element content after inspecting only the first char of the tag.
1099 Assumptions at that stage could crash the receive process on malformed
1102 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1103 While running the DKIM ACL we operate on the Permanent memory pool so that
1104 variables created with "set" persist to the DATA ACL. Also (at any time)
1105 DNS lookups that fail create cache records using the Permanent pool. But
1106 expansions release any allocations made on the current pool - so a dnsdb
1107 lookup expansion done in the DKIM ACL releases the memory used for the
1108 DNS negative-cache, and bad things result. Solution is to switch to the
1109 Main pool for expansions.
1110 While we're in that code, add checks on the DNS cache during store_reset,
1111 active in the testsuite.
1112 Problem spotted, and debugging aided, by Wolfgang Breyha.
1114 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1115 When none of the hosts presented to a transport match an already-open
1116 connection, close it and proceed with the list. Previously we would
1117 queue the message. Spotted by Lena with Yahoo, probably involving
1120 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1121 Previously a spurious "250 OK id=" response was appended to the proper
1124 JH/08 The "support for" informational output now, which built with Content
1125 Scanning support, has a line for the malware scanner interfaces compiled
1126 in. Interface can be individually included or not at build time.
1128 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1129 by the template makefile "src/EDITME". The "STREAM" support for an older
1130 ClamAV interface method is removed.
1132 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1133 rows affected is given instead).
1135 JH/11 The runtime Berkeley DB library version is now additionally output by
1136 "exim -d -bV". Previously only the compile-time version was shown.
1138 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1139 SMTP connection. Previously, when one had more recipients than the
1140 first, an abortive onward connection was made. Move to full support for
1141 multiple onward connections in sequence, handling cutthrough connection
1142 for all multi-message initiating connections.
1144 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1145 routers. Previously, a multi-recipient message would fail to match the
1146 onward-connection opened for the first recipient, and cause its closure.
1148 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1149 a timeout on read on a GnuTLS initiating connection, resulting in the
1150 initiating connection being dropped. This mattered most when the callout
1151 was marked defer_ok. Fix to keep the two timeout-detection methods
1154 JH/15 Relax results from ACL control request to enable cutthrough, in
1155 unsupported situations, from error to silently (except under debug)
1156 ignoring. This covers use with PRDR, frozen messages, queue-only and
1159 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1161 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1162 metadata, resulting in a crash in free().
1164 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1165 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1166 Broken also in d185889f4, with init system revamp.
1168 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1169 Previously we abruptly closed the connection after reading a malware-
1170 found indication; now we go on to read the "scan ok" response line,
1173 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1174 ACL. Previously, a crash would result.
1176 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1177 macro text after a replacement (previously it was only once per line) and
1178 by skipping builtin macros when searching for an uppercase lead character.
1180 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1181 for the build is renamed.
1183 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1184 was allocated for every new TLS startup, meaning one per message. Fix
1185 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1187 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1188 reported the original. Fix to report (as far as possible) the ACL
1189 result replacing the original.
1191 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1192 OpenSSL. Certificate information is loaded for every new TLS startup,
1193 and the resources needed to be freed.
1195 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1197 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1200 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1201 DATA response info to the (existing) per-recipient response info for
1202 the "C=" log element. It can have useful tracking info from the
1203 destination system. Patch from Simon Arlott.
1205 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1206 length value. Previously this would segfault.
1208 HS/02 Support Avast multiline protoocol, this allows passing flags to
1209 newer versions of the scanner.
1211 JH/28 Ensure that variables possibly set during message acceptance are marked
1212 dead before release of memory in the daemon loop. This stops complaints
1213 about them when the debug_store option is enabled. Discovered specifically
1214 for sender_rate_period, but applies to a whole set of variables.
1215 Do the same for the queue-runner and queue-list loops, for variables set
1216 from spool message files. Do the same for the SMTP per-message loop, for
1217 certain variables indirectly set in ACL operations.
1219 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1220 as a multi-recipient message from a mailinglist manager). The coding had
1221 an arbitrary cutoff number of characters while checking for more input;
1222 enforced by writing a NUL into the buffer. This corrupted long / fast
1223 input. The problem was exposed more widely when more pipelineing of SMTP
1224 responses was introduced, and one Exim system was feeding another.
1225 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1226 receiving system, and refused recipients seen by the sending system
1227 (propating to people being dropped from mailing lists).
1228 Discovered and pinpointed by David Carter.
1230 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1231 replaced by the ${authresults } expansion.
1233 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1235 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1236 allows proper process termination in container environments.
1238 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1239 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1241 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1242 and "err_perm", deprecated since 4.83 when the RFC-defined words
1243 "temperror" and "permerror" were introduced.
1245 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1246 transport-filters or DKIM-signing. The restriction was lost in the
1247 consolidation of verify-callout and delivery SMTP handling.
1248 Extend the restriction to also cover ARC-signing.
1250 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1251 in defer=pass mode supply a 450 to the initiator. Previously the message
1254 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1255 tls_require_ciphers is used as before.
1257 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1258 "pass_unscanned". Only tmpfails from the scanner are written to
1259 the paniclog, as they may require admin intervention (permission
1260 denied, license issues). Other scanner errors (like decompression
1261 bombs) do not cause a paniclog entry.
1263 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1264 Previously it was possible to log spurious information in receive log
1267 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1268 triggered odd behaviour from Outlook Express clients.
1270 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1273 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1274 since the IETF WG has not yet settled on that versus the original
1275 "bare" representation.
1277 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1278 Previously the millisecond value corrupted the output.
1279 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1280 corrupted the output.
1286 JH/01 Rework error string handling in TLS interface so that the caller in
1287 more cases is responsible for logging. This permits library-sourced
1288 string to be attached to addresses during delivery, and collapses
1289 pairs of long lines into single ones.
1291 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1292 during configuration. Wildcards are allowed and expanded.
1294 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1295 This permits better logging.
1297 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1298 do not close it down and have the child transport start it up again on
1299 the passed-on TCP connection. Instead, proxy the child (and any
1300 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1301 affected: the continued delivery log lines do not have any DNSSEC, TLS
1302 Certificate or OCSP information. TLS cipher information is still logged.
1304 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1305 identical IP addresses on different listening ports. Will also affect
1308 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1309 add noisy ifdef guards to special-case this sillyness.
1310 Patch from Bernd Kuhls.
1312 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1313 than 255 are no longer allowed.
1315 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1316 Disable the session-cache too, which might reduce our load. Since we
1317 currrectly use a new context for every connection, both as server and
1318 client, there is no benefit for these.
1319 GnuTLS appears to not support tickets server-side by default (we don't
1320 call gnutls_session_ticket_enable_server()) but client side is enabled
1321 by default on recent versions (3.1.3 +) unless the PFS priority string
1324 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1325 <https://reproducible-builds.org/specs/source-date-epoch/>.
1327 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1328 the check for any unsuccessful recipients did not notice the limit, and
1329 erroneously found still-pending ones.
1331 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1332 MSG_MORE. Only in-clear (not on TLS connections).
1334 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1335 transport-filter is involved we can buffer the headers in memory for
1336 creating the signature, and read the spool data file once for the
1337 signature and again for transmission.
1339 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1340 4.77 as the kernel support then wasn't solid, having issues in 64bit
1341 mode. Now, it's been long enough. Add support for FreeBSD also.
1343 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1344 case where the routing stage had gathered several addresses to send to
1345 a host before calling the transport for the first, we previously failed
1346 to close down TLS in the old transport process before passing the TCP
1347 connection to the new process. The new one sent a STARTTLS command
1348 which naturally failed, giving a failed delivery and bloating the retry
1349 database. Investigation and fix prototype from Wolfgang Breyha.
1351 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1352 false-negatives in the check that the sender had not preempted a response
1353 or prompt from Exim (running as a server), due to that code's lack of
1354 awareness of the SMTP input buffering.
1356 PP/04 Add commandline_checks_require_admin option.
1357 Exim drops privileges sanely, various checks such as -be aren't a
1358 security problem, as long as you trust local users with access to their
1359 own account. When invoked by services which pass untrusted data to
1360 Exim, this might be an issue. Set this option in main configuration
1361 AND make fixes to the calling application, such as using `--` to stop
1364 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1365 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1366 responses to those, into a single TLS record each way (this usually means
1367 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1370 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1371 affects you only if you're dancing at the edge of the param size limits.
1372 If you are, and this message makes sense to you, then: raise the
1373 configured limit or use OpenSSL 1.1. Nothing we can do for older
1376 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1377 cmdline element to get the documented default one. Previously it was
1380 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1381 are unsuable for non-message-specific reasons - eg. connection timeout,
1382 banner-time rejection.
1384 JH/16 Fix logging of delivery remote port, when specified by router, under
1387 PP/06 Repair manualroute's ability to take options in any order, even if one
1388 is the name of a transport.
1391 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1393 JH/17 Change the list-building routines interface to use the expanding-string
1394 triplet model, for better allocation and copying behaviour.
1396 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1397 Previously it was constructed the first time a possibly-matching string
1398 was met in the configuration file input during startup; now it is done
1401 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1402 compatible one, to avoid the (poorly documented) possibility of a config
1403 file in the working directory redirecting the DB files, possibly correpting
1404 some existing file. CVE-2017-10140 assigned for BDB.
1406 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1407 cache-hot. Previously, although the result was properly cached, the
1408 initial verify call returned a defer.
1410 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1411 the main verify for receipient in uncached-mode.
1413 JH/22 Retire historical build files to an "unsupported" subdir. These are
1414 defined as "ones for which we have no current evidence of testing".
1416 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1417 if present. Previously it was ignored.
1419 JH/24 Start using specified-initialisers in C structure init coding. This is
1420 a C99 feature (it's 2017, so now considered safe).
1422 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1423 if was a fixed-sized field and bitmask ops via macros; it is now more
1426 PP/07 GitHub PR 56: Apply MariaDB build fix.
1427 Patch provided by Jaroslav Škarvada.
1429 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1430 during Coverity cleanups [4.87 JH/47]
1431 Diagnosis and fix provided by Michael Fischer v. Mollard.
1433 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1434 the right size to place the terminating semicolon on its own folded
1435 line, the header hash was calculated to an incorrect value thanks to
1436 the (relaxed) space the fold became.
1438 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1439 and confused the parent.
1441 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1442 which could crash as a result. This could lead to undeliverable messages.
1444 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1447 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1448 matters on fast-turnover and PID-randomising systems, which were getting
1449 out-of-order delivery.
1451 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1452 a possibly-overlapping copy. The symptom was that "Remote host closed
1453 connection in response to HELO" was logged instead of the actual 4xx
1456 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1457 Previously only that bufferd was discarded, resulting in SYMTP command
1460 JH/32 DKIM: when a message has multiple signatures matching an identity given
1461 in dkim_verify_signers, run the dkim acl once for each. Previously only
1462 one run was done. Bug 2189.
1464 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1465 "panic the current process" to "deliberately defer". The panic log is
1466 still written with the problem list name; the mail and reject logs now
1467 get a temp-reject line for the message that was being handled, saying
1468 something like "domains check lookup or other defer". The SMTP 451
1469 message is still "Temporary local problem".
1471 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1472 A crafted sequence of BDAT commands could result in in-use memory beeing
1473 freed. CVE-2017-16943.
1475 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1476 from SMTP input. Previously it was always done; now only done for DATA
1477 and not BDAT commands. CVE-2017-16944.
1479 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1480 to the message (such as an overlong header line). Previously this was
1481 not done and we did not exit BDAT mode. Followon from the previous item
1482 though a different problem.
1488 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1489 than -2003 did; needs libidn2 in addition to libidn.
1491 JH/02 The path option on a pipe transport is now expanded before use.
1493 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1494 Patch provided by "Björn", documentation fix added too.
1496 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1497 missing a wire-to-host endian conversion.
1499 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1500 close after a BDAT command line could be taken as a following command,
1501 giving a synch failure. Fix by only checking for synch immediately
1502 before acknowledging the chunk.
1504 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1505 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1506 macro. Patches provided by Josh Soref.
1508 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1509 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1510 by IANA as a service extension. However, John Klensin suggests that we
1513 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1514 the dkim code may be unix-mode line endings rather than smtp wire-format
1515 CRLF, so prepend a CR to any bare LF.
1517 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1518 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1520 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1521 the first were themselves being wrongly included in the feed into dkim
1522 processing; with most chunk sizes in use this resulted in an incorrect
1523 body hash calculated value.
1525 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1526 DKIM signature block, for verification. Although advised against by
1527 standards it is specifically not ruled illegal.
1529 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1531 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1532 missing a body hash (the bh= tag).
1534 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1535 It seems that HAProxy sends the Proxy Protocol information in clear and
1536 only then does a TLS startup, so do the same.
1538 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1539 TCP connections (such as for Spamd) unless the daemon successfully set
1540 Fast Open mode on its listening sockets. This fixes breakage seen on
1541 too-old kernels or those not configured for Fast Open, at the cost of
1542 requiring both directions being enabled for TFO, and TFO never being used
1543 by non-daemon-related Exim processes.
1545 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1546 endings, at least on the first header line. Try to canonify any that get
1547 past that check, despite the cost.
1549 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1550 now limited to an arbitrary five deep, while parsing addresses with the
1551 strip_excess_angle_brackets option enabled.
1553 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1554 instead leave the unprompted TLS handshake in socket buffer for the
1555 TLS library to consume.
1557 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1559 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1561 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1562 region in one operation, for speed, and this leaves assigned pointers
1563 dangling. Add checks run only under the testsuite which checks all
1564 variables at a store-reset and panics on a dangling pointer; add code
1565 explicitly nulling out all the variables discovered. Fixes one known
1566 bug: a transport crash, where a dangling pointer for $sending_ip_address
1567 originally assigned in a verify callout, is re-used.
1569 PP/06 Drop '.' from @INC in various Perl scripts.
1571 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1573 PP/08 Reduce a number of compilation warnings under clang; building with
1574 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1575 should be warning-free.
1577 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1579 HS/01 Fix portability problems introduced by PP/08 for platforms where
1580 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1582 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1583 chunk. This allows us to accept broken chunked messages. We need a more
1584 general solution here.
1586 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1587 already-broken messages in the queue.
1589 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1591 JH/19 Fix reference counting bug in routing-generated-address tracking.
1597 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1598 supports it and a size is available (ie. the sending peer gave us one).
1600 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1601 years of being deprecated). The replacements are the ACLs
1602 acl_smtp_mime and acl_not_smtp_mime.
1604 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1605 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1606 or even an in-clear connection were permitted. Now, if the host lookup
1607 was dnssec and dane was requested then the host is only used if the
1608 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1609 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1610 if one fails this test.
1611 This means that a poorly-configured remote DNS will make it incommunicado;
1612 but it protects against a DNS-interception attack on it.
1614 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1615 non-noisy when a race steals the message being considered.
1617 JH/05 If main configuration option tls_certificate is unset, generate a
1618 self-signed certificate for inbound TLS connections.
1620 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1621 in rewrites and routers.
1623 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1624 and logged a warning sing 4.83; now they are a configuration file error.
1626 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1627 (lacking @domain). Apply the same qualification processing as RCPT.
1629 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1631 JH/10 Support ${sha256:} applied to a string (as well as the previous
1634 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1635 a cutthrough deliver is pending, as we always want to make a connection.
1636 This also avoids re-routing the message when later placing the cutthrough
1637 connection after a verify cache hit.
1638 Do not update it with the verify result either.
1640 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1641 when routing results in more than one destination address.
1643 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1644 signing (which inhibits the cutthrough capability). Previously only
1645 the presence of an option was tested; now an expansion evaluating as
1646 empty is permissible (obviously it should depend only on data available
1647 when the cutthrough connection is made).
1649 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1650 the relevant preceding SMTP command did not note the pipelining mode.
1652 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1653 Previously they were not counted.
1655 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1656 as one having no matching records. Previously we deferred the message
1657 that needed the lookup.
1659 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1660 distinguished as "(=".
1662 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1663 for missing MX records. Previously it only worked for missing A records.
1665 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1667 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1668 after the data-go-ahead and data-ack. Patch from Jason Betts.
1670 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1671 even for a "none" policy. Patch from Tony Meyer.
1673 JH/22 Fix continued use of a connection for further deliveries. If a port was
1674 specified by a router, it must also match for the delivery to be
1677 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1678 When one of the recipients of a message was unsuitable for the connection
1679 (has no matching addresses), we lost track of needing to mark it
1680 deferred. As a result mail would be lost.
1682 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1684 JH/25 Decoding ACL controls is now done using a binary search; the source code
1685 takes up less space and should be simpler to maintain. Merge the ACL
1686 condition decode tables also, with similar effect.
1688 JH/26 Fix problem with one_time used on a redirect router which returned the
1689 parent address unchanged. A retry would see the parent address marked as
1690 delivered, so not attempt the (identical) child. As a result mail would
1693 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1694 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1695 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1698 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1699 Main config options check_{log,spool}_{inodes,space} are now
1700 100 inodes, 10MB unless set otherwise in the configuration.
1702 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1703 Previously it only applied to the main-section connection policy
1706 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1708 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1709 by me. Added RFC7919 DH primes as an alternative.
1711 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1712 are not in the system include path.
1714 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1715 GnuTLS, when a session startup failed (eg because the client disconnected)
1716 Exim did stdio operations after fclose. This was exposed by a recent
1717 change which nulled out the file handle after the fclose.
1719 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1720 signed directly by the cert-signing cert, rather than an intermediate
1721 OCSP-signing cert. This is the model used by LetsEncrypt.
1723 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1725 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1726 an incoming connection.
1728 HS/02 Bug 1802: Do not half-close the connection after sending a request
1731 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1732 fallback to "prime256v1".
1734 JH/34 SECURITY: Use proper copy of DATA command in error message.
1735 Could leak key material. Remotely exploitable. CVE-2016-9963.
1741 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1742 and 3.4.4 - once the server is enabled to respond to an OCSP request
1743 it does even when not requested, resulting in a stapling non-aware
1744 client dropping the TLS connection.
1746 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1747 support variable-length bit vectors. No functional change.
1749 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1750 The I= interface field on outgoing lines is now after the H= remote
1751 host field, same as incoming lines. There is a separate
1752 outgoing_interface log selector which allows you to disable the
1755 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1756 If not running log_selector +smtp_connection the mainlog would be held
1757 open indefinitely after a "too many connections" event, including to a
1758 deleted file after a log rotate. Leave the per net connection logging
1759 leaving it open for efficiency as that will be quickly detected by the
1760 check on the next write.
1762 HS/01 Bug 1671: Fix post transport crash.
1763 Processing the wait-<transport> messages could crash the delivery
1764 process if the message IDs didn't exist for some reason. When
1765 using 'split_spool_directory=yes' the construction of the spool
1766 file name failed already, exposing the same netto behaviour.
1768 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1769 mime_regex ACL conditions.
1771 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1772 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1773 to HELO, local diagnostic string.
1775 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1776 log line to debug. Even when configured with a tls authenticator many
1777 client connections are expected to not authenticate in this way, so
1778 an authenticate fail is not an error.
1780 HS/02 Add the Exim version string to the process info. This way exiwhat
1781 gives some more detail about the running daemon.
1783 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1784 matter for fast-change records such as DNSBLs.
1786 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1787 retry record, even if constant. There may be multiple transports with
1788 different interface settings and the retry behaviour needs to be kept
1791 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1793 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1795 JH/10 Bug 840: fix log_defer_output option of pipe transport
1797 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1798 hosts_randomize. This matters a lot when combined with mua_wrapper.
1800 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1801 ${quote_pgsql:<string>} operator.
1803 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1805 JH/14 Change header file name for freeradius-client. Relevant if compiling
1806 with Radius support; from the Gentoo tree and checked under Fedora.
1808 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1810 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1811 option transport_filter, meaning no filtering.
1813 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1815 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1816 defaults to "*" (all hosts). The variable is now available when not built
1817 with TLS, default unset, mainly to enable keeping the testsuite sane.
1818 If a server certificate is not supplied (via tls_certificate) an error is
1819 logged, and clients will find TLS connections fail on startup. Presumably
1820 they will retry in-clear.
1821 Packagers of Exim are strongly encouraged to create a server certificate
1822 at installation time.
1824 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1825 with the $config_file variable.
1827 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1828 in transport context, after the attempt, and per-recipient. The latter type
1829 is per host attempted. The event data is the error message, and the errno
1830 information encodes the lookup type (A vs. MX) used for the (first) host,
1831 and the trailing two digits of the smtp 4xx response.
1833 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1834 to write to mainlog (or rejectlog, paniclog) in the window between file
1835 creation and permissions/ownership being changed. Particularly affects
1836 installations where exicyclog is run as root, rather than exim user;
1837 result is that the running daemon panics and dies.
1839 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1841 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1842 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1843 "pri" and "weight". Note that the previous implicit priority given by the
1844 list order is no longer honoured.
1846 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1847 for DKIM processing.
1849 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1850 by defining SUPPORT_SOCKS.
1852 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1853 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1854 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1855 variables are renamed to proxy_{local,external}_{address,port}.
1857 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1858 for a build by defining SUPPORT_I18N
1860 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1861 of the query string, and make ${quote_redis:} do that quoting.
1863 JH/29 Move Events support from Experimental to mainline, enabled by default
1864 and removable for a build by defining DISABLE_EVENT.
1866 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1868 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1869 cached by the daemon.
1871 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1872 by defining LOOKUP_REDIS. The libhiredis library is required.
1874 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1875 keys are given for lookup.
1877 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1878 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1879 only supported when built with TLS support. The PolarSSL SHA routines
1880 are still used when the TLS library is too old for convenient support.
1882 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1883 openssl_options), for security. OpenSSL forces this from version 1.1.0
1884 server-side so match that on older versions.
1886 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1887 allocation for $value could be released as the expansion processing
1888 concluded, but leaving the global pointer active for it.
1890 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1891 and to use the domains and local_parts ACL conditions.
1893 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1894 incorrectly not doubled on cutthrough transmission, hence seen as a
1895 body-termination at the receiving system - resulting in truncated mails.
1896 Commonly the sender saw a TCP-level error, and retransmitted the message
1897 via the normal store-and-forward channel. This could result in duplicates
1898 received - but deduplicating mailstores were liable to retain only the
1899 initial truncated version.
1901 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1903 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1905 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1906 we're in there, support oversigning also; bug 1309.
1908 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1910 HS/04 Add support for keep_environment and add_environment options.
1912 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1913 either intentional arithmetic overflow during PRNG, or testing config-
1916 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1917 delivery resulted in actual delivery. Cancel cutthrough before DATA
1920 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1921 rejects a recipient: pass the reject to the originator.
1923 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1924 Many were false-positives and ignorable, but it's worth fixing the
1927 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1928 for the new environment-manipulation done at startup. Move the routines
1929 from being local to tls.c to being global via the os.c file.
1931 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1932 an extract embedded as result-arg for a map, the first arg for extract
1933 is unavailable so we cannot tell if this is a numbered or keyed
1934 extraction. Accept either.
1940 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1943 JH/02 The smtp transport option "multi_domain" is now expanded.
1945 JH/03 The smtp transport now requests PRDR by default, if the server offers
1948 JH/04 Certificate name checking on server certificates, when exim is a client,
1949 is now done by default. The transport option tls_verify_cert_hostnames
1950 can be used to disable this per-host. The build option
1951 EXPERIMENTAL_CERTNAMES is withdrawn.
1953 JH/05 The value of the tls_verify_certificates smtp transport and main options
1954 default to the word "system" to access the system default CA bundle.
1955 For GnuTLS, only version 3.0.20 or later.
1957 JH/06 Verification of the server certificate for a TLS connection is now tried
1958 (but not required) by default. The verification status is now logged by
1959 default, for both outbound TLS and client-certificate supplying inbound
1962 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1965 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1966 Status Notification (bounce) messages are now MIME format per RFC 3464.
1967 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1968 under the control of the dsn_advertise_hosts option, and routers may
1969 have a dsn_lasthop option.
1971 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1972 default, modifiable by a malware= option. The list separator for
1973 the options can now be changed in the usual way. Bug 68.
1975 JH/10 The smtp_receive_timeout main option is now expanded before use.
1977 JH/11 The incoming_interface log option now also enables logging of the
1978 local interface on delivery outgoing connections.
1980 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1981 if the interface and destination host and port all match.
1983 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1986 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1987 Patch from Andrew Lewis.
1989 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1990 now supports optional time-restrictions, weighting, and priority
1991 modifiers per server. Patch originally by <rommer@active.by>.
1993 JH/16 The spamd_address main option now supports a mixed list of local
1994 and remote servers. Remote servers can be IPv6 addresses, and
1995 specify a port-range.
1997 JH/17 Bug 68: The spamd_address main option now supports an optional
1998 timeout value per server.
2000 JH/18 Bug 1581: Router and transport options headers_add/remove can
2001 now have the list separator specified.
2003 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2006 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2009 JH/21 Support for the A6 type of dns record is withdrawn.
2011 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2012 rather than the verbs used.
2014 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2015 from 255 to 1024 chars.
2017 JH/24 Verification callouts now attempt to use TLS by default.
2019 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2020 are generic router options now. The defaults didn't change.
2022 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2023 Original patch from Alexander Shikoff, worked over by JH.
2025 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2026 files if ZCAT_COMMAND is not executable.
2028 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2030 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2032 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2033 Normally benign, it bites when the pair was led to by a CNAME;
2034 modern usage is to not canonicalize the domain to a CNAME target
2035 (and we were inconsistent anyway for A-only vs AAAA+A).
2037 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2039 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2040 when evaluating $sender_host_dnssec.
2042 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2043 $sender_helo_dnssec variable.
2045 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2047 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2049 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2051 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2052 documented as working, but never had. Support all but $spam_report.
2054 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2055 added for tls authenticator.
2057 HS/03 Add perl_taintmode main config option
2063 TL/01 When running the test suite, the README says that variables such as
2064 no_msglog_check are global and can be placed anywhere in a specific
2065 test's script, however it was observed that placement needed to be near
2066 the beginning for it to behave that way. Changed the runtest perl
2067 script to read through the entire script once to detect and set these
2068 variables, reset to the beginning of the script, and then run through
2069 the script parsing/test process like normal.
2071 TL/02 The BSD's have an arc4random API. One of the functions to induce
2072 adding randomness was arc4random_stir(), but it has been removed in
2073 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2074 function when detected.
2076 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2077 cause callback expansion.
2079 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2080 syntax errors in an expansion can be treated as a string instead of
2081 logging or causing an error, due to the internal use of bool_lax
2082 instead of bool when processing it.
2084 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2085 server certificates when making smtp deliveries.
2087 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2089 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2091 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2093 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2094 Merged patch from Sebastian Wiedenroth.
2096 JH/05 Fix results-pipe from transport process. Several recipients, combined
2097 with certificate use, exposed issues where response data items split
2098 over buffer boundaries were not parsed properly. This eventually
2099 resulted in duplicates being sent. This issue only became common enough
2100 to notice due to the introduction of connection certificate information,
2101 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2103 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2104 size buffer was used, resulting in syntax errors when an expansion
2107 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2108 version 3.3.6 or later.
2110 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2111 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2112 both become "event_action", the variables become $event_name, $event_data
2113 and $event_defer_errno. There is a new variable $verify_mode, usable in
2114 routers, transports and related events. The tls:cert event is now also
2115 raised for inbound connections, if the main configuration event_action
2118 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2119 early OCSP support, but no stapling (appears to be less than 1.0.0).
2121 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2122 server certificate names available under the smtp transport option
2123 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2126 JH/10 Time-related extraction expansions from certificates now use the main
2127 option "timezone" setting for output formatting, and are consistent
2128 between OpenSSL and GnuTLS compilations. Bug 1541.
2130 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2131 encoded parameter in the incoming message. Bug 1558.
2133 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2134 include certificate info, eximon was claiming there were spoolfile
2137 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2139 JH/14 Log delivery-related information more consistently, using the sequence
2140 "H=<name> [<ip>]" wherever possible.
2142 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2143 are problematic for Debian distribution, omit them from the release
2146 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2148 JH/16 Fix string representation of time values on 64bit time_t architectures.
2151 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2152 output list separator was used.
2157 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2158 checkers that were complaining about end of non-void function with no
2161 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2162 This was a regression introduced in 4.83 by another bugfix.
2164 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2166 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2167 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2173 TF/01 Correctly close the server side of TLS when forking for delivery.
2175 When a message was received over SMTP with TLS, Exim failed to clear up
2176 the incoming connection properly after forking off the child process to
2177 deliver the message. In some situations the subsequent outgoing
2178 delivery connection happened to have the same fd number as the incoming
2179 connection previously had. Exim would try to use TLS and fail, logging
2180 a "Bad file descriptor" error.
2182 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2183 utilities have not been installed.
2185 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2186 temporary space as the ACL may create new global variables.
2188 TL/01 LDAP support uses per connection or global context settings, depending
2189 upon the detected version of the libraries at build time.
2191 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2192 to extract and use the src ip:port in logging and expansions as if it
2193 were a direct connection from the outside internet. PPv2 support was
2194 updated based on HAProxy spec change in May 2014.
2196 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2198 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2199 Properly escape header and check for NULL return.
2201 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2204 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2206 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2207 characters in header names, implemented as a verify condition.
2208 Contributed by Michael Fischer v. Mollard.
2210 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2211 results permerror and temperror. Previous values are deprecated but
2212 still accepted. In a future release, err_perm and err_temp will be
2213 completely removed, which will be a backward incompatibility if the
2214 ACL tests for either of these two old results. Patch contributed by
2215 user bes-internal on the mailing list.
2217 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2219 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2220 selectors, in both main and reject logs.
2222 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2225 JH/07 Add malware type "sock" for talking to simple daemon.
2227 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2229 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2230 routers/transports under cutthrough routing.
2232 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2233 numbers. Touch up "bool" conditional to keep the same definition.
2235 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2237 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2239 JH/12 Expand items in router/transport headers_add or headers_remove lists
2240 individually rather than the list as a whole. Bug 1452.
2242 Required for reasonable handling of multiple headers_ options when
2243 they may be empty; requires that headers_remove items with embedded
2244 colons must have them doubled (or the list-separator changed).
2246 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2247 view the policy declared in the DMARC record. Currently, $dmarc_status
2248 is a combined value of both the record presence and the result of the
2251 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2253 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2254 dnslookup router and the smtp transport (applying to the forward
2257 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2258 of ldap servers used for a specific lookup. Patch provided by Heiko
2261 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2262 New variable $lookup_dnssec_authenticated for observability.
2264 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2265 Patch submitted by Lars Timman.
2267 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2269 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2270 Requires trusted mode and valid format message id, aborts otherwise.
2271 Patch contributed by Heiko Schlichting.
2273 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2274 certextract with support for various fields. Bug 1358.
2276 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2277 is requested by default, modifiable by smtp transport option
2280 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2281 operate on certificate variables to give certificate fingerprints
2282 Also new ${sha256:cert_variable}.
2284 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2286 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2287 Christian Aistleitner.
2289 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2291 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2292 file. Patch from Wolfgang Breyha.
2294 JH/25 Expand the coverage of the delivery $host and $host_address to
2295 client authenticators run in verify callout. Bug 1476.
2297 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2298 align with daemon_smtp_ports. Bug 72.
2300 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2301 support and error reporting did not work properly.
2303 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2304 and is readable. Patch from Andrew Colin Kissa.
2306 TL/14 Enhance documentation of ${run expansion and how it parses the
2307 commandline after expansion, particularly in the case when an
2308 unquoted variable expansion results in an empty value.
2310 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2312 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2313 by renaming away from it. Addresses GH issue 15, reported by
2316 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2318 TL/15 SECURITY: prevent double expansion in math comparison functions
2319 (can expand unsanitized data). Not remotely exploitable.
2326 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2328 PP/02 Make -n do something, by making it not do something.
2329 When combined with -bP, the name of an option is not output.
2331 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2334 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2335 $sender_host_name and config options to manage this, and basic check
2338 PP/05 DSCP support for outbound connections and control modifier for inbound.
2340 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2341 (Only plugin which currently uses this is kerberos4, which nobody should
2342 be using, but we should make it available and other future plugins might
2343 conceivably use it, even though it would break NAT; stuff *should* be
2344 using channel bindings instead).
2346 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2347 name; added for Sendmail compatibility; requires admin caller.
2348 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2349 just ignore it); requires trusted caller.
2350 Also parse but ignore: -Ac -Am -X<logfile>
2353 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2355 TL/02 Add +smtp_confirmation as a default logging option.
2357 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2358 Patch by Magnus Holmgren from 2007-02-20.
2360 TL/04 Bugzilla 1281 - Spec typo.
2361 Bugzilla 1283 - Spec typo.
2362 Bugzilla 1290 - Spec grammar fixes.
2364 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2366 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2368 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2369 mailing list by by Dmitry Isaikin.
2371 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2373 JH/02 Support "G" suffix to numbers in ${if comparisons.
2375 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2377 NM/01 Bugzilla 1197 - Spec typo
2378 Bugzilla 1196 - Spec examples corrections
2380 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2382 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2383 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2386 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2387 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2389 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2390 "acl {{name}{arg}...}", and optional args on acl condition
2393 JH/05 Permit multiple router/transport headers_add/remove lines.
2395 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2397 JH/07 Avoid using a waiting database for a single-message-only transport.
2398 Performance patch from Paul Fisher. Bugzilla 1262.
2400 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2403 JH/09 Add $headers_added variable, with content from use of ACL modifier
2404 add_header (but not yet added to the message). Bugzilla 199.
2406 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2407 Pulled from Bugzilla 817 by Wolfgang Breyha.
2409 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2411 (nb: this is the same fix as in Exim 4.80.1)
2413 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2416 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2419 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2421 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2422 advertise SMTP AUTH mechanism to us, instead of a generic
2423 protocol violation error. Also, make Exim more robust to bad
2424 data from the Dovecot auth socket.
2426 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2428 When a queue runner is handling a message, Exim first routes the
2429 recipient addresses, during which it prunes them based on the retry
2430 hints database. After that it attempts to deliver the message to
2431 any remaining recipients. It then updates the hints database using
2434 So if a recipient address works intermittently, it can get repeatedly
2435 deferred at routing time. The retry hints record remains fresh so the
2436 address never reaches the final cutoff time.
2438 This is a fairly common occurrence when a user is bumping up against
2439 their storage quota. Exim had some logic in its local delivery code
2440 to deal with this. However it did not apply to per-recipient defers
2441 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2443 This change adds a proper retry rule check during routing so that the
2444 final cutoff time is checked against the message's age. We only do
2445 this check if there is an address retry record and there is not a
2446 domain retry record; this implies that previous attempts to handle
2447 the address had the retry_use_local_parts option turned on. We use
2448 this as an approximation for the destination being like a local
2449 delivery, as in LMTP.
2451 I suspect this new check makes the old local delivery cutoff check
2452 redundant, but I have not verified this so I left the code in place.
2454 TF/02 Correct gecos expansion when From: is a prefix of the username.
2456 Test 0254 submits a message to Exim with the header
2460 When I ran the test suite under the user fanf2, Exim expanded
2461 the header to contain my full name, whereas it should have added
2462 a Resent-Sender: header. It erroneously treats any prefix of the
2463 username as equal to the username.
2465 This change corrects that bug.
2467 GF/01 DCC debug and logging tidyup
2468 Error conditions log to paniclog rather than rejectlog.
2469 Debug lines prefixed by "DCC: " to remove any ambiguity.
2471 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2473 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2474 Bug spotted by Jeremy Harris; was flawed since initial commit.
2475 Would have resulted in OCSP responses post-SNI triggering an Exim
2476 NULL dereference and crash.
2478 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2480 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2481 Bug detection, analysis and fix by Samuel Thibault.
2482 Bugzilla 1331, Debian bug #698092.
2484 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2486 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2487 Server implementation by Todd Lyons, client by JH.
2488 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2489 config variable "prdr_enable" controls whether the server
2490 advertises the facility. If the client requests PRDR a new
2491 acl_data_smtp_prdr ACL is called once for each recipient, after
2492 the body content is received and before the acl_smtp_data ACL.
2493 The client is controlled by both of: a hosts_try_prdr option
2494 on the smtp transport, and the server advertisement.
2495 Default client logging of deliveries and rejections involving
2496 PRDR are flagged with the string "PRDR".
2498 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2499 fclose(). Diagnosis by Todd Lyons.
2501 PP/17 Update configure.default to handle IPv6 localhost better.
2502 Patch by Alain Williams (plus minor tweaks).
2505 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2506 This is now consistent with GnuTLS, and is now documented: the
2507 previous undocumented portable approach to treating the option as
2508 unset was to force an expansion failure. That still works, and
2509 an empty string is now equivalent.
2511 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2512 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2513 not performing validation itself.
2515 PP/20 Added force_command boolean option to pipe transport.
2516 Patch from Nick Koston, of cPanel Inc.
2518 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2521 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2523 PP/21 Fix eximon continuous updating with timestamped log-files.
2524 Broken in a format-string cleanup in 4.80, missed when I repaired the
2525 other false fix of the same issue.
2526 Report and fix from Heiko Schlichting.
2529 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2530 Report from Prashanth Katuri.
2532 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2533 It's SecureTransport, so affects any MacOS clients which use the
2534 system-integrated TLS libraries, including email clients.
2536 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2537 using a MIME ACL for non-SMTP local injection.
2538 Report and assistance in diagnosis by Warren Baker.
2540 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2542 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2544 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2545 last id that failed so it may be referenced in subsequent ACL's.
2547 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2550 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2551 ldap library initialization, allowing self-signed CA's to be
2552 used. Also properly sets require_cert option later in code by
2553 using NULL (global ldap config) instead of ldap handle (per
2554 session). Bug diagnosis and testing by alxgomz.
2556 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2557 the src/util/ subdirectory.
2559 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2560 renamed to Transport Post Delivery Action by Jeremy Harris, as
2563 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2564 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2565 redis_servers = needs to be configured which will be used by the redis
2566 lookup. Patch from Warren Baker, of The Packet Hub.
2568 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2570 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2571 hostname or reverse DNS when processing a host list. Used suggestions
2572 from multiple comments on this bug.
2574 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2576 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2577 Missed a few lines, added it to make the runtest require no keyboard
2580 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2581 contains upper case chars. Make router use caseful_local_part.
2583 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2584 support when GnuTLS has been built with p11-kit.
2590 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2592 This, or similar/improved, will also be change PP/11 of 4.82.
2598 PP/01 Handle short writes when writing local log-files.
2599 In practice, only affects FreeBSD (8 onwards).
2600 Bugzilla 1053, with thanks to Dmitry Isaikin.
2602 NM/01 Bugzilla 949 - Documentation tweak
2604 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2607 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2609 PP/02 Implemented gsasl authenticator.
2611 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2613 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2614 `pkg-config foo` for cflags/libs.
2616 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2617 with rest of GSASL and with heimdal_gssapi.
2619 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2620 `pkg-config foo` for cflags/libs for the TLS implementation.
2622 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2623 properties get this fed in as external SSF. A number of robustness
2624 and debugging improvements to the cyrus_sasl authenticator.
2626 PP/08 cyrus_sasl server now expands the server_realm option.
2628 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2629 Patch by Jeremy Harris.
2631 PP/10 Added dbmjz lookup type.
2633 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2635 PP/12 MAIL args handles TAB as well as SP, for better interop with
2636 non-compliant senders.
2637 Analysis and variant patch by Todd Lyons.
2639 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2640 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2641 Patch from Dirk Mueller <dmueller@suse.com>
2643 PP/13 tls_peerdn now print-escaped for spool files.
2644 Observed some $tls_peerdn in wild which contained \n, which resulted
2645 in spool file corruption.
2647 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2648 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2649 or write after TLS renegotiation, which otherwise led to messages
2652 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2653 as a tracking header (ie: a signed header comes before the signature).
2654 Patch from Wolfgang Breyha.
2656 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2657 comma-sep list; embedded commas doubled.
2659 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2661 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2663 Report and patch from Dmitry Banschikov.
2665 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2666 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2667 failures. We appear to now support TLS1.1+ with Exim.
2669 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2670 lets Exim select keys and certificates based upon TLS SNI from client.
2671 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2672 before an outbound SMTP session. New log_selector, +tls_sni.
2674 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2675 NULL dereference. Report and patch from Alun Jones.
2677 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2678 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2679 Not seeing resolver debug output on NetBSD, but suspect this is a
2680 resolver implementation change.
2682 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2683 Left warnings. Added "eximon gdb" invocation mode.
2685 PP/21 Defaulting "accept_8bitmime" to true, not false.
2687 PP/22 Added -bw for inetd wait mode support.
2689 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2690 locate the relevant includes and libraries. Made this the default.
2692 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2693 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2695 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2696 This may cause build issues on older platforms.
2698 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2699 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2700 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2701 Added SNI support via GnuTLS too.
2702 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2704 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2706 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2707 Applied second patch from Janne, implementing suggestion to default
2708 multiple-strings-in-record handling to match SPF spec.
2710 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2712 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2713 read-only, out of scope).
2714 Patch from Wolfgang Breyha, report from Stuart Northfield.
2716 PP/29 Fix three issues highlighted by clang analyser static analysis.
2717 Only crash-plausible issue would require the Cambridge-specific
2718 iplookup router and a misconfiguration.
2719 Report from Marcin Mirosław.
2721 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2723 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2724 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2725 As part of this, removing so much warning spew let me fix some minor
2726 real issues in debug logging.
2728 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2729 assignment on my part. Fixed.
2731 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2732 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2733 Janne Snabb (who went above and beyond: thank you).
2735 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2736 string otherwise requires a connection and a bunch more work and it's
2737 relatively easy to get wrong. Should also expose TLS library linkage
2740 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2741 64-bit ${eval} (JH/03).
2743 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2744 GNU libc to support some of the 64-bit stuff, should not lead to
2745 conflicts. Defined before os.h is pulled in, so if a given platform
2746 needs to override this, it can.
2748 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2749 protection layer was required, which is not implemented.
2750 Bugzilla 1254, patch from Wolfgang Breyha.
2752 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2753 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2754 tls_dhparam take prime identifiers. Also unbreak combination of
2755 OpenSSL+DH_params+TLSSNI.
2757 PP/39 Disable SSLv2 by default in OpenSSL support.
2763 PP/01 Solaris build fix for Oracle's LDAP libraries.
2764 Bugzilla 1109, patch from Stephen Usher.
2766 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2768 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2771 TF/02 Fix a couple more cases where we did not log the error message
2772 when unlink() failed. See also change 4.74-TF/03.
2774 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2775 lock up or crash if it happened to be inside a call to libc when it
2776 got a SIGUSR1 from exiwhat.
2778 The SIGUSR1 handler appends the current process status to the process
2779 log which is later printed by exiwhat. It used to use the general
2780 purpose logging code to do this, but several functions it calls are
2781 not safe for signals.
2783 The new output code in the SIGUSR1 handler is specific to the process
2784 log, and simple enough that it's easy to inspect for signal safety.
2785 Removing some special cases also simplifies the general logging code.
2786 Removing the spurious timestamps from the process log simplifies
2789 TF/04 Improved ratelimit ACL condition.
2791 The /noupdate option has been deprecated in favour of /readonly which
2792 has clearer semantics. The /leaky, /strict, and /readonly update modes
2793 are mutually exclusive. The update mode is no longer included in the
2794 database key; it just determines when the database is updated. (This
2795 means that when you upgrade Exim will forget old rate measurements.)
2797 Exim now checks that the per_* options are used with an update mode that
2798 makes sense for the current ACL. For example, when Exim is processing a
2799 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2800 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2801 must specify per_mail/readonly. If you omit the update mode it defaults to
2802 /leaky where that makes sense (as before) or /readonly where required.
2804 The /noupdate option is now undocumented but still supported for
2805 backwards compatibility. It is equivalent to /readonly except that in
2806 ACLs where /readonly is required you may specify /leaky/noupdate or
2807 /strict/noupdate which are treated the same as /readonly.
2809 A useful new feature is the /count= option. This is a generalization
2810 of the per_byte option, so that you can measure the throughput of other
2811 aggregate values. For example, the per_byte option is now equivalent
2812 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2814 The per_rcpt option has been generalized using the /count= mechanism
2815 (though it's more complicated than the per_byte equivalence). When it is
2816 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2817 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2818 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2819 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2820 Note that using per_rcpt with a non-readonly update mode in more than
2821 one ACL will cause the recipients to be double-counted. (The per_mail
2822 and per_byte options don't have this problem.)
2824 The handling of very low rates has changed slightly. If the computed rate
2825 is less than the event's count (usually one) then this event is the first
2826 after a long gap. In this case the rate is set to the same as this event's
2827 count, so that the first message of a spam run is counted properly.
2829 The major new feature is a mechanism for counting the rate of unique
2830 events. The new per_addr option counts the number of different
2831 recipients that someone has sent messages to in the last time period. It
2832 behaves like per_rcpt if all the recipient addresses are different, but
2833 duplicate recipient addresses do not increase the measured rate. Like
2834 the /count= option this is a general mechanism, so the per_addr option
2835 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2836 example, measure the rate that a client uses different sender addresses
2837 with the options per_mail/unique=$sender_address. There are further
2838 details in the main documentation.
2840 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2842 TF/06 Removed a few PCRE remnants.
2844 TF/07 Automatically extract Exim's version number from tags in the git
2845 repository when doing development or release builds.
2847 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2848 Bugzilla 879. Patch from Paul Fisher.
2850 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2851 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2854 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2856 PP/05 Variable $av_failed, true if the AV scanner deferred.
2857 Bugzilla 1078. Patch from John Horne.
2859 PP/06 Stop make process more reliably on build failure.
2860 Bugzilla 1087. Patch from Heiko Schlittermann.
2862 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2863 Bugzilla 1089. Patch from Heiko Schlittermann.
2865 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2866 Bugzilla 1131. Patch from Holger Weiß.
2868 PP/09 Handle IPv6 addresses with SPF.
2869 Bugzilla 860. Patch from Wolfgang Breyha.
2871 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2873 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2876 PP/11 match_* no longer expand right-hand-side by default.
2877 New compile-time build option, EXPAND_LISTMATCH_RHS.
2878 New expansion conditions, "inlist", "inlisti".
2880 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2882 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2884 PP/14 fix log_write() format string regression from TF/03.
2885 Bugzilla 1152. Patch from Dmitry Isaikin.
2891 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2893 PP/02 Harmonised TLS library version reporting; only show if debugging.
2894 Layout now matches that introduced for other libraries in 4.74 PP/03.
2896 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2898 PP/04 New "dns_use_edns0" global option.
2900 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2903 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2904 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2906 TK/01 Updated PolarSSL code to 0.14.2.
2907 Bugzilla 1097. Patch from Andreas Metzler.
2909 PP/07 Catch divide-by-zero in ${eval:...}.
2910 Fixes bugzilla 1102.
2912 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2915 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2916 format-string attack -- SECURITY: remote arbitrary code execution.
2918 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2919 time unintentionally subject to list matching rules, letting the header
2920 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2921 arbitrary string expansion). This allowed for information disclosure.
2923 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2924 INT_MIN/-1 -- value coerced to INT_MAX.
2930 NM/01 Workaround for PCRE version dependency in version reporting
2933 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2934 This fixes portability to compilers other than gcc, notably
2935 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2937 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2938 makefiles for portability to HP-UX and POSIX correctness.
2940 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2941 Also via indented variable definition in the Makefile.
2942 (Debugging by Oliver Heesakkers).
2944 PP/02 Restore caching of spamd results with expanded spamd_address.
2945 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2947 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2948 Improves build reliability. Fix from: Frank Elsner
2950 NM/02 Fix wide character breakage in the rfc2047 coding
2951 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2953 NM/03 Allow underscore in dnslist lookups
2954 Fixes bug 1026. Patch from Graeme Fowler
2956 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2957 Code patches from Adam Ciarcinski of NetBSD.
2959 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2962 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2963 is logged, to avoid truncation. Patch from John Horne.
2965 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2966 Patch from Jakob Hirsch.
2968 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2969 SQL string expansion failure details.
2970 Patch from Andrey Oktyabrski.
2972 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2973 Patch from Simon Arlott.
2975 PP/09 New lookups functionality failed to compile on old gcc which rejects
2976 extern declarations in function scope.
2977 Patch from Oliver Fleischmann
2979 PP/10 Use sig_atomic_t for flags set from signal handlers.
2980 Check getgroups() return and improve debugging.
2981 Fixed developed for diagnosis in bug 927 (which turned out to be
2984 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2985 Patch from Mark Zealey.
2987 PP/12 Bugzilla 1056: Improved spamd server selection.
2988 Patch from Mark Zealey.
2990 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2991 Based on patch from Heiko Schlittermann.
2993 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2994 Patch from Uwe Doering, sign-off by Michael Haardt.
2996 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2997 variable declaration deep within a block. Bug and patch from
3000 PP/15 lookups-Makefile IRIX compatibility coercion.
3002 PP/16 Make DISABLE_DKIM build knob functional.
3004 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3005 Patch by Simon Arlott
3007 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3008 variable argument macros. Our copy now differs from upstream.
3014 TF/01 Failure to get a lock on a hints database can have serious
3015 consequences so log it to the panic log.
3017 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3018 controlled using the smtp_confirmation log selector.
3020 TF/03 Include the error message when we fail to unlink a spool file.
3022 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3023 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3024 for maintaining out-of-tree patches for some time.
3026 PP/01 Bugzilla 139: Documentation and portability issues.
3027 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3028 Handle per-OS dynamic-module compilation flags.
3030 PP/02 Let /dev/null have normal permissions.
3031 The 4.73 fixes were a little too stringent and complained about the
3032 permissions on /dev/null. Exempt it from some checks.
3033 Reported by Andreas M. Kirchwitz.
3035 PP/03 Report version information for many libraries, including
3036 Exim version information for dynamically loaded libraries. Created
3037 version.h, now support a version extension string for distributors
3038 who patch heavily. Dynamic module ABI change.
3040 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3041 privilege escalation vulnerability whereby the Exim run-time user
3042 can cause root to append content of the attacker's choosing to
3045 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3048 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3049 If dropping privileges for untrusted macros, we disabled normal logging
3050 on the basis that it would fail; for the Exim run-time user, this is not
3051 the case, and it resulted in successful deliveries going unlogged.
3052 Fixed. Reported by Andreas Metzler.
3058 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3059 only prepend for the Resent-* case. Fixes regression introduced in
3060 Exim 4.70 by NM/22 for Bugzilla 607.
3062 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3063 increasing numbers of administrators be bitten by this.
3065 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3067 PP/03 Bugzilla 994: added openssl_options main configuration option.
3069 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3071 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3073 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3075 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3076 without a peer certificate, leading to a segfault because of an
3077 assumption that peers always have certificates. Be a little more
3078 paranoid. Problem reported by Martin Tscholak.
3080 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3081 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3082 NB: ClamAV planning to remove STREAM in "middle of 2010".
3083 CL also introduces -bmalware, various -d+acl logging additions and
3084 more caution in buffer sizes.
3086 PP/09 Implemented reverse_ip expansion operator.
3088 PP/10 Bugzilla 937: provide a "debug" ACL control.
3090 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3092 PP/12 Bugzilla 973: Implement --version.
3094 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3096 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3098 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3100 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3101 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3102 ignore trailing whitespace.
3104 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3106 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3109 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3110 Notification from Dr Andrew Aitchison.
3112 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3113 ExtendedDetectionInfo response format.
3114 Notification from John Horne.
3116 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3119 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3120 XSL and documented dependency on system catalogs, with examples of how
3123 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3126 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3127 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3128 configuration file which is writeable by the Exim user or group.
3130 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3131 of configuration files to cover files specified with the -C option if
3132 they are going to be used with root privileges, not just the default
3135 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3136 option (effectively making it always true).
3138 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3139 files to be used while preserving root privileges.
3141 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3142 that rogue child processes cannot use them.
3144 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3145 run-time user, instead of root.
3147 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3148 Exim run-time user without dropping privileges.
3150 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3151 result string, instead of calling string_vformat() twice with the same
3154 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3155 for other users. Others should always drop root privileges if they use
3156 -C on the command line, even for a whitelisted configure file.
3158 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3160 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3166 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3167 $data_path, and $header_path variables; fixed documentation bugs and
3170 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3171 exipick to access non-standard spools, including the "frozen" queue
3174 NM/01 Bugzilla 965: Support mysql stored procedures.
3175 Patch from Alain Williams
3177 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3179 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3180 Patch from Andreas Metzler
3182 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3183 Patch from Kirill Miazine
3185 NM/05 Bugzilla 671: Added umask to procmail example.
3187 JJ/03 installed exipick 20100323.0, fixing doc bug
3189 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3190 directory. Notification and patch from Dan Rosenberg.
3192 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3194 TK/02 Improve log output when DKIM signing operation fails.
3196 MH/01 Treat the transport option dkim_domain as a colon separated
3197 list, not as a single string, and sign the message with each element,
3198 omitting multiple occurences of the same signer.
3200 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3201 Bugzilla 985, 986. Patch by Simon Arlott
3203 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3204 Patch by Simon Arlott
3206 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3207 MBX locking. Notification from Dan Rosenberg.
3213 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3215 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3217 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3219 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3221 NM/04 Bugzilla 915: Flag broken perl installation during build.
3227 TK/01 Added patch by Johannes Berg that expands the main option
3228 "spamd_address" if it starts with a dollar sign.
3230 TK/02 Write list of recipients to X-Envelope-Sender header when building
3231 the mbox-format spool file for content scanning (suggested by Jakob
3234 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3235 (http://www.dcc-servers.net/) support via dccifd. Activated by
3236 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3238 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3239 by Mark Daniel Reidel <mr@df.eu>.
3241 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3242 When building exim an external PCRE library is now needed -
3243 PCRE is a system library on the majority of modern systems.
3244 See entry on PCRE_LIBS in EDITME file.
3246 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3247 conversation. Added nologin parameter to request.
3248 Patch contributed by Kirill Miazine.
3250 TF/01 Do not log submission mode rewrites if they do not change the address.
3252 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3254 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3255 log files in place. Contributed by Roberto Lima.
3257 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3259 TF/03 Bugzilla 615: When checking the local_parts router precondition
3260 after a local_part_suffix or local_part_prefix option, Exim now
3261 does not use the address's named list lookup cache, since this
3262 contains cached lookups for the whole local part.
3264 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3265 Robert Millan. Documentation is in experimental-spec.txt.
3267 TF/04 Bugzilla 668: Fix parallel build (make -j).
3269 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3271 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3272 Patch provided by Jan Srzednicki.
3274 TF/05 Leading white space used to be stripped from $spam_report which
3275 wrecked the formatting. Now it is preserved.
3277 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3278 that they are available at delivery time.
3280 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3282 TF/08 TLS error reporting now respects the incoming_interface and
3283 incoming_port log selectors.
3285 TF/09 Produce a more useful error message if an SMTP transport's hosts
3286 setting expands to an empty string.
3288 NM/06 Bugzilla 744: EXPN did not work under TLS.
3289 Patch provided by Phil Pennock.
3291 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3292 Patch provided by Richard Godbee.
3294 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3295 acl_smtp_notquit, added index entry.
3297 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3298 Patch provided by Eugene Bujak.
3300 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3301 accept(). Patch provided by Maxim Dounin.
3303 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3304 Patch provided by Phil Pennock.
3306 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3308 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3309 Patch provided by Brad "anomie" Jorsch.
3311 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3312 Patch provided by Dean Brooks.
3314 TK/05 Add native DKIM support (does not depend on external libraries).
3316 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3317 Patch provided by Graeme Fowler.
3319 NM/16 Bugzilla 851: Documentation example syntax fix.
3321 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3323 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3326 NM/19 Bugzilla 745: TLS version reporting.
3327 Patch provided by Phil Pennock.
3329 NM/20 Bugzilla 167: bool: condition support.
3330 Patch provided by Phil Pennock.
3332 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3333 clients. Patch provided by Phil Pennock.
3335 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3336 Patch provided by Brad "anomie" Jorsch.
3338 NM/23 Bugzilla 687: Fix misparses in eximstats.
3339 Patch provided by Heiko Schlittermann.
3341 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3342 Patch provided by Heiko Schlittermann.
3344 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3345 plus update to original patch.
3347 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3349 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3350 Patch provided by David Brownlee.
3352 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3354 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3356 NM/30 Bugzilla 888: TLS documentation bugfixes.
3358 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3360 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3361 Unlike the original bugzilla I have changed all shell scripts in src tree.
3363 NM/33 Bugzilla 898: Transport filter timeout fix.
3364 Patch by Todd Rinaldo.
3366 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3367 Patch by Serge Demonchaux.
3369 NM/35 Bugzilla 39: Base64 decode bug fixes.
3370 Patch by Jakob Hirsch.
3372 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3374 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3376 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3378 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3384 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3385 ALT-N's libdkim that I have put here:
3386 http://duncanthrax.net/exim-experimental/
3388 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3389 were called 'true' and it seems that C99 defines that as a reserved
3390 keyword to be used with 'bool' variable types. That means you could
3391 not include C99-style headers which use bools without triggering
3392 build errors in sieve.c.
3394 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3395 as mailq or other aliases. Changed the --help handling significantly
3396 to do whats expected. exim_usage() emits usage/help information.
3398 SC/01 Added the -bylocaldomain option to eximstats.
3400 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3402 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3404 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3410 PH/01 Another patch from the Sieve maintainer.
3412 PH/02 When an IPv6 address is converted to a string for single-key lookup
3413 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3414 dots are used instead of colons so that keys in lsearch files need not
3415 contain colons. This was done some time before quoting was made available
3416 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3417 (notated using the quote facility) so as to distinguish them from IPv4
3418 keys. This meant that lookups for IP addresses in host lists did not work
3419 for iplsearch lookups.
3421 This has been fixed by arranging for IPv6 addresses to be expressed with
3422 colons if the lookup type is iplsearch. This is not incompatible, because
3423 previously such lookups could never work.
3425 The situation is now rather anomalous, since one *can* have colons in
3426 ordinary lsearch keys. However, making the change in all cases is
3427 incompatible and would probably break a number of configurations.
3429 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3432 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3433 conversion specification without a maximum field width, thereby enabling
3434 a rogue spamd server to cause a buffer overflow. While nobody in their
3435 right mind would setup Exim to query an untrusted spamd server, an
3436 attacker that gains access to a server running spamd could potentially
3437 exploit this vulnerability to run arbitrary code as the Exim user.
3439 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3440 $primary_hostname instead of what libspf2 thinks the hosts name is.
3442 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3443 a directory entry by the name of the lookup key. Previously, if a
3444 symlink pointed to a non-existing file or a file in a directory that
3445 Exim lacked permissions to read, a lookup for a key matching that
3446 symlink would fail. Now it is enough that a matching directory entry
3447 exists, symlink or not. (Bugzilla 503.)
3449 PH/03 The body_linecount and body_zerocount variables are now exported in the
3452 PH/04 Added the $dnslist_matched variable.
3454 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3455 This means they are set thereafter only if the connection becomes
3458 PH/06 Added the client_condition to authenticators so that some can be skipped
3459 by clients under certain conditions.
3461 PH/07 The error message for a badly-placed control=no_multiline_responses left
3462 "_responses" off the end of the name.
3464 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3466 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3467 (without spaces) instead of just copying the configuration text.
3469 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3471 PH/11 Added $max_received_linelength.
3473 PH/12 Added +ignore_defer and +include_defer to host lists.
3475 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3476 way in which PCRE > 7.0 is built.
3478 PH/14 Implemented queue_only_load_latch.
3480 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3481 MAIL command. The effect was to mangle the value on 64-bit systems.
3483 PH/16 Another patch from the Sieve maintainer.
3485 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3487 PH/18 If a system quota error occurred while trying to create the file for
3488 a maildir delivery, the message "Mailbox is full" was not appended to the
3489 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3490 only to a quota excession during the actual writing of the file.
3492 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3493 characters?) which causes problems in log lines. The DN values are now
3494 passed through string_printing() before being added to log lines.
3496 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3497 and InterBase are left for another time.)
3499 PH/21 Added message_body_newlines option.
3501 PH/22 Guard against possible overflow in moan_check_errorcopy().
3503 PH/23 POSIX allows open() to be a macro; guard against that.
3505 PH/24 If the recipient of an error message contained an @ in the local part
3506 (suitably quoted, of course), incorrect values were put in $domain and
3507 $local_part during the evaluation of errors_copy.
3513 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3514 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3517 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3518 issue a MAIL command.
3520 PH/02 In an ACL statement such as
3522 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3524 if a client was not listed at all, or was listed with a value other than
3525 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3526 the condition was not true (as it should be), so access was not denied.
3527 The bug was that the ! inversion was incorrectly passed on to the second
3528 item. This has been fixed.
3530 PH/03 Added additional dnslists conditions == and =& which are different from
3531 = and & when the dns lookup returns more than one IP address.
3533 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3534 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3536 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3537 FSYNC, which compiles an option called disable_fsync that allows for
3538 bypassing fsync(). The documentation is heavily laced with warnings.
3540 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3542 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3543 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3544 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3545 including adding "make clean"; (3) Added -fPIC when compiling the test
3546 dynamically loaded module, to get rid of a warning.
3548 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3549 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3550 = 0s. The bug is otherwise harmless.
3552 PH/07 There was a bug in the dovecot authenticator such that the value of
3553 $auth1 could be overwritten, and so not correctly preserved, after a
3554 successful authentication. This usually meant that the value preserved by
3555 the server_setid option was incorrect.
3557 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3559 PH/09 Installed PCRE release 7.0.
3561 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3562 run for batched SMTP input. It is now run at the start of every message
3563 in the batch. While fixing this I discovered that the process information
3564 (output by running exiwhat) was not always getting set for -bs and -bS
3565 input. This is fixed, and it now also says "batched" for BSMTP.
3567 PH/11 Added control=no_pipelining.
3569 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3570 patch, slightly modified), and move the expansion of helo_data till after
3571 the connection is made in the smtp transport (so it can use these
3574 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3576 PH/14 Added log_selector = +pid.
3578 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3580 PH/16 Add ${if forany and ${if forall.
3582 PH/17 Added dsn_from option to vary the From: line in DSNs.
3584 PH/18 Flush SMTP output before performing a callout, unless control =
3585 no_callout_flush is set.
3587 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3588 was true (the default) a successful delivery failed to delete the retry
3589 item, thus causing premature timeout of the address. The bug is now
3592 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3594 PH/21 Long custom messages for fakedefer and fakereject are now split up
3595 into multiline responses in the same way that messages for "deny" and
3596 other ACL rejections are.
3598 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3599 with slight modification.
3601 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3602 draft, changing the syntax and factoring some duplicate code".
3604 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3605 for deliveries of the second and subsequent messages over the same SMTP
3608 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3609 ${reduce, with only minor "tidies".
3611 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3613 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3614 expansion side effects.
3616 PH/27 When a message times out after an over-quota error from an Exim-imposed
3617 quota, the bounce message says "mailbox is full". This message was not
3618 being given when it was a system quota that was exceeded. It now should
3621 MH/03 Made $recipients available in local_scan(). local_scan() already has
3622 better access to the recipient list through recipients_list[], but
3623 $recipients can be useful in postmaster-provided expansion strings.
3625 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3626 in the case of a MAIL command with additional options following the
3627 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3628 were accidentally chopped off.
3630 PH/29 SMTP synchronization checks are implemented when a command is read -
3631 there is a check that no more input is waiting when there shouldn't be
3632 any. However, for some commands, a delay in an ACL can mean that it is
3633 some time before the response is written. In this time, more input might
3634 arrive, invalidly. So now there are extra checks after an ACL has run for
3635 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3636 pipelining has not been advertised.
3638 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3640 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3641 returned in the SMTP response when smtp_return_error_details was false.
3642 This has been fixed.
3644 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3645 instead of the C I/O that was originally supplied, because problems were
3646 reported on Solaris.
3648 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3649 Exim which did not show up earlier: it was assuming that a call to
3650 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3651 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3652 was a macro that became an assignment, so it seemed to work. This has
3653 changed to a proper function call with a void return, hence the compile
3654 error. Exim's code has been fixed.
3656 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3659 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3661 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3663 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3664 because it turns out to be too big or there is a timeout) there is no
3665 "Completed" line in the log. When some messages of this type were
3666 selected by exigrep, they were listed as "not completed". Others were
3667 picked up by some special patterns. I have improved the selection
3668 criteria to be more general.
3670 PH/38 The host_find_failed option in the manualroute router can now be set
3671 to "ignore", to completely ignore a host whose IP address cannot be
3672 found. If all hosts are ignored, the behaviour is controlled by the new
3673 host_all_ignored option.
3675 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3676 homing or because of multiple MX records with /mx) generated more than
3677 one IP address, and the following item turned out to be the local host,
3678 all the secondary addresses of the first item were incorrectly removed
3679 from the list, along with the local host and any following hosts (which
3680 is what is supposed to happen).
3682 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3683 whoever called Exim into the -H file. In the case of the daemon it was
3684 behaving confusingly. When first started, it used values for whoever
3685 started the daemon, but after a SIGHUP it used the Exim user (because it
3686 calls itself on a restart). I have changed the code so that it now always
3689 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3690 message are rejected with the same error (e.g. no authentication or bad
3691 sender address), and a DATA command is nevertheless sent (as can happen
3692 with PIPELINING or a stupid MUA), the error message that was given to the
3693 RCPT commands is included in the rejection of the DATA command. This is
3694 intended to be helpful for MUAs that show only the final error to their
3697 PH/42 Another patch from the Sieve maintainer.
3699 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3700 Eximstats - Fixed some broken HTML links and added missing column headers
3702 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3703 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3705 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3707 PH/43 Yet another patch from the Sieve maintainer.
3709 PH/44 I found a way to check for a TCP/IP connection going away before sending
3710 the response to the final '.' that terminates a message, but only in the
3711 case where the client has not sent further data following the '.'
3712 (unfortunately, this is allowed). However, in many cases there won't be
3713 any further data because there won't be any more messages to send. A call
3714 to select() can be used: if it shows that the input is "ready", there is
3715 either input waiting, or the socket has been closed. An attempt to read
3716 the next input character can distinguish the two cases. Previously, Exim
3717 would have sent an OK response which the client would never have see.
3718 This could lead to message repetition. This fix should cure that, at
3719 least in a lot of common cases.
3721 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3722 advertised in response to EHLO.
3728 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3729 fixed by 4.65/MH/01 (is this a record?) are fixed:
3731 (i) An empty string was always treated as zero by the numeric comparison
3732 operators. This behaviour has been restored.
3734 (ii) It is documented that the numeric comparison operators always treat
3735 their arguments as decimal numbers. This was broken in that numbers
3736 starting with 0 were being interpreted as octal.
3738 While fixing these problems I realized that there was another issue that
3739 hadn't been noticed. Values of message_size_limit (both the global option
3740 and the transport option) were treated as octal if they started with 0.
3741 The documentation was vague. These values are now always treated as
3742 decimal, and I will make that clear in the documentation.
3748 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3749 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3752 MH/01 Don't check that the operands of numeric comparison operators are
3753 integers when their expansion is in "skipping" mode (fixes bug
3754 introduced by 4.64-PH/07).
3756 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3757 child addresses, Exim now panics and dies. Previously, because the count
3758 is held in a short int, deliveries were likely to be lost. As such a
3759 large number of recipients for a single message is ridiculous
3760 (performance will be very, very poor), I have chosen to impose a limit
3761 rather than extend the field.
3767 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3768 leftover -K file (the existence of which was triggered by #402).
3769 While we were at it, introduced process PID as part of the -K
3770 filename. This should rule out race conditions when creating
3773 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3774 processing considerably. Previous code took too long for large mails,
3775 triggering a timeout which in turn triggers #401.
3777 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3778 in the DK code in transports.c. sendfile() is not really portable,
3779 hence the _LINUX specificness.
3781 TF/01 In the add_headers option to the mail command in an Exim filter,
3782 there was a bug that Exim would claim a syntax error in any
3783 header after the first one which had an odd number of characters
3786 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3787 callout verification, Exim cached a "reject" for the entire domain. This
3788 is correct for most verifications, but it is not correct for a recipient
3789 verification with use_sender or use_postmaster set, because in that case
3790 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3791 case of MAIL FROM:<> rejection from other early rejections (e.g.
3792 rejection of HELO). When verifying a recipient using a non-null MAIL
3793 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3794 Whatever the result of the callout, the value of the domain cache is
3795 left unchanged (for any other kind of callout, getting as far as trying
3796 RCPT means that the domain itself is ok).
3798 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3801 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3802 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3803 session, the daemon ignores ECONNECT errors and logs others; it now
3804 ignores EPIPE as well.
3806 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3807 (quoted-printable decoding).
3809 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3810 later the small subsequent patch to fix an introduced bug.
3812 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3814 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3816 PH/08 An error is now given if message_size_limit is specified negative.
3818 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3819 to be given (somewhat) arbitrary names.
3821 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3824 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3825 miscellaneous code fixes
3827 PH/10 Added the log_reject_target ACL modifier to specify where to log
3830 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3831 hostname. This is wrong, because it relates to the incoming message (and
3832 probably the interface on which it is arriving) and not to the outgoing
3833 callout (which could be using a different interface). This has been
3834 changed to use the value of the helo_data option from the smtp transport
3835 instead - this is what is used when a message is actually being sent. If
3836 there is no remote transport (possible with a router that sets up host
3837 addresses), $smtp_active_hostname is used.
3839 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3840 tweaks were necessary in order to get it to work (see also 21 below):
3841 (a) The code assumed that strncpy() returns a negative number on buffer
3842 overflow, which isn't the case. Replaced with Exim's string_format()
3844 (b) There were several signed/unsigned issues. I just did the minimum
3845 hacking in of casts. There is scope for a larger refactoring.
3846 (c) The code used strcasecmp() which is not a standard C function.
3847 Replaced with Exim's strcmpic() function.
3848 (d) The code set only $1; it now sets $auth1 as well.
3849 (e) A simple test gave the error "authentication client didn't specify
3850 service in request". It would seem that Dovecot has changed its
3851 interface. Fortunately there's a specification; I followed it and
3852 changed what the client sends and it appears to be working now.
3854 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3857 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3858 address A is aliased to B and C, where B exists and C does not. Without
3859 -v the output is "A verified" because verification stops after a
3860 successful redirection if more than one address is generated. However,
3861 with -v the child addresses are also verified. Exim was outputting "A
3862 failed to verify" and then showing the successful verification for C,
3863 with its parentage. It now outputs "B failed to verify", showing B's
3864 parentage before showing the successful verification of C.
3866 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3867 look up a TXT record in a specific list after matching in a combined
3870 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3871 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3872 they consult the DNS. I had assumed they would set it the way they
3873 wanted; and indeed my experiments on Linux seem to show that in some
3874 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3875 To be on the safe side, however, I have now made the interface to
3876 host_find_byname() similar to host_find_bydns(), with an argument
3877 containing the DNS resolver options. The host_find_byname() function now
3878 sets these options at its start, just as host_find_bydns() does. The smtp
3879 transport options dns_qualify_single and dns_search_parents are passed to
3880 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3881 of host_find_byname() use the default settings of RES_DEFNAMES
3882 (qualify_single) but not RES_DNSRCH (search_parents).
3884 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3885 spool_read_header() do less string testing, by means of a preliminary
3886 switch on the second character of optional "-foo" lines. (This is
3887 overdue, caused by the large number of possibilities that now exist.
3888 Originally there were few.) While I was there, I also converted the
3889 str(n)cmp tests so they don't re-test the leading "-" and the first
3890 character, in the hope this might squeeze out yet more improvement.
3892 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3893 flag allowing group syntax was set by the header_syntax check but not
3894 turned off, possible causing trouble later; (2) The flag was not being
3895 set at all for the header_verify test, causing "group"-style headers to
3896 be rejected. I have now set it in this case, and also caused header_
3897 verify to ignore an empty address taken from a group. While doing this, I
3898 came across some other cases where the code for allowing group syntax
3899 while scanning a header line wasn't quite right (mostly, not resetting
3900 the flag correctly in the right place). These bugs could have caused
3901 trouble for malformed header lines. I hope it is now all correct.
3903 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3904 with the "reply" argument non-NULL. The code, however (which originally
3905 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3906 but it didn't always do it. This confused somebody who was copying the
3907 code for some other use. I have removed all the tests.
3909 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3910 feature that was used to support insecure browsers during the U.S. crypto
3911 embargo. It requires special client support, and Exim is probably the
3912 only MTA that supported it -- and would never use it because real RSA is
3913 always available. This code has been removed, because it had the bad
3914 effect of slowing Exim down by computing (never used) parameters for the
3915 RSA_EXPORT functionality.
3917 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3918 authenticator to fail if there's a tab character in the incoming data
3919 (there should never be unless someone is messing about, as it's supposed
3920 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3921 if the connection is using TLS or if the remote IP is the same as the
3922 local IP, and the "valid-client-cert option" if a client certificate has
3925 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3926 authenticators. This can be used for authorization after authentication
3927 succeeds. (In the case of plaintext, it servers for both authentication
3930 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3931 if any retry times were supplied.
3933 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3934 connection, where there is no client IP address to check. In this
3935 situation, the verify now always succeeds.
3937 PH/25 Applied John Jetmore's -Mset patch.
3939 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3941 PH/27 In a string expansion for a processed (not raw) header when multiple
3942 headers of the same name were present, leading whitespace was being
3943 removed from all of them, but trailing whitespace was being removed only
3944 from the last one. Now trailing whitespace is removed from each header
3945 before concatenation. Completely empty headers in a concatenation (as
3946 before) are ignored.
3948 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3949 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3951 PH/29 [Removed. This was a change that I later backed out, and forgot to
3952 correct the ChangeLog entry (that I had efficiently created) before
3953 committing the later change.]
3955 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3956 address errors (4xx response to RCPT) over the same connection as other
3957 messages routed to the same hosts. Such deliveries are always "forced",
3958 so retry times are not inspected. This resulted in far too many retries
3959 for the affected addresses. The effect occurred only when there were more
3960 hosts than the hosts_max_try setting in the smtp transport when it had
3961 the 4xx errors. Those hosts that it had tried were not added to the list
3962 of hosts for which the message was waiting, so if all were tried, there
3963 was no problem. Two fixes have been applied:
3965 (i) If there are any address or message errors in an SMTP delivery, none
3966 of the hosts (tried or untried) are now added to the list of hosts
3967 for which the message is waiting, so the message should not be a
3968 candidate for sending over the same connection that was used for a
3969 successful delivery of some other message. This seems entirely
3970 reasonable: after all the message is NOT "waiting for some host".
3971 This is so "obvious" that I'm not sure why it wasn't done
3972 previously. Hope I haven't missed anything, but it can't do any
3973 harm, as the worst effect is to miss an optimization.
3975 (ii) If, despite (i), such a delivery is accidentally attempted, the
3976 routing retry time is respected, so at least it doesn't keep
3977 hammering the server.
3979 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3980 in ${readsocket because some servers need this prod.
3982 PH/32 Added some extra debug output when updating a wait-xxx database.
3984 PH/33 The hint "could be header name not terminated by colon", which has been
3985 given for certain expansion errors for a long time, was not being given
3986 for the ${if def:h_colon_omitted{... case.
3988 PH/34 The spec says: "With one important exception, whenever a domain list is
3989 being scanned, $domain contains the subject domain." There was at least
3990 one case where this was not true.
3992 PH/35 The error "getsockname() failed: connection reset by peer" was being
3993 written to the panic log as well as the main log, but it isn't really
3994 panic-worthy as it just means the connection died rather early on. I have
3995 removed the panic log writing for the ECONNRESET error when getsockname()
3998 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3999 runs only) independently of the message's sender address. This meant
4000 that, if the 4xx error was in fact related to the sender, a different
4001 message to the same recipient with a different sender could confuse
4002 things. In particular, this can happen when sending to a greylisting
4003 server, but other circumstances could also provoke similar problems.
4004 I have changed the default so that the retry time for these errors is now
4005 based a combination of the sender and recipient addresses. This change
4006 can be overridden by setting address_retry_include_sender=false in the
4009 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4010 remote server are returned as part of bounce messages. This was not
4011 happening for LMTP over a pipe (the lmtp transport), but now it is the
4012 same for both kinds of LMTP.
4014 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4015 in header lines that were in fact CNAMEs. This is no longer the case.
4017 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4018 and queue runs started by the daemon processed all messages. This has
4019 been fixed so that -R and -S can now usefully be given with -q<time>.
4021 PH/40 Import PCRE release 6.7 (fixes some bugs).
4023 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4025 PH/42 Give an error if -q is specified more than once.
4027 PH/43 Renamed the variables $interface_address and $interface_port as
4028 $received_ip_address and $received_port, to make it clear that these
4029 values apply to message reception, and not to the outgoing interface when
4030 a message is delivered. (The old names remain recognized, of course.)
4032 PH/44 There was no timeout on the connect() call when using a Unix domain
4033 socket in the ${readsocket expansion. There now is.
4035 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4036 be meaningful with "accept".
4038 SC/01 Eximstats V1.43
4039 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4041 SC/02 Eximstats V1.44
4042 Use a glob alias rather than an array ref in the generated
4043 parser. This improves both readability and performance.
4045 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4046 Collect SpamAssassin and rejection statistics.
4047 Don't display local sender or destination tables unless
4048 there is data to show.
4049 Added average volumes into the top table text output.
4051 SC/04 Eximstats V1.46
4052 Collect data on the number of addresses (recipients)
4053 as well as the number of messages.
4055 SC/05 Eximstats V1.47
4056 Added 'Message too big' to the list of mail rejection
4057 reasons (thanks to Marco Gaiarin).
4059 SC/06 Eximstats V1.48
4060 Mainlog lines which have GMT offsets and are too short to
4061 have a flag are now skipped.
4063 SC/07 Eximstats V1.49 (Alain Williams)
4064 Added the -emptyok flag.
4066 SC/08 Eximstats V1.50
4067 Fixes for obtaining the IP address from reject messages.
4069 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4070 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4071 whitespace changes from 4.64-PH/27
4073 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4076 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4077 are found, allow negative numbers in numeric criteria)
4079 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4081 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4082 to match changes made in 4.64-PH/43
4084 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4086 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4087 "domain" is actually an IP address, and force a failure. This locks out
4088 those revolvers/nameservers that support "A-for-A" lookups, in
4089 contravention of the specifications.
4091 PH/48 When a host name was looked up from an IP address, and the subsequent
4092 forward lookup of the name timed out, the host name was left in
4093 $sender_host_name, contrary to the specification.
4095 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4096 restricted to single-key lookups, Exim was not diagnosing an error if
4097 * or *@ was used with a query-style lookup.
4099 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4101 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4102 long ago, but noone interested enough thought of it. Let's just say that
4103 the "1.1" means that there are some new functions that weren't there at
4104 some point in the past.
4106 PH/51 Error processing for expansion failure of helo_data from an smtp
4107 transport during callout processing was broken.
4109 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4110 tested/used via the -bh/-bhc/-bs options.
4112 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4113 bug, fixed in subsequent PCRE releases).
4115 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4116 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4118 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4124 SC/01 Use a glob alias rather than an array ref in eximstats generated
4125 parser. This improves both readability and performance.
4127 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4128 Don't display local sender or destination tables in eximstats unless
4129 there is data to show.
4130 Added average volumes into the eximstats top table text output.
4132 SC/03 Collect data on the number of addresses (recipients) as well
4133 as the number of messages in eximstats.
4135 TF/01 Correct an error in the documentation for the redirect router. Exim
4136 does (usually) call initgroups() when daemonizing.
4138 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4139 with consistent privilege compared to when running as a daemon.
4141 TF/03 Note in the spec that $authenticated_id is not set for local
4142 submissions from trusted users.
4144 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4145 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4147 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4148 by adding some example configuration directives to the default
4149 configuration file. A little bit of work is required to uncomment the
4150 directives and define how usernames and passwords are checked, but
4151 there is now a framework to start from.
4153 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4154 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4155 without this. I don't know how relevant this is to other LDAP libraries.
4157 PH/02 Add the verb name to the "unknown ACL verb" error.
4159 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4161 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4163 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4164 directory not expanded when it should be if an expanded home directory
4165 was set for the address (which is overridden by the transport).
4167 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4170 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4171 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4172 because it is too late at that time, and has no effect.
4174 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4175 security issue with \' (bugzilla #107). I could not use the
4176 PQescapeStringConn() function, because it needs a PGconn value as one of
4179 PH/08 When testing addresses using -bt, indicate those final addresses that
4180 are duplicates that would not cause an additional delivery. At least one
4181 person was confused, thinking that -bt output corresponded to deliveries.
4182 (Suppressing duplicates isn't a good idea as you lose the information
4183 about possibly different redirections that led to the duplicates.)
4185 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4186 systems where poll() doesn't work, in particular OS X.
4188 PH/10 Added more information to debugging output for retry time not reached.
4190 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4191 operations in malware.c.
4193 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4196 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4197 syslog_duplication set false, log lines that would normally be written
4198 both the the main log and to the reject log were not written to syslog at
4201 PH/14 In the default configuration, change the use of "message" in ACL warn
4202 statements to "add_header".
4204 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4205 not followed by a command (e.g. "seen endif").
4207 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4208 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4211 PH/17 Added extra conditions to the default value of delay_warning_condition
4215 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4216 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4217 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4220 The Auto-Submitted: and various List- headers are standardised, whereas I
4221 don't think Precedence: ever was.
4223 PH/18 Refactored debugging code in route_finduser() to show more information,
4224 in particular, the error code if getpwnam() issues one.
4226 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4227 This is apparently needed in addition to the PH/07 change above to avoid
4228 any possible encoding problems.
4230 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4231 but not after initializing Perl.
4233 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4234 output them only if debugging. By default they are written stderr,
4235 apparently, which is not desirable.
4237 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4240 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4243 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4245 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4246 authenticated or an ident call has been made. Suppress the default
4247 values for $authenticated_id and $authenticated_sender (but permit -oMai
4248 and -oMas) when testing with -bh.
4250 PH/24 Re-jigged the order of the tests in the default configuration so that the
4251 tests for valid domains and recipients precede the DNS black list and CSA
4252 tests, on the grounds that those ones are more expensive.
4254 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4255 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4256 This bug exists in every version of Exim that I still have, right back to
4259 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4260 However, an attempt to turn on case-sensitivity in a regex key by
4261 including (?-i) didn't work because the subject string was already
4262 lowercased, and the effects were non-intuitive. It turns out that a
4263 one-line patch can be used to allow (?-i) to work as expected.
4269 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4270 other effects) broke the use of negated acl sub-conditions.
4272 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4275 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4276 "Deny" causes Exim to reject the incoming connection with a 554 error.
4277 Unfortunately, if there is a major crisis, such as a disk failure,
4278 tcp-wrappers gives "deny", whereas what one would like would be some
4279 kind of temporary error. A kludge has been added to help with this.
4280 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4281 554 error is used if errno is still zero or contains ENOENT (which occurs
4282 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4285 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4287 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4288 errors. Otherwise a message that provokes a temporary error (when other
4289 messages do not) can cause a whole host to time out.
4291 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4292 addresses were routed directly to files or pipes from a redirect router.
4293 File deliveries just didn't batch; pipe deliveries might have suffered
4296 PH/06 A failure to get a lock for a hints database would erroneously always say
4297 "Failed to get write lock", even when it was really a read lock.
4299 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4300 of 0600. This has been changed to use the value of the lockfile_mode
4301 option (which defaults to 0600).
4303 PH/08 Applied small patch from the Sieve maintainer.
4305 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4306 folder from quota calculations, a direct delivery into this folder messed
4307 up the contents of the maildirsize file. This was because the regex was
4308 used only to exclude .Trash (or whatever) when the size of the mailbox
4309 was calculated. There was no check that a delivery was happening into an
4310 excluded directory. This bug has been fixed by ignoring all quota
4311 processing for deliveries into excluded directories.
4313 PH/10 Added the maildirfolder_create_regex option to appendfile.
4319 PH/01 The code for finding all the local interface addresses on a FreeBSD
4320 system running IPv6 was broken. This may well have applied to all BSD
4321 systems, as well as to others that have similar system calls. The broken
4322 code found IPv4 interfaces correctly, but gave incorrect values for the
4323 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4324 that it would not match correctly against @[] and not recognize the IPv6
4327 PH/02 The ipliteral router was not recognizing addresses of the form user@
4328 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4330 PH/03 Added disable_ipv6.
4332 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4333 decoded form, because it is most often used to construct To: headers
4334 lines in autoreplies, and the decoded form may well be syntactically
4335 invalid. However, $reply_address has leading white space removed, and all
4336 newlines turned into spaces so that the autoreply transport does not
4339 PH/05 If group was specified without a user on a router, and no group or user
4340 was specified on a transport, the group from the router was ignored.
4342 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4343 for visible compile-time settings that can be used to change these
4344 numbers, for those that want even more. Backwards compatibility with old
4345 spool files has been maintained. However, going back to a previous Exim
4346 release will lost any variables that are in spool files.
4348 PH/07 Two small changes when running in the test harness: increase delay when
4349 passing a TCP/IP connection to a new process, in case the original
4350 process has to generate a bounce, and remove special handling of
4351 127.0.0.2 (sic), which is no longer necessary.
4353 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4354 be the same on different OS.
4356 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4359 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4360 whether --show-vars was specified or not
4362 JJ/02 exipick: Added support for new ACL variable spool format introduced
4365 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4366 syntactically invalid From: or Reply-to: line, and a filter used this to
4367 generate an autoreply, and therefore failed to obtain an address for the
4368 autoreply, Exim could try to deliver to a non-existent relative file
4369 name, causing unrelated and misleading errors. What now happens is that
4370 it logs this as a hard delivery error, but does not attempt to create a
4373 PH/11 The exinext utility has a -C option for testing purposes, but although
4374 the given file was scanned by exinext itself; it wasn't being passed on
4375 when Exim was called.
4377 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4378 an end-of-file indication when reading a command response.
4380 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4381 compiled. In many other places in Exim, IPv6 addresses are always
4382 recognized, so I have changed this. It also means that IPv4 domain
4383 literals of the form [IPV4:n.n.n.n] are now always recognized.
4385 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4386 used if the router is not running as root, for example, when verifying at
4387 ACL time, or when using -bh. The debugging output from this situation was
4388 non-existent - all you got was a failure to exec. I have made two
4391 (a) Failures to set uid/gid, the current directory, or a process leader
4392 in a subprocess such as that created by queryprogram now generate
4393 suitable debugging output when -d is set.
4395 (b) The queryprogram router detects when it is not running as root,
4396 outputs suitable debugging information if -d is set, and then runs
4397 the subprocess without attempting to change uid/gid.
4399 PH/15 Minor change to Makefile for building test_host (undocumented testing
4402 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4403 additional section of a DNS packet that returns MX or SRV records.
4404 Instead, it always explicitly searches for A/AAAA records. This avoids
4405 major problems that occur when a DNS server includes only records of one
4406 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4407 fixed another bug: if SRV records were looked up and the corresponding
4408 address records were *not* found in the additional section, the port
4409 values from the SRV records were lost.
4411 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4412 using the correct key (the original address) when searching the retry
4413 rules in order to find which one to use for generating the retry hint.
4415 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4416 adding the default one. Similarly, if it contains a Reply-To: header, the
4417 errors_reply_to option, if set, is not used.
4419 PH/19 When calculating a retry time, Exim used to measure the "time since
4420 failure" by looking at the "first failed" field in the retry record. Now
4421 it does not use this if it is later than than the arrival time of the
4422 message. Instead it uses the arrival time. This makes for better
4423 behaviour in cases where some deliveries succeed, thus re-setting the
4424 "first failed" field. An example is a quota failure for a huge message
4425 when small messages continue to be delivered. Without this change, the
4426 "time since failure" will always be short, possible causing more frequent
4427 delivery attempts for the huge message than are intended.
4428 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4430 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4431 $1, $2, $3) because the numerical variables can be reset during some
4432 expansion items (e.g. "match"), thereby losing the authentication data.
4434 PH/21 Make -bV show the size of off_t variables so that the test suite can
4435 decide whether to run tests for quotas > 2G.
4437 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4438 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4439 filecount value is greater than 2G or if a quota value is greater than 2G
4440 on a system where the size of off_t is not greater than 4, a panic error
4443 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4444 never match. The debug and -bh output now contains an explicit error
4445 message indicating a malformed IPv4 address or mask.
4447 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4448 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4449 PH/23 above applies.
4451 PH/25 Do not write to syslog when running in the test harness. The only
4452 occasion when this arises is a failure to open the main or panic logs
4453 (for which there is an explicit test).
4455 PH/26 Added the /no_tell option to "control=freeze".
4457 PH/27 If a host name lookup failed very early in a connection, for example, if
4458 the IP address matched host_lookup and the reverse lookup yielded a name
4459 that did not have a forward lookup, an error message of the form "no IP
4460 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4461 could be logged. Now it outputs the IP address instead of "NULL".
4463 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4464 allows the sender and the authenticated sender to be set when
4465 submitting a message from within Exim. Since child_open_exim() is
4466 documented for local_scan(), the new function should be too.
4468 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4469 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4470 results in an empty string is now treated as unset.
4472 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4474 PH/31 Added sender_verify_fail logging option.
4476 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4477 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4478 tidied the source and removed it altogether.
4480 PH/33 When a queue run was abandoned because the load average was too high, a
4481 log line was always written; now it is written only if the queue_run log
4482 selector is set. In addition, the log line for abandonment now contains
4483 information about the queue run such as the pid. This is always present
4484 in "start" and "stop" lines but was omitted from the "abandon" line.
4486 PH/34 Omit spaces between a header name and the colon in the error message that
4487 is given when verify = headers_syntax fails (if there are lots of them,
4488 the message gets confusing).
4490 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4491 names, as there are now some PTR records that contain slashes. This check
4492 is only to protect against broken name servers that fall over on strange
4493 characters, so the fact that it applies to all lookups doesn't matter.
4495 PH/36 Now that the new test suite is complete, we can remove some of the
4496 special code in Exim that was needed for the old test suite. For example,
4497 sorting DNS records because real resolvers return them in an arbitrary
4498 order. The new test suite's fake resolver always returns records in the
4501 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4502 bounces) except when queue_only is set, to avoid logging races between
4503 the different processes.
4505 PH/38 Panic-die if .include specifies a non-absolute path.
4507 PH/39 A tweak to the "H" retry rule from its user.
4509 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4510 a label. They prevented compilation on older perls.
4512 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4513 a warning to be raised on newish perls.
4515 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4516 on queue. Changes to match documented behaviour of showing count of
4517 messages matching specified criteria.
4519 PH/40 Changed the default ident timeout from 30s to 5s.
4521 PH/41 Added support for the use of login_cap features, on those BSD systems
4522 that have them, for controlling the resources used by pipe deliveries.
4524 PH/42 The content-scanning code uses fopen() to create files in which to put
4525 message data. Previously it was not paying any attention to the mode of
4526 the files. Exim runs with umask(0) because the rest of the code creates
4527 files with open(), and sets the required mode explicitly. Thus, these
4528 files were ending up world-writeable. This was not a big issue, because,
4529 being within the spool directory, they were not world-accessible. I have
4530 created a function called modefopen, which takes an additional mode
4531 argument. It sets umask(777), creates the file, chmods it to the required
4532 mode, then resets the umask. All the relevant calls to fopen() in the
4533 content scanning code have been changed to use this function.
4535 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4536 to 24 hours. This avoids potential overflow problems when processing G
4537 and H retry rules. I suspect nobody ever tinkers with this value.
4539 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4541 PH/45 When the plaintext authenticator is running as a client, the server's
4542 challenges are checked to ensure they are valid base64 strings. By
4543 default, the authentication attempt is cancelled if an invalid string is
4544 received. Setting client_ignore_invalid_base64 true ignores these errors.
4545 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4546 they are received. Thus, the responses can be made to depend on the
4547 challenges. If an invalid string is ignored, an empty string is placed in
4550 PH/46 Messages that are created by the autoreply transport now contains a
4551 References: header, in accordance with RFCs 2822 and 3834.
4553 PH/47 Added authenticated_sender_force to the smtp transport.
4555 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4557 PH/49 Installed latest patch from the Sieve maintainer.
4559 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4560 was also set, the appendfile transport was unnecessarily scanning a
4561 directory of message files (e.g. for maildir delivery) to find the count
4562 of files (along with the size), even though it did not need this
4563 information. It now does the scan only if it needs to find either the
4564 size of the count of files.
4566 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4568 PH/52 Two bugs concerned with error handling when the smtp transport is
4571 (i) Exim was not creating retry information for temporary errors given
4572 for individual recipients after the DATA command when the smtp transport
4573 was used in LMTP mode. This meant that they could be retried too
4574 frequently, and not timed out correctly.
4576 (ii) Exim was setting the flag that allows error details to be returned
4577 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4578 recipients that were returned after the DATA command.
4580 PH/53 This is related to PH/52, but is more general: for any failing address,
4581 when detailed error information was permitted to be returned to the
4582 sender, but the error was temporary, then after the final timeout, only
4583 "retry timeout exceeded" was returned. Now it returns the full error as
4584 well as "retry timeout exceeded".
4586 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4587 do this, and (what is worse) MTAs that accept it.
4589 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4590 will now be deprecated.
4592 PH/56 New os.c-cygwin from the Cygwin maintainer.
4594 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4595 formats (previously only available in exim formats via -bpr, -bpru,
4596 and -bpra. Now also available in native and exiqgrep formats)
4598 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4599 with very large, slow to parse queues
4601 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4603 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4605 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4606 responses to authentication challenges, though it was showing the
4607 challenges; (ii) I've removed the CR characters from the debug output for
4610 PH/58 Allow for the insertion of a newline as well as a space when a string
4611 is turned into more than one encoded-word during RFC 2047 encoding. The
4612 Sieve code now uses this.
4614 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4615 data_4xx, lost_connection, tls_required.
4617 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4618 message was being sent as an SMTP response.
4620 PH/61 Add -l and -k options to exicyclog.
4622 PH/62 When verifying, if an address was redirected to one new address, so that
4623 verification continued, and the new address failed or deferred after
4624 having set something in $address_data, the value of $address_data was not
4625 passed back to the ACL. This was different to the case when no
4626 redirection occurred. The value is now passed back in both cases.
4628 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4629 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4630 use login_cap.h, so on its own it isn't the distinguishing feature. The
4631 new name refers directly to the setclassresources() function.
4633 PH/65 Added configuration files for NetBSD3.
4635 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4637 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4638 is preferred over IPv4.
4640 PH/68 The bounce_return_message and bounce_return_body options were not being
4641 honoured for bounces generated during the reception of non-SMTP messages.
4642 In particular, this applied to messages rejected by the ACL. This bug has
4643 been fixed. However, if bounce_return_message is true and bounce_return_
4644 body is false, the headers that are returned for a non-SMTP message
4645 include only those that have been read before the error was detected.
4646 (In the case of an ACL rejection, they have all been read.)
4648 PH/69 The HTML version of the specification is now built in a directory called
4649 spec_html instead of spec.html, because the latter looks like a path with
4650 a MIME-type, and this confuses some software.
4652 PH/70 Catch two compiler warnings in sieve.c.
4654 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4655 function verify_get_ident() calls ip_connect() to connect a socket, but
4656 if the "connect()" function timed out, ip_connect() used to close the
4657 socket. However, verify_get_ident() also closes the socket later, and in
4658 between Exim writes to the log, which may get opened at this point. When
4659 the socket was closed in ip_connect(), the log could get the same file
4660 descriptor number as the socket. This naturally causes chaos. The fix is
4661 not to close the socket in ip_connect(); the socket should be closed by
4662 the function that creates it. There was only one place in the code where
4663 this was missing, in the iplookup router, which I don't think anybody now
4664 uses, but I've fixed it anyway.
4666 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4667 well as to direct DNS lookups. Otherwise the handling of names in host
4668 lists is inconsistent and therefore confusing.
4674 PH/01 Two changes to the default runtime configuration:
4676 (1) Move the checks for relay_from_hosts and authenticated clients from
4677 after to before the (commented out) DNS black list checks.
4679 (2) Add control=submission to the relay_from_hosts and authenticated
4680 clients checks, on the grounds that messages accepted by these
4681 statements are most likely to be submissions.
4683 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4685 (1) Generate an error if the third argument for the ${prvs expansion is
4688 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4691 (3) Reset the variables that are obtained from the first argument of
4692 ${prvscheck and used in the second argument before leaving the code,
4693 because their memory is reclaimed, so using them afterwards may do
4696 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4697 one (it's much easier than Tom thought :-).
4699 (5) Because of (4), we can now allow for the use of $prvscheck_result
4700 inside the third argument.
4702 PH/03 For some reason, the default setting of PATH when running a command from
4703 a pipe transport was just "/usr/bin". I have changed it to
4706 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4707 anything to be listed in the output from -bV.
4709 PH/05 When a filter generated an autoreply, the entire To: header line was
4710 quoted in the delivery log line, like this:
4712 => >A.N.Other <ano@some.domain> <original@ddress> ...
4714 This has been changed so that it extracts the operative address. There
4715 may be more than one such address. If so, they are comma-separated, like
4718 => >ano@some.domain,ona@other.domain <original@ddress> ...
4720 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4721 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4722 not being looked up in the rDNS to get a host name, Exim was showing the
4723 IP address twice in Received: lines, even though the IP addresses were
4724 identical. For example:
4726 Received: from [1.2.3.4] (helo=[1.2.3.4])
4728 However, if the real host name was known, it was omitting the HELO data
4729 if it matched the actual IP address. This has been tidied up so that it
4730 doesn't show the same IP address twice.
4732 PH/07 When both +timestamp and +memory debugging was on, the value given by
4733 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4734 called by the debug printing, thereby overwriting the timestamp buffer.
4735 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4737 PH/08 When the original message was included in an autoreply transport, it
4738 always said "this is a copy of the message, including all the headers",
4739 even if body_only or headers_only was set. It now gives an appropriate
4742 PH/09 Applied a patch from the Sieve maintainer which:
4744 o fixes some comments
4745 o adds the (disabled) notify extension core
4746 o adds some debug output for the result of if/elsif tests
4747 o points to the current vacation draft in the documentation
4748 and documents the missing references header update
4752 o fixes a bug in processing the envelope test (when testing
4753 multiple envelope elements, the last element determined the
4756 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4757 Electronic Mail") by including:
4759 Auto-submitted: auto-generated
4761 in the messages that it generates (bounce messages and others, such as
4762 warnings). In the case of bounce messages for non-SMTP messages, there was
4763 also a typo: it was using "Auto_submitted" (underscore instead of
4764 hyphen). Since every message generated by Exim is necessarily in response
4765 to another message, thes have all been changed to:
4767 Auto-Submitted: auto-replied
4769 in accordance with these statements in the RFC:
4771 The auto-replied keyword:
4773 - SHOULD be used on messages sent in direct response to another
4774 message by an automatic process,
4776 - MUST NOT be used on manually-generated messages,
4778 - MAY be used on Delivery Status Notifications (DSNs) and Message
4779 Disposition Notifications (MDNs),
4781 - MUST NOT be used on messages generated by automatic or periodic
4782 processes, except for messages which are automatic responses to
4785 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4786 to the default Received: header definition.
4788 PH/12 Added log selector acl_warn_skipped (default on).
4790 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4791 variables because (a) they are in the wrong storage pool and (b) even if
4792 they were copied, it wouldn't work properly because of the caching.
4794 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4795 checking when decoding. Apparently there are clients that generate
4796 overlong encoded strings. Why am I not surprised?
4798 PH/15 If the first argument of "${if match_address" was not empty, but did not
4799 contain an "@" character, Exim crashed. Now it writes a panic log message
4800 and treats the condition as false.
4802 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4804 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4805 extension "envelope-auth". The code is finished and in agreement with
4806 other implementations, but there is no documentation so far and in fact,
4807 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4808 not changing the active code.
4810 Print executed "if" and "elsif" statements when debugging is used. This
4811 helps a great deal to understand what a filter does.
4813 Document more things not specified clearly in RFC3028. I had all this
4814 sorted out, when out of a sudden new issues came to my mind. Oops."
4816 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4819 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4820 canonical form (as documented). However, after a host name lookup from
4821 the IP address, check_host() was doing a simple string comparison with
4822 addresses acquired from the DNS when checking that the found name did
4823 have the original IP as one of its addresses. Since any found IPv6
4824 addresses are likely to be in abbreviated form, the comparison could
4825 fail. Luckily, there already exists a function for doing the comparison
4826 by converting both addresses to binary, so now that is used instead of
4827 the text comparison.
4829 PH/20 There was another similar case to PH/19, when a complete host name was
4830 given in a host list; looking up its IP address could give an abbreviated
4831 form, whereas the current host's name might or might not be abbreviated.
4832 The same fix has been applied.
4838 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4839 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4842 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4843 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4845 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4847 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4848 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4849 Also, for "Auto-Submitted", treat anything other than "no" as
4850 non-personal, in accordance with RFC 3834. (Previously it treated
4851 anything starting "auto-" as non-personal.)
4853 TF/01 The control=submission/name=... option had a problem with syntax
4854 errors if the name included a slash character. The /name= option
4855 now slurps the rest of the string, so it can include any characters
4856 but it must come last in the list of options (after /sender_retain
4859 PH/05 Some modifications to the interface to the fake nameserver for the new
4867 TK/01 Added the "success_on_redirect" address verification option. See
4868 NewStuff for rationale and an example.
4870 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4872 PH/02 Patch to exigrep to allow it to work on syslog lines.
4874 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4875 fread() to skip over the body file's header line, because in Cygwin the
4876 header line is locked and is inaccessible.
4878 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4879 co-exist for some time) to make it clear that it is the Exim ID that is
4880 referenced, not the Message-ID: header line.
4882 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4883 string_format() function, because snprintf() does not exist on all
4886 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4887 ${stat: expansion item.
4889 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4890 protocol synchronization error", to keep the pedants happy.
4892 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4893 well as for IRIX systems, when gcc is being used. See the host.c source
4896 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4898 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4901 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4902 transport and to the smtp transport in LMTP mode.
4904 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4906 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4907 run a filter in a subprocess. This could lead to confusion in subsequent
4908 lookups in the parent process. There should also be a search_tidyup() at
4909 the end of the subprocess.
4911 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4912 only if the host matched helo_try_verify_hosts, which caused the
4913 verification to occur when the EHLO/HELO command was issued. The ACL just
4914 tested the remembered result. Now, if a previous verification attempt has
4915 not happened, "verify = helo" does it there and then.
4917 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4919 TK/03 Fix log output including CR from clamd.
4921 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4922 exist provoked a memory error which could cause a segfault.
4924 PH/15 Installed PCRE 6.2
4926 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4928 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4929 of the problem. Specifically, suggested +O2 rather than +O1 for the
4932 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4934 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4935 record, the port was not correctly logged when the outgoing_port log
4936 selector was set (it logged the transort's default port).
4938 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4939 fallback_hosts, and "hosts" in the smtp transport.
4941 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4942 host errors such as "Connection refused".
4944 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4945 authentication with radiusclient 0.4.9:
4947 - Error returned from rc_read_config was caught wrongly
4948 - Username/password not passed on to radius server due to wrong length.
4950 The presumption is that some radiusclient API changes for 4.51/PH/17
4951 were not taken care of correctly. The code is still untested by me (my
4952 Linux distribution still has 0.3.2 of radiusclient), but it was
4953 contributed by a Radius user.
4955 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4956 expanding the "port" option of the smtp transport.
4958 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4959 while reading a MIME header. Thanks to Tom Hughes for a patch.
4961 PH/24 Include config.h inside local_scan.h so that configuration settings are
4964 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4965 that in an ACL for RCPT (for example), you can examine exactly what was
4968 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4969 commands, but it was not correctly comparing the address with the actual
4970 client host address. Thus, it would show the EHLO address in Received:
4971 header lines when this was not necessary.
4973 PH/27 Added the % operator to ${eval:}.
4975 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4976 it should be ignoring failures (because with -C, for example, it has lost
4977 privilege). It wasn't ignoring creation failures other than "already
4980 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4983 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4984 because an input line was too long, either on its own, or by virtue of
4985 too many continuations, the temporary file was not being removed, and the
4986 return code was incorrect.
4988 PH/31 Missing "BOOL" in function definition in filtertest.c.
4990 PH/32 Applied Sieve patches from the maintainer.
4992 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4994 PH/33 Added "verify = not_blind".
4996 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4997 Local/Makefile (with some defaults set). These are used in built scripts
4998 such as exicyclog, but they have never been used in the exim_install
4999 script (though there are many overriding facilities there). I have
5000 arranged that the exim_install script now takes note of these two
5003 PH/35 Installed configuration files for Dragonfly.
5005 PH/36 When a locally submitted message by a trusted user did not contain a
5006 From: header, and the sender address was obtained from -f or from an SMTP
5007 MAIL command, and the trusted user did not use -F to supply a sender
5008 name, $originator_name was incorrectly used when constructing a From:
5009 header. Furthermore, $originator_name was used for submission mode
5010 messages from external hosts without From: headers in a similar way,
5011 which is clearly wrong.
5013 PH/37 Added control=suppress_local_fixups.
5015 PH/38 When log_selector = +received_sender was set, and the addition of the
5016 sender made the log line's construction buffer exactly full, or one byte
5017 less than full, an overflow happened when the terminating "\n" was
5020 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5021 when the result of a list match is failure because a DNS lookup failed.
5023 PH/40 RM_COMMAND is now used in the building process.
5025 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5026 the "build-* directories that it finds.
5028 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5029 address in a domain literal was a prefix of an interface address.
5031 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5032 when verifying a sender address, unless rewrite_headers is false.
5034 PH/44 Wrote a long comment about why errors_to addresses are verified as
5035 recipients, not senders.
5037 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5038 the ratelimit ACL was added.
5040 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5042 PH/46 Added extra information about PostgreSQL errors to the error string.
5044 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5045 avoiding the need to install special zones in a real server. This is
5046 backwards compatible; if it can't find the fake resolver, it drops back.
5047 Thus, both old and new test suites can be run.
5049 TF/02 Added util/ratelimit.pl
5051 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5052 clock is set back in time.
5054 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5055 Candler <B.Candler@pobox.com>.
5057 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5058 OK for addresses that are the result of redirections.
5060 PH/48 A number of further additions for the benefit of the new test suite,
5061 including a fake gethostbyname() that interfaces to the fake DNS resolver
5064 TF/06 The fix for widen_domains has also been applied to qualify_single and
5065 search_parents which are the other dnslookup options that can cause
5068 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5071 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5073 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5074 service" if it encountered an MX record with an empty target hostname.
5075 The message is now "an MX or SRV record indicated no SMTP service".
5077 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5078 if the DNS of the sending site is misconfigured. This is quite a
5079 common situation. This change restores the behaviour of treating a
5080 helo verification defer as a failure.
5082 PH/51 If self=fail was set on a router, the bounce message did not include the
5083 actual error message.
5089 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5091 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5092 command itself ended in error, the underlying message about the transport
5093 filter timeout was being overwritten with the pipe command error. Now the
5094 underlying error message should be appended to the second error message.
5096 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5098 PH/02 Reduce the amount of output that "make" produces by default. Full output
5099 can still be requested.
5101 PH/03 The warning log line about a condition test deferring for a "warn" verb
5102 was being output only once per connection, rather than after each
5103 occurrence (because it was using the same function as for successful
5104 "warn" verbs). This seems wrong, so I have changed it.
5106 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5107 it should not have, which might have caused a crash in the right
5108 circumstances, but probably never did.
5110 PH/04 Installed a modified version of Tony Finch's patch to make submission
5111 mode fix the return path as well as the Sender: header line, and to
5112 add a /name= option so that you can make the user's friendly name appear
5115 TF/03 Added the control = fakedefer ACL modifier.
5117 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5118 Mark Lowes for thorough testing.
5120 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5122 TK/03 Merged latest SRS patch from Miles Wilton.
5124 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5125 with the definition in sysexits.h (which is #included earlier).
5126 Fortunately, Exim does not actually use EX_OK. The code used to try to
5127 preserve the sysexits.h value, by assuming that macro definitions were
5128 scanned for macro replacements. I have been disabused of this notion,
5129 so now the code just undefines EX_OK before #including unistd.h.
5131 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5132 in the smtp transport. When a block could not be written in a single
5133 write() function, the timeout was being re-applied to each part-write.
5134 This seems wrong - if the receiver was accepting one byte at a time it
5135 would take for ever. The timeout is now adjusted when this happens. It
5136 doesn't have to be particularly precise.
5138 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5139 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5141 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5142 without a domain if the check to <postmaster@domain> fails.
5144 SC/01 Eximstats: added -xls and the ability to specify output files
5145 (patch written by Frank Heydlauf).
5147 SC/02 Eximstats: use FileHandles for outputting results.
5149 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5151 SC/04 Eximstats: fixed display of large numbers with -nvr option
5153 SC/05 Eximstats: fixed merging of reports with empty tables.
5155 SC/06 Eximstats: added the -include_original_destination flag
5157 SC/07 Eximstats: removed tabs and trailing whitespace.
5159 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5161 TK/06 MBOX spool code: Add real "From " MBOX separator line
5162 so the .eml file is really in mbox format (even though
5163 most programs do not really care). Patch from Alex Miller.
5165 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5166 The latter is generated from $received_to and is only set if the
5167 message has one envelope recipient. SA can use these headers,
5168 obviously out-of-the-box. Patch from Alex Miller.
5170 PH/08 The ${def test on a variable was returning false if the variable's
5171 value was "0", contrary to what the specification has always said!
5172 The result should be true unless the variable is empty.
5174 PH/09 The syntax error of a character other than { following "${if
5175 def:variable_name" (after optional whitespace) was not being diagnosed.
5176 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5177 accidental colon was present, for example, could give incorrect results.
5179 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5180 result is used (not including appendfile, where other changes are about
5183 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5184 This involved changing a lot of size variables from int to off_t. It
5185 should work with maildirs and everything.
5187 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5188 spamd dying while we are connected to it.
5190 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5193 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5194 function for BDB 4.3.
5196 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5198 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5199 the sources for PCRE have been re-arranged and more files are now
5202 PH/15 The code I had for printing potentially long long variables in PH/11
5203 above was not the best (it lost precision). The length of off_t variables
5204 is now inspected at build time, and an appropriate printing format (%ld
5205 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5206 to be "long long int" or "long int". This is needed for the internal
5207 formatting function string_vformat().
5209 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5210 the configuration file to be ":syslog", then the script "guesses" where
5211 the logs files are, rather than using the compiled in default. In our
5212 case the guess is not the same as the compiled default, so the script
5213 suddenly stopped working when I started to use syslog. The patch checks
5214 to see if log_file_path is "". If so, it attempts to read it from exim
5215 with no configuration file to get the compiled in version, before it
5216 falls back to the previous guessing code."
5218 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5219 implementing BATV in an Exim configuration. See NewStuff for the gory
5222 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5223 Makefile that are specific to HP-UX.
5225 PH/18 If the "use_postmaster" option was set for a recipient callout together
5226 with the "random" option, the postmaster address was used as the MAIL
5227 FROM address for the random test, but not for the subsequent recipient
5228 test. It is now used for both.
5230 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5231 patch removes a few documentation additions to RFC 3028, because the
5232 latest draft now contains them. It adds the new en;ascii-case comparator
5233 and a new error check for 8bit text in MIME parts. Comparator and
5234 require names are now matched exactly. I enabled the subaddress
5235 extension, but it is not well tested yet (read: it works for me)."
5237 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5238 rework some of the code of TK/09 above to avoid the hardwired use of
5239 "%lld" and "long long". Replaced the call to snprintf() with a call to
5242 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5243 records point to non-existent hosts", "retry timeout exceeded", and
5244 "retry time not reached for any host after a long failure period".
5246 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5247 experimental DomainKeys support:
5249 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5250 (2) On an error such as an illegally used "control", the wrong name for
5251 the control was given.
5253 These problems did NOT occur unless DomainKeys support was compiled.
5255 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5257 PH/24 Added ${if match_ip condition.
5259 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5260 databases so that it will be absolutely obvious if a crash occurs in the
5261 DB library. This is a regular occurrence (often caused by mis-matched
5264 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5265 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5266 picked up on a user's system that detects such things. There doesn't seem
5267 to be a gcc warning option for this - only an attribute that has to be
5268 put on the function's prototype. It seems that in Fedora Core 4 they have
5269 set this on a number of new functions. No doubt there will be more in due
5272 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5273 specify a transport. However, if an address that was verified by such a
5274 router was the subject of a callout, Exim crashed because it tried to
5275 read the rcpt_include_affixes from the non-existent transport. Now it
5276 just assumes that the setting of that option is false. This bug was
5277 introduced by 4.51/PH/31.
5279 PH/28 Changed -d+all to exclude +memory, because that information is very
5280 rarely of interest, but it makes the output a lot bigger. People tend to
5281 do -d+all out of habit.
5283 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5284 code in os-type was giving problems when libc.so lives in lib64, like on
5287 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5288 aren't the modern standard, and it seems that some systems' include files
5289 don't always have them. Exim was already checking for some of the newer
5290 ones like T_AAAA, and defining it itself. I've added checks for all the
5291 record types that Exim uses.
5293 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5294 not automatically generating a new one, as it is supposed to. This
5295 prevented TLS from working. If the file did exist, but contained invalid
5296 data, a new version was generated, as expected. It was only the case of a
5297 non-existent file that was broken.
5299 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5300 with a change in libdomainkeys > 0.64.
5302 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5303 from DNS. If the selector record carries the flag, it now has
5304 precedence over the domain-wide flag.
5306 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5308 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5309 the use of a port name that isn't defined in /etc/services) occurred, the
5310 message was deferred as in a normal delivery, and thus remained on the
5311 spool, instead of being failed because of the mua_wrapper setting. This
5312 is now fixed, and I tidied up some of the mua_wrapper messages at the
5315 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5316 the messages in a hash of arrays rather than using individual hashes.
5317 This is a bit cleaner and results in dramatic memory savings, albeit
5318 at a slight CPU cost.
5320 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5321 as requested by Marc Sherman.
5323 SC/10 Eximstats: added histograms for user specified patterns as requested
5326 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5328 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5329 fopen() in the content-scanning modules that did not already have it.
5335 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5336 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5338 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5340 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5342 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5343 to test Sieve filters that use "vacation".
5345 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5346 that changes the way the GnuTLS parameters are stored in the cache file.
5347 The new format can be generated externally. For backward compatibility,
5348 if the data in the cache doesn't make sense, Exim assumes it has read an
5349 old-format file, and it generates new data and writes a new file. This
5350 means that you can't go back to an older release without removing the
5353 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5354 work if there are any delivery delays because "one_time" forces the
5355 parent to be marked "delivered", so its unseen clone is never tried
5356 again. For this reason, Exim now forbids the simultaneous setting of
5359 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5360 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5361 entry if you want to know the details. The fix, however, neglected to
5362 consider the case when local delivery batching is involved. The test for
5363 "previously delivered" was not happening when checking to see if an
5364 address could be batched with a previous (undelivered) one; under
5365 certain circumstances this could lead to multiple deliveries to the same
5368 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5369 in its include files, and this causes problems building Exim.
5371 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5372 header_syntax) but Exim was just ignoring anything given after a slash.
5373 In particular, this caused confusion with an attempt to use "verify =
5374 reverse_host_lookup/defer_ok". An error is now given when options are
5375 supplied for verify items that do not have them. (Maybe reverse_host_
5376 lookup should have a defer_ok option, but that's a different point.)
5378 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5379 defined by RFC 821) to 2048, because there were problems with some AUTH
5380 commands, and RFC 1869 says the size should be increased for extended
5381 SMTP commands that take arguments.
5383 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5386 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5387 "unknown" error; now it says that the functionality isn't in the binary.
5389 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5390 an address' error message when a string expansion fails (syntax or
5391 whatever). Otherwise the password may appear in the log. Following change
5392 PH/42 below, there is no longer a chance of it appearing in a bounce
5395 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5397 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5398 of its addresses was ever tried. (Bugzilla bug #2.)
5400 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5401 the result incorrectly in the debug output. (It correctly added a newline
5402 to what was transported.)
5404 TF/01 Added $received_time.
5406 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5407 commented out examples of how to interface to a virus scanner and to
5408 SpamAssassin. Also added commented examples of av_scanner and
5409 spamd_address settings.
5411 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5412 and controls are allowed in which ACLs. There were a couple of minor
5413 errors. Some of the entries in the conditions table (which is a table of
5414 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5415 negation of where the condition IS allowed.
5417 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5419 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5420 header file does not have a version number, so I've had to invent a new
5421 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5422 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5423 radiusclient), but it was contributed by a Radius user.
5425 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5426 files or directories, for OpenSSL.
5428 PH/19 When an Exim process that is running as root has to create an Exim log
5429 file, it does so in a subprocess that runs as exim:exim so as to get the
5430 ownership right at creation (otherwise, other Exim processes might see
5431 the file with the wrong ownership). There was no test for failure of this
5432 fork() call, which would lead to the process getting stuck as it waited
5433 for a non-existent subprocess. Forks do occasionally fail when resources
5434 run out. I reviewed all the other calls to fork(); they all seem to check
5437 PH/20 When checking for unexpected SMTP input at connect time (before writing
5438 the banner), Exim was not dealing correctly with a non-positive return
5439 from the read() function. If the client had disconnected by this time,
5440 the result was a log entry for a synchronization error with an empty
5441 string after "input=" when read() returned zero. If read() returned -1
5442 (an event I could not check), uninitialized data bytes were printed.
5443 There were reports of junk text (parts of files, etc) appearing after
5446 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5448 PH/22 Added support for macro redefinition, and (re)definition in between
5449 driver and ACL definitions.
5451 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5452 forgetting to use the resulting value; it was using the unexpanded value.
5454 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5455 hadn't been configured. The fix is from Juergen Kreileder, who
5456 understands it better than I do:
5458 "Here's what I see happening with three configured cyrus_sasl
5459 authenticators configured (plain, login, cram-md5):
5461 On startup auth_cyrus_sasl_init() gets called for each of these.
5462 This means three calls to sasl_listmech() without a specified mech_list.
5463 => SASL tests which mechs of all available mechs actually work
5464 => three warnings about OTP not working
5465 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5467 With the patch, sasl_listmech() also gets called three times. But now
5468 SASL's mech_list option is set to the server_mech specified in the the
5469 authenticator. Or in other words, the answer from sasl_listmech()
5470 gets limited to just the mech you're testing for (which is different
5472 => the return list contains just 'plain' or 'login', 'cram-md5' or
5473 nothing depending on the value of ob->server_mech.
5475 I've just tested the patch: Authentication still works fine,
5476 unavailable mechs specified in the exim configuration are still
5477 caught, and the auth.log warnings about OTP are gone."
5479 PH/25 When debugging is enabled, the contents of the command line are added
5480 to the debugging output, even when log_selector=+arguments is not
5483 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5484 answer is "GNU", and only if the return is "GNU/something" is the answer
5487 PH/27 $acl_verify_message is now set immediately after the failure of a
5488 verification in an ACL, and so is available in subsequent modifiers. In
5489 particular, the message can be preserved by coding like this:
5491 warn !verify = sender
5492 set acl_m0 = $acl_verify_message
5494 Previously, $acl_verify_message was set only while expanding "message"
5495 and "log_message" when a very denied access.
5497 PH/28 Modified OS/os.c-Linux with
5499 -#ifndef OS_LOAD_AVERAGE
5500 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5502 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5503 nomenclature these days.)
5505 PH/29 Installed patch from the Sieve maintainer that adds the options
5506 sieve_useraddress and sieve_subaddress to the redirect router.
5508 PH/30 In these circumstances:
5509 . Two addresses routed to the same list of hosts;
5510 . First host does not offer TLS;
5511 . First host accepts first address;
5512 . First host gives temporary error to second address;
5513 . Second host offers TLS and a TLS session is established;
5514 . Second host accepts second address.
5515 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5516 and peerdn, if requested) that were in fact used only for the second
5519 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5520 attention to any local part prefix or suffix that was matched by the
5521 router that accepted the address. It now behaves in the same way as it
5522 does for delivery: the affixes are removed from the local part unless
5523 rcpt_include_affixes is set on the transport.
5525 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5526 timeout during the DATA phase of an incoming message.
5528 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5529 applied a patch sanctioned by the Sieve maintainer.
5531 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5532 the uid or gid is negative. A case of a negative gid caused this to be
5533 noticed. The fix allows for either to be negative.
5535 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5536 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5539 PH/36 The change PH/12 above was broken. Fixed it.
5541 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5542 the grounds that routing the same address twice would always produce the
5543 same answer. This might have been true once, but it is certainly no
5544 longer true now. Routing a child address may depend on the previous
5545 routing that produced that child. Some complicated redirection strategies
5546 went wrong when messages had multiple recipients, and made Exim's
5547 behaviour dependent on the order in which the addresses were given.
5549 I have moved the duplicate checking until after the routing is complete.
5550 Exim scans the addresses that are assigned to local and remote
5551 transports, and removes any duplicates. This means that more work will be
5552 done, as duplicates will always all be routed, but duplicates are
5553 presumably rare, so I don't expect this is of any significance.
5555 For deliveries to pipes, files, and autoreplies, the duplicate checking
5556 still happens during the routing process, since they are not going to be
5559 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5560 It corrects a timeout issue with spamd. This is Ian's comment: "The
5561 background is that sometimes spamd either never reads data from a
5562 connection it has accepted, or it never writes response data. The exiscan
5563 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5564 blindly assumes that writes won't block so it may never time out."
5566 PH/39 Allow G after quota size as well as K and M.
5568 PH/40 The value set for $authenticated_id in an authenticator may not contain
5569 binary zeroes or newlines because the value is written to log lines and
5570 to spool files. There was no check on this. Now the value is run through
5571 the string_printing() function so that such characters are converted to
5572 printable escape sequences.
5574 PH/41 $message_linecount is a new variable that contains the total number of
5575 lines in the message. Compare $body_linecount, which is the count for the
5578 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5579 bounce and delay warning messages, except in certain special cases, which
5582 (a) An SMTP error message from a remote host;
5583 (b) A message specified in a :fail: redirection;
5584 (c) A message specified in a "fail" command in a system filter;
5585 (d) A message specified in a FAIL return from the queryprogram router;
5586 (e) A message specified by the cannot_route_message router option.
5588 In these cases only, Exim does include the error details in bounce and
5589 warning messages. There are also a few cases where bland messages such
5590 as "unrouteable address" or "local delivery error" are given.
5592 PH/43 $value is now also set for the "else" part of a ${run expansion.
5594 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5595 being worked on, but at least Exim now implements the latest version to
5598 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5599 process to complete was treated as a delivery failure, a timeout while
5600 writing the message to the pipe was logged, but erroneously treated as a
5601 successful delivery. Such timeouts include transport filter timeouts. For
5602 consistency with the overall process timeout, these timeouts are now
5603 treated as errors, giving rise to delivery failures by default. However,
5604 there is now a new Boolean option for the pipe transport called
5605 timeout_defer, which, if set TRUE, converts the failures into defers for
5606 both kinds of timeout. A transport filter timeout is now identified in
5609 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5610 systems where "make" and "gmake" are different, calling "gmake" at top
5611 level broke things. I've arranged for the value of $(MAKE) to be passed
5612 from the Makefile to this script so that it can call the same version of
5616 A note about Exim versions 4.44 and 4.50
5617 ----------------------------------------
5619 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5620 changes of various kinds. As a consequence, a big documentation update was
5621 needed. This delayed the release for rather longer than seemed good, especially
5622 in the light of a couple of (minor) security issues. Therefore, the changes
5623 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5624 release. So 4.44 and 4.50 are in effect two different branches that both start
5627 I have left the 4.50 change log unchanged; it contains all the changes since
5628 4.43. The change log for 4.44 is below; many of its items are identical to
5629 those for 4.50. This seems to be the most sensible way to preserve the
5630 historical information.
5636 1. Minor wording change to the doc/README.SIEVE file.
5638 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5639 computation of the current number of files was incorrect.
5641 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5642 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5645 4. Give more explanation in the error message when the command for a transport
5646 filter fails to execute.
5648 5. There are several places where Exim runs a non-Exim command in a
5649 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5650 was being done only for the command run by the queryprogram router. It is
5651 now done for all such subprocesses. The other cases are: ${run, transport
5652 filters, and the commands run by the lmtp and pipe transports.
5654 6. Added CONFIGURE_GROUP build-time option.
5656 7. Some older OS have a limit of 256 on the maximum number of file
5657 descriptors. Exim was using setrlimit() to set 1000 as a large value
5658 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5659 systems. I've change it so that if it can't get 1000, it tries for 256.
5661 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5662 was an oversight, and furthermore, ever since the addition of extra
5663 controls (e.g. 4.43/32), the checks on when to allow different forms of
5664 "control" were broken. There should now be diagnostics for all cases when a
5665 control that does not make sense is encountered.
5667 9. Added the /retain_sender option to "control=submission".
5669 10. $recipients is now available in the predata ACL (oversight).
5671 11. Tidy the search cache before the fork to do a delivery from a message
5672 received from the command line. Otherwise the child will trigger a lookup
5673 failure and thereby defer the delivery if it tries to use (for example) a
5674 cached ldap connection that the parent has called unbind on.
5676 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5677 of $address_data from the recipient verification was clobbered by the
5678 sender verification.
5680 13. The value of address_data from a sender verification is now available in
5681 $sender_address_data in subsequent conditions in the ACL statement.
5683 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5685 15. Added a new option "connect=<time>" to callout options, to set a different
5688 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5689 was its contents. (It was OK if the option was not defined at all.)
5691 17. A "Completed" log line is now written for messages that are removed from
5692 the spool by the -Mrm option.
5694 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5695 information about exactly what failed.
5697 19. Added -dd to debug only the daemon process.
5699 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5700 handles timeouts, both on the server side and network timeouts. Renamed the
5701 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5703 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5704 It is now set to "smtps".
5706 22. $host_address is now set to the target address during the checking of
5707 ignore_target_hosts.
5709 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5710 being passed; this would have caused $sender_host_name to have been used if
5711 matching the list had actually called for a host name (not very likely,
5712 since this list is usually IP addresses). A host name is now passed as
5715 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5716 code that specifies a non-restarting handler (typically sigaction() in
5717 modern systems) in an attempt to fix a rare and obscure crash bug.
5719 25. Narrowed the window for a race in the daemon that could cause it to ignore
5720 SIGCHLD signals. This is not a major problem, because they are used only to
5721 wake it up if nothing else does.
5723 26. A malformed maildirsize file could cause Exim to calculate negative values
5724 for the mailbox size or file count. Odd effects could occur as a result.
5725 The maildirsize information is now recalculated if the size or filecount
5728 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5729 support for a long time. Removed HAVE_SYS_VFS_H.
5731 28. Installed the latest version of exipick from John Jetmore.
5733 29. In an address list, if the pattern was not a regular expression, an empty
5734 subject address (from a bounce message) matched only if the pattern was an
5735 empty string. Non-empty patterns were not even tested. This was the wrong
5736 because it is perfectly reasonable to use an empty address as part of a
5737 database query. An empty address is now tested by patterns that are
5738 lookups. However, all the other forms of pattern expect the subject to
5739 contain a local part and a domain, and therefore, for them, an empty
5740 address still always fails if the pattern is not itself empty.
5742 30. Exim went into a mad DNS loop when attempting to do a callout where the
5743 host was specified on an smtp transport, and looking it up yielded more
5744 than one IP address.
5746 31. Re-factored the code for checking spool and log partition space into a
5747 function that finds that data and another that does the check. The former
5748 is then used to implement four new variables: $spool_space, $log_space,
5749 $spool_inodes, and $log_inodes.
5751 32. The RFC2047 encoding function was originally intended for short strings
5752 such as real names; it was not keeping to the 75-character limit for
5753 encoded words that the RFC imposes. It now respects the limit, and
5754 generates multiple encoded words if necessary. To be on the safe side, I
5755 have increased the buffer size for the ${rfc2047: expansion operator from
5758 33. It is now permitted to omit both strings after an "if" condition; if the
5759 condition is true, the result is "true". As before, when the second string
5760 is omitted, a false condition yields an empty string. This makes it less
5761 cumbersome to write custom ACL and router conditions.
5763 34. Failure to deliver a bounce message always caused it to be frozen, even if
5764 there was an errors_to setting on the router. The errors_to setting is now
5767 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5768 canonical form (fully expanded) before being placed in
5769 $sender_host_address.
5771 36. The table in the code that translates DNS record types into text (T_A to
5772 "A" for instance) was missing entries for NS and CNAME. It is just possible
5773 that this could have caused confusion if both these types were looked up
5774 for the same domain, because the text type is used as part of Exim's
5775 per-process caching. But the chance of anyone hitting this buglet seems
5778 37. The dnsdb lookup has been extended in a number of ways.
5780 (1) There is a new type, "zns", which walks up the domain tree until it
5781 finds some nameserver records. It should be used with care.
5783 (2) There is a new type, "mxh", which is like "mx" except that it returns
5784 just the host names, not the priorities.
5786 (3) It is now possible to give a list of domains (or IP addresses) to be
5787 looked up. The behaviour when one of the lookups defers can be
5788 controlled by a keyword.
5790 (4) It is now possible to specify the separator character for use when
5791 multiple records are returned.
5793 38. The dnslists ACL condition has been extended: it is now possible to supply
5794 a list of IP addresses and/or domains to be looked up in a particular DNS
5797 39. Added log_selector=+queue_time_overall.
5799 40. When running the queue in the test harness, wait just a tad after forking a
5800 delivery process, to get repeatability of debugging output.
5802 41. Include certificate and key file names in error message when GnuTLS fails
5803 to set them up, because the GnuTLS error message doesn't include the name
5804 of the failing file when there is a problem reading it.
5806 42. Allow both -bf and -bF in the same test run.
5808 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5810 44. The "Exiscan patch" is now merged into the mainline Exim source.
5812 45. Sometimes the final signoff response after QUIT could fail to get
5813 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5814 < 0 before doing a fflush(). This bug looks as though it goes back to the
5815 introduction of TLS in release 3.20, but "sometimes" must have been rare
5816 because the tests only now provoked it.
5818 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5819 (this can affect the format of dates).
5821 47. exim_tidydb, when checking for the continued existence of a message for
5822 which it has found a message-specific retry record, was not finding
5823 messages that were in split spool directories. Consequently, it was
5824 deleting retry records that should have stayed in existence.
5826 48. Steve fixed some bugs in eximstats.
5828 49. The SPA authentication driver was not abandoning authentication and moving
5829 on to the next authenticator when an expansion was forced to fail,
5830 contradicting the general specification for all authenticators. Instead it
5831 was generating a temporary error. It now behaves as specified.
5833 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5834 (the order specifies the preference for clients). The order is now AES256,
5835 AES128, 3DES, ARCFOUR128.
5837 51. Small patch to Sieve code - explicitly set From: when generating an
5840 52. Exim crashed if a remote delivery caused a very long error message to be
5841 recorded - for instance if somebody sent an entire SpamAssassin report back
5842 as a large number of 550 error lines. This bug was coincidentally fixed by
5843 increasing the size of one of Exim's internal buffers (big_buffer) that
5844 happened as part of the Exiscan merge. However, to be on the safe side, I
5845 have made the code more robust (and fixed the comments that describe what
5848 53. Now that there can be additional text after "Completed" in log lines (if
5849 the queue_time_overall log selector is set), a one-byte patch to exigrep
5850 was needed to allow it to recognize "Completed" as not the last thing in
5853 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5854 patch that reportedly fixes this has been added. I am not expert enough to
5855 create a test for it. This is what the patch creator wrote:
5857 "I found a little strange behaviour of ldap code when working with
5858 Windows 2003 AD Domain, where users was placed in more than one
5859 Organization Units. When I tried to give exim partial DN, the exit code
5860 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5861 But simultaneously result of request was absolutely normal ldap result,
5862 so I produce this patch..."
5864 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5865 REFERENCE, so I have modified the code to exclude the patch when that macro
5868 55. Some experimental protocols are using DNS PTR records for new purposes. The
5869 keys for these records are domain names, not reversed IP addresses. The
5870 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5871 leaves it alone. Component reversal etc. now happens only for IP addresses.
5874 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5876 57. Double the size of the debug message buffer (to 2048) so that more of very
5877 long debug lines gets shown.
5879 58. The exicyclog utility now does better if the number of log files to keep
5880 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5882 59. Two changes related to the smtp_active_hostname option:
5884 (1) $smtp_active_hostname is now available as a variable.
5885 (2) The default for smtp_banner uses $smtp_active_hostname instead
5886 of $primary_hostname.
5888 60. The host_aton() function is supposed to be passed a string that is known
5889 to be a valid IP address. However, in the case of IPv6 addresses, it was
5890 not checking this. This is a hostage to fortune. Exim now panics and dies
5891 if the condition is not met. A case was found where this could be provoked
5892 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5893 components; fortuitously, this particular loophole had already been fixed
5894 by change 4.50/55 above.
5896 If there are any other similar loopholes, the new check in host_aton()
5897 itself should stop them being exploited. The report I received stated that
5898 data on the command line could provoke the exploit when Exim was running as
5899 exim, but did not say which command line option was involved. All I could
5900 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5901 running as the user.
5904 61. There was a buffer overflow vulnerability in the SPA authentication code
5905 (which came originally from the Samba project). I have added a test to the
5906 spa_base64_to_bits() function which I hope fixes it.
5909 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5910 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5912 63. The daemon start-up calls getloadavg() while still root for those OS that
5913 need the first call to be done as root, but it missed one case: when
5914 deliver_queue_load_max is set with deliver_drop_privilege. This is
5915 necessary for the benefit of the queue runner, because there is no re-exec
5916 when deliver_drop_privilege is set.
5918 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5919 This has been fixed.
5921 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5922 was in use, was not putting the data itself into the right store pool;
5923 consequently, it could be overwritten for a subsequent message in the same
5924 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5927 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5929 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5930 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5931 the function were treating the return as a boolean value, which happened to
5932 work because 0=false and not-0=true, but is not correct code.
5934 68. The host_aton() function was not handling scoped IPv6 addresses (those
5935 with, for example, "%eth0" on the end) correctly.
5937 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5938 negated items (that is, ~something) in unsigned ints. Some compilers
5939 apparently mutter when there is no cast.
5941 70. If an address verification called from an ACL failed, and did not produce a
5942 user-specific message (i.e. there was only a "system" message), nothing was
5943 put in $acl_verify_message. In this situation, it now puts the system
5946 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5947 session; change 4.31/43 added the unwanted input to the log line - except
5948 that it did not do this in the start of session case. It now does.
5950 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5951 This is wrong and can cause the other end to generate a synchronization
5952 error if it is another Exim or anything else that does the synchronization
5953 check. A QUIT command is no longer sent after a timeout.
5955 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5956 during host lookups.
5958 74. The defer_ok option of callout verification was not working if it was used
5959 when verifying addresses in header lines, that is, for this case:
5961 verify = header_sender/callout=defer_ok
5963 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5964 those file descriptors could be used for SMTP connections. If anything
5965 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5966 could be sent to the SMTP client, causing chaos. The daemon now opens
5967 stdin, stdout, and stderr to /dev/null when it puts itself into the
5970 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5971 log by default. The user can override this with suitable Perl magic.
5973 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5974 the log message when discard triggers, was not working for the DATA ACL or
5975 for the non-SMTP ACL.
5977 78. Error message wording change in sieve.c.
5979 79. If smtp_accept_max_per_host was set, the number of connections could be
5980 restricted to fewer than expected, because the daemon was trying to set up
5981 a new connection before checking whether the processes handling previous
5982 connections had finished. The check for completed processes is now done
5983 earlier. On busy systems, this bug wouldn't be noticed because something
5984 else would have woken the daemon, and it would have reaped the completed
5987 80. If a message was submitted locally by a user whose login name contained one
5988 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5989 It caused a spool format error. I have fixed the spool reading code. A
5990 related problem was that the "from" clause in the Received: line became
5991 illegal because of the space(s). It is now covered by ${quote_local_part.
5993 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5996 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5998 83. Patch from Sieve maintainer to fix unterminated string problem in
5999 "vacation" handling.
6001 84. Some minor changes to the Linux configuration files to help with other
6002 OS variants using glibc.
6004 85. One more patch for Sieve to update vacation handling to latest spec.
6007 ----------------------------------------------------
6008 See the note above about the 4.44 and 4.50 releases.
6009 ----------------------------------------------------
6015 1. Change 4.43/35 introduced a bug that caused file counts to be
6016 incorrectly computed when quota_filecount was set in an appendfile
6019 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6020 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6023 3. Give more explanation in the error message when the command for a transport
6024 filter fails to execute.
6026 4. There are several places where Exim runs a non-Exim command in a
6027 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6028 was being done only for the command run by the queryprogram router. It is
6029 now done for all such subprocesses. The other cases are: ${run, transport
6030 filters, and the commands run by the lmtp and pipe transports.
6032 5. Some older OS have a limit of 256 on the maximum number of file
6033 descriptors. Exim was using setrlimit() to set 1000 as a large value
6034 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6035 systems. I've change it so that if it can't get 1000, it tries for 256.
6037 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6038 was an oversight, and furthermore, ever since the addition of extra
6039 controls (e.g. 4.43/32), the checks on when to allow different forms of
6040 "control" were broken. There should now be diagnostics for all cases when a
6041 control that does not make sense is encountered.
6043 7. $recipients is now available in the predata ACL (oversight).
6045 8. Tidy the search cache before the fork to do a delivery from a message
6046 received from the command line. Otherwise the child will trigger a lookup
6047 failure and thereby defer the delivery if it tries to use (for example) a
6048 cached ldap connection that the parent has called unbind on.
6050 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6051 of $address_data from the recipient verification was clobbered by the
6052 sender verification.
6054 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6055 was its contents. (It was OK if the option was not defined at all.)
6057 11. A "Completed" log line is now written for messages that are removed from
6058 the spool by the -Mrm option.
6060 12. $host_address is now set to the target address during the checking of
6061 ignore_target_hosts.
6063 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6064 being passed; this would have caused $sender_host_name to have been used if
6065 matching the list had actually called for a host name (not very likely,
6066 since this list is usually IP addresses). A host name is now passed as
6069 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6070 code that specifies a non-restarting handler (typically sigaction() in
6071 modern systems) in an attempt to fix a rare and obscure crash bug.
6073 15. Narrowed the window for a race in the daemon that could cause it to ignore
6074 SIGCHLD signals. This is not a major problem, because they are used only to
6075 wake it up if nothing else does.
6077 16. A malformed maildirsize file could cause Exim to calculate negative values
6078 for the mailbox size or file count. Odd effects could occur as a result.
6079 The maildirsize information is now recalculated if the size or filecount
6082 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6083 support for a long time. Removed HAVE_SYS_VFS_H.
6085 18. Updated exipick to current release from John Jetmore.
6087 19. Allow an empty sender to be matched against a lookup in an address list.
6088 Previously the only cases considered were a regular expression, or an
6091 20. Exim went into a mad DNS lookup loop when doing a callout where the
6092 host was specified on the transport, if the DNS lookup yielded more than
6095 21. The RFC2047 encoding function was originally intended for short strings
6096 such as real names; it was not keeping to the 75-character limit for
6097 encoded words that the RFC imposes. It now respects the limit, and
6098 generates multiple encoded words if necessary. To be on the safe side, I
6099 have increased the buffer size for the ${rfc2047: expansion operator from
6102 22. Failure to deliver a bounce message always caused it to be frozen, even if
6103 there was an errors_to setting on the router. The errors_to setting is now
6106 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6107 canonical form (fully expanded) before being placed in
6108 $sender_host_address.
6110 24. Updated eximstats to version 1.33
6112 25. Include certificate and key file names in error message when GnuTLS fails
6113 to set them up, because the GnuTLS error message doesn't include the name
6114 of the failing file when there is a problem reading it.
6116 26. Expand error message when OpenSSL has problems setting up cert/key files.
6119 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6120 (this can affect the format of dates).
6122 28. exim_tidydb, when checking for the continued existence of a message for
6123 which it has found a message-specific retry record, was not finding
6124 messages that were in split spool directories. Consequently, it was
6125 deleting retry records that should have stayed in existence.
6127 29. eximstats updated to version 1.35
6128 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6129 1.35 - bugfix such that pie charts by volume are generated correctly
6131 30. The SPA authentication driver was not abandoning authentication and moving
6132 on to the next authenticator when an expansion was forced to fail,
6133 contradicting the general specification for all authenticators. Instead it
6134 was generating a temporary error. It now behaves as specified.
6136 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6137 (the order specifies the preference for clients). The order is now AES256,
6138 AES128, 3DES, ARCFOUR128.
6140 31. Small patch to Sieve code - explicitly set From: when generating an
6143 32. Exim crashed if a remote delivery caused a very long error message to be
6144 recorded - for instance if somebody sent an entire SpamAssassin report back
6145 as a large number of 550 error lines. This bug was coincidentally fixed by
6146 increasing the size of one of Exim's internal buffers (big_buffer) that
6147 happened as part of the Exiscan merge. However, to be on the safe side, I
6148 have made the code more robust (and fixed the comments that describe what
6151 33. Some experimental protocols are using DNS PTR records for new purposes. The
6152 keys for these records are domain names, not reversed IP addresses. The
6153 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6154 leaves it alone. Component reversal etc. now happens only for IP addresses.
6157 34. The host_aton() function is supposed to be passed a string that is known
6158 to be a valid IP address. However, in the case of IPv6 addresses, it was
6159 not checking this. This is a hostage to fortune. Exim now panics and dies
6160 if the condition is not met. A case was found where this could be provoked
6161 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6162 components; fortuitously, this particular loophole had already been fixed
6163 by change 4.50/55 or 4.44/33 above.
6165 If there are any other similar loopholes, the new check in host_aton()
6166 itself should stop them being exploited. The report I received stated that
6167 data on the command line could provoke the exploit when Exim was running as
6168 exim, but did not say which command line option was involved. All I could
6169 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6170 running as the user.
6173 35. There was a buffer overflow vulnerability in the SPA authentication code
6174 (which came originally from the Samba project). I have added a test to the
6175 spa_base64_to_bits() function which I hope fixes it.
6178 36. The daemon start-up calls getloadavg() while still root for those OS that
6179 need the first call to be done as root, but it missed one case: when
6180 deliver_queue_load_max is set with deliver_drop_privilege. This is
6181 necessary for the benefit of the queue runner, because there is no re-exec
6182 when deliver_drop_privilege is set.
6184 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6185 was in use, was not putting the data itself into the right store pool;
6186 consequently, it could be overwritten for a subsequent message in the same
6187 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6190 38. Sometimes the final signoff response after QUIT could fail to get
6191 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6192 < 0 before doing a fflush(). This bug looks as though it goes back to the
6193 introduction of TLS in release 3.20, but "sometimes" must have been rare
6194 because the tests only now provoked it.
6200 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6201 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6202 Now it returns an integer. A number of calls were still expecting a T/F
6203 return. Fortuitously, in all cases, the tests worked in OK situations,
6204 which is the norm. However, things would have gone wrong on any write
6205 failures on the smtp file descriptor. This function is used when sending
6206 messages over SMTP and also when doing verify callouts.
6208 2. When Exim is called to do synchronous delivery of a locally submitted
6209 message (the -odf or -odi options), it no longer closes stderr before doing
6212 3. Implemented the mua_wrapper option.
6214 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6216 5. Implemented the functions header_remove(), header_testname(),
6217 header_add_at_position(), and receive_remove_recipient(), and exported them
6220 6. If an ACL "warn" statement specified the addition of headers, Exim already
6221 inserted X-ACL-Warn: at the start if there was no header name. However, it
6222 was not making this test for the second and subsequent header lines if
6223 there were newlines in the string. This meant that an invalid header could
6224 be inserted if Exim was badly configured.
6226 7. Allow an ACL "warn" statement to add header lines at the start or after all
6227 the Received: headers, as well as at the end.
6229 8. Added the rcpt_4xx retry error code.
6231 9. Added postmaster_mailfrom=xxx to callout verification option.
6233 10. Added mailfrom=xxxx to the callout verification option, for verify=
6236 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6237 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6239 12. Inserted some casts to stop certain compilers warning when using pointer
6240 differences as field lengths or precisions in printf-type calls (mostly
6241 affecting debugging statements).
6243 13. Added optional readline() support for -be (dynamically loaded).
6245 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6246 same clock tick as a message's arrival, so that its received time was the
6247 same as the "first fail" time on the retry record, and that message
6248 remained on the queue past the ultimate address timeout, every queue runner
6249 would try a delivery (because it was past the ultimate address timeout) but
6250 after another failure, the ultimate address timeout, which should have then
6251 bounced the address, did not kick in. This was a "< instead of <=" error;
6252 in most cases the first failure would have been in the next clock tick
6253 after the received time, and all would be well.
6255 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6256 being recognized when the domain list was tested by the match_domain
6257 condition in an expansion string.
6259 16. Added the ${str2b64: operator.
6261 17. Exim was always calling setrlimit() to set a large limit for the number of
6262 processes, without checking whether the existing limit was already
6263 adequate. (It did check for the limit on file descriptors.) Furthermore,
6264 errors from getrlimit() and setrlimit() were being ignored. Now they are
6265 logged to the main and panic logs, but Exim does carry on, to try to do its
6266 job under whatever limits there are.
6268 18. Imported PCRE 5.0.
6270 19. Trivial typo in log message " temporarily refused connection" (the leading
6273 20. If the log selector return_path_on_delivery was set and an address was
6274 redirected to /dev/null, the delivery process crashed because it assumed
6275 that a return path would always be set for a "successful" delivery. In this
6276 case, the whole delivery is bypassed as an optimization, and therefore no
6279 21. Internal re-arrangement: the function for sending a challenge and reading
6280 a response while authentication was assuming a zero-terminated challenge
6281 string. It's now changed to take a pointer and a length, to allow for
6282 binary data in such strings.
6284 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6286 23. Exim was not respecting finduser_retries when seeking the login of the
6287 uid under which it was called; it was always trying 10 times. (The default
6288 setting of finduser_retries is zero.) Also, it was sleeping after the final
6289 failure, which is pointless.
6291 24. Implemented tls_on_connect_ports.
6293 25. Implemented acl_smtp_predata.
6295 26. If the domain in control=submission is set empty, Exim assumes that the
6296 authenticated id is a complete email address when it generates From: or
6297 Sender: header lines.
6299 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6300 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6301 chown and chgrp in /bin and hostname in /usr/bin.
6303 28. Exim was keeping the "process log" file open after each use, just as it
6304 does for the main log. This opens the possibility of it remaining open for
6305 long periods when the USR1 signal hits a daemon. Occasional processlog
6306 errors were reported, that could have been caused by this. Anyway, it seems
6307 much more sensible not to leave this file open at all, so that is what now
6310 29. The long-running daemon process does not normally write to the log once it
6311 has entered its main loop, and it closes the log before doing so. This is
6312 so that log files can straightforwardly be renamed and moved. However,
6313 there are a couple of unusual error situations where the daemon does write
6314 log entries, and I had neglected to close the log afterwards.
6316 30. The text of an SMTP error response that was received during a remote
6317 delivery was being truncated at 512 bytes. This is too short for some of
6318 the long messages that one sometimes sees. I've increased the limit to
6321 31. It is now possible to make retry rules that apply only when a message has a
6322 specific sender, in particular, an empty sender.
6324 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6325 it possible to be selective about when SMTP synchronization is enforced.
6327 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6329 32. Implemented hosts_connection_nolog.
6331 33. Added an ACL for QUIT.
6333 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6336 35. Added mailbox_size and mailbox_filecount to appendfile.
6338 36. Added control = no_multiline_responses to ACLs.
6340 37. There was a bug in the logic of the code that waits for the clock to tick
6341 in the case where the clock went backwards by a substantial amount such
6342 that the microsecond fraction of "now" was more than the microsecond
6343 fraction of "then" (but the whole seconds number was less).
6345 38. Added support for the libradius Radius client library this is found on
6346 FreeBSD (previously only the radiusclient library was supported).
6352 1. When certain lookups returned multiple values in the form name=value, the
6353 quoting of the values was not always being done properly. Specifically:
6354 (a) If the value started with a double quote, but contained no whitespace,
6356 (b) If the value contained whitespace other than a space character (i.e.
6357 tabs or newlines or carriage returns) it was not quoted.
6358 This fix has been applied to the mysql and pgsql lookups by writing a
6359 separate quoting function and calling it from the lookup code. The fix
6360 should probably also be applied to nisplus, ibase and oracle lookups, but
6361 since I cannot test any of those, I have not disturbed their existing code.
6363 2. A hit in the callout cache for a specific address caused a log line with no
6364 reason for rejecting RCPT. Now it says "Previous (cached) callout
6365 verification failure".
6367 3. There was an off-by-one bug in the queryprogram router. An over-long
6368 return line was truncated at 256 instead of 255 characters, thereby
6369 overflowing its buffer with the terminating zero. As well as fixing this, I
6370 have increased the buffer size to 1024 (and made a note to document this).
6372 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6373 when Exim is waiting for an SMTP response from a remote server, Exim
6374 restarts its select() call on the socket, thereby resetting its timeout.
6375 This is not a problem when such interrupts are rare. Somebody set up a cron
6376 job to run exiwhat every 2 minutes, which is less than the normal select()
6377 timeout (5 or 10 minutes). This meant that the select() timeout never
6378 kicked in because it was always reset. I have fixed this by comparing the
6379 time when an interrupt arrives with the time at the start of the first call
6380 to select(). If more time than the timeout has elapsed, the interrupt is
6381 treated as a timeout.
6383 5. Some internal re-factoring in preparation for the addition of Sieve
6384 extensions (by MH). In particular, the "personal" test is moved to a
6385 separate function, and given an option for scanning Cc: and Bcc: (which is
6386 not set for Exim filters).
6388 6. When Exim created an email address using the login of the caller as the
6389 local part (e.g. when creating a From: or Sender: header line), it was not
6390 quoting the local part when it contained special characters such as @.
6392 7. Installed new OpenBSD configuration files.
6394 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6395 try to make them clearer.
6397 9. Callout options, other than the timeout value, were being ignored when
6398 verifying sender addresses in header lines. For example, when using
6400 verify = header_sender/callout=no_cache
6402 the cache was (incorrectly) being used.
6404 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6405 only the Cygwin environment.
6407 11. When return_path_on_delivery was set as a log selector, if different remote
6408 addresses in the same message used different return paths and parallel
6409 remote delivery occurred, the wrong values would sometimes be logged.
6410 (Whenever a remote delivery process finished, the return path value from
6411 the most recently started remote delivery process was logged.)
6413 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6414 lines when AUTH and/or TLS are in use. This is the "received protocol"
6415 field. Exim used to use "asmtp" for authenticated SMTP, without any
6416 indication (in the protocol name) for TLS use. Now it follows the RFC and
6417 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6418 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6419 names appear in log lines as well as in Received: header lines.
6421 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6422 extensions, and comparison tests, and to fix some bugs.
6424 14. Changes to the "personal" filter test:
6426 (1) The test was buggy in that it was just doing the equivalent of
6427 "contains" tests on header lines. For example, if a user's address was
6428 anne@some.where, the "personal" test would incorrectly be true for
6430 To: susanne@some.where
6432 This test is now done by extracting each address from the header in turn,
6433 and checking the entire address. Other tests that are part of "personal"
6434 are now done using regular expressions (for example, to check local parts
6435 of addresses in From: header lines).
6437 (2) The list of non-personal local parts in From: addresses has been
6438 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6439 taken from the Sieve specification recommendations.
6441 (3) If the message contains any header line starting with "List-" it is
6442 treated as non-personal.
6444 (4) The test for "circular" in the Subject: header line has been removed
6445 because it now seems ill-conceived.
6447 15. Minor typos in src/EDITME comments corrected.
6449 16. Installed latest exipick from John Jetmore.
6451 17. If headers_add on a router specified a text string that was too long for
6452 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6453 of string_sprintf() is now avoided.
6455 18. $message_body_size was not set (it was always zero) when running the DATA
6456 ACL and the local_scan() function.
6458 19. For the "mail" command in an Exim filter, no default was being set for
6459 the once_repeat time, causing a random time value to be used if "once" was
6460 specified. (If the value happened to be <= 0, no repeat happened.) The
6461 default is now 0s, meaning "never repeat". The "vacation" command was OK
6462 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6463 (I found it when inspecting the code).
6465 20. There is now an overall timeout for performing a callout verification. It
6466 defaults to 4 times the callout timeout, which applies to individual SMTP
6467 commands during the callout. The overall timeout applies when there is more
6468 than one host that can be tried. The timeout is checked before trying the
6469 next host. This prevents very long delays if there are a large number of
6470 hosts and all are timing out (e.g. when the network connections are timing
6471 out). The value of the overall timeout can be changed by specifying an
6472 additional sub-option for "callout", called "maxwait". For example:
6474 verify = sender/callout=5s,maxwait=20s
6476 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6477 to the end before writing, but this should make it even safer).
6479 22. Exim was forgetting that it had advertised PIPELINING for the second and
6480 subsequent messages on an SMTP connection. It was also not resetting its
6481 memory on STARTTLS and an internal HELO.
6483 23. When Exim logs an SMTP synchronization error within a session, it now
6484 records whether PIPELINING has been advertised or not.
6486 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6487 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6488 rather than long int.
6490 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6492 26. Added the never_mail option to autoreply.
6498 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6499 crash if the getsockname() call failed; this can happen if a connection is
6500 closed very soon after it is established. The problem was simply in the
6501 order in which certain operations were done, causing Exim to try to write
6502 to the SMTP stream before it had set up the file descriptor. The bug has
6503 been fixed by making things happen in the correct order.
6509 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6510 before the connection was closed, thus losing the rejection response.
6512 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6513 some early Solaris releases, but causes trouble in current releases where
6514 socklen_t is defined.
6516 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6519 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6522 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6523 that releasing the top of it at the end releases what was used for sub-
6524 expansions (unless the block got too big). However, discard this block if
6525 the first thing is a variable or header, so that we can use its block when
6526 it is dynamic (useful for very large $message_headers, for example).
6528 6. Lookups now cache *every* query, not just the most recent. A new, separate
6529 store pool is used for this. It can be recovered when all lookup caches are
6530 flushed. Lookups now release memory at the end of their result strings.
6531 This has involved some general refactoring of the lookup sources.
6533 7. Some code has been added to the store_xxx() functions to reduce the amount
6534 of flapping under certain conditions.
6536 8. log_incoming_interface used to affect only the <= reception log lines. Now
6537 it causes the local interface and port to be added to several more SMTP log
6538 lines, for example "SMTP connection from", and rejection lines.
6540 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6542 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6544 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6545 could be overwritten at the end of the current message (or the start of a
6546 new message if it was set in a HELO ACL). The value is now preserved for
6547 the duration of the SMTP connection.
6549 12. If a transport had a headers_rewrite setting, and a matching header line
6550 contained an unqualified address, that address was qualified, even if it
6551 did not match any rewriting rules. The underlying bug was that the values
6552 of the flags that permit the existence of unqualified sender and recipient
6553 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6554 non-local messages, and by -bnq for local messages) were not being
6555 preserved with the message after it was received.
6557 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6558 "next input=" as part of the text comprising the host identity instead of
6559 the correct text. The code was using the same buffer for two different
6560 strings. However, depending on which order the printing function evaluated
6561 its arguments, the bug did not always show up. Under Linux, for example, my
6562 test suite worked just fine.
6564 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6565 doesn't work with some older versions of Perl. It has been changed to "my",
6566 which in any case is probably the better facility to use.
6568 15. A really picky compiler found some instances of statements for creating
6569 error messages that either had too many or two few arguments for the format
6572 16. The size of the buffer for calls to the DNS resolver has been increased
6573 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6574 for addresses that have a lot of PTR records. This alleviates a problem; it
6575 does not fully solve it.
6577 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6578 buffer now truncates the list and logs the incident, which is the same
6579 action as happens when Exim is looking up a host name and its aliases.
6580 Previously in this situation something unpredictable would happen;
6581 sometimes it was "internal error: store_reset failed".
6583 18. If a server dropped the connection unexpectedly when an Exim client was
6584 using GnuTLS and trying to read a response, the client delivery process
6585 crashed while trying to generate an error log message.
6587 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6588 string, for example:
6590 warn message = H1: something\nH2: something
6592 the text was added as a single header line from Exim's point of view
6593 though it ended up OK in the delivered message. However, searching for the
6594 second and subsequent header lines using $h_h2: did not work. This has been
6595 fixed. Similarly, if a system filter added multiple headers in this way,
6596 the routers could not see them.
6598 20. Expanded the error message when iplsearch is called with an invalid key to
6599 suggest using net-iplsearch in a host list.
6601 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6602 ACLs are no longer actually imposed (and a message to that effect is
6605 22. If a "gecos" field in a passwd entry contained escaped characters, in
6606 particular, if it contained a \" sequence, Exim got it wrong when building
6607 a From: or a Sender: header from that name. A second bug also caused
6608 incorrect handling when an unquoted " was present following a character
6609 that needed quoting.
6611 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6612 was not being matched caselessly.
6614 24. Arranged for all hyphens in the exim.8 source to be escaped with
6617 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6618 properly. Recipient callout cache records were still being keyed to include
6619 the sender, even when use_sender was set false. This led to far more
6620 callouts that were necessary. The sender is no longer included in the key
6621 when use_sender is false.
6623 26. Added "control = submission" modifier to ACLs.
6625 27. Added the ${base62d: operator to decode base 62 numbers.
6627 28. dnsdb lookups can now access SRV records.
6629 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6630 the configuration file.
6632 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6633 (-v) mode. This makes the output for a verbose queue run more intelligible.
6635 31. Added a use_postmaster feature to recipient callouts.
6637 32. Added the $body_zerocount variable, containing the number of binary zero
6638 bytes in the message body.
6640 33. The time of last modification of the "new" subdirectory is now used as the
6641 "mailbox time last read" when there is a quota error for a maildir
6644 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6646 35. Added +ignore_unknown as a special item in host lists.
6648 36. Code for decoding IPv6 addresses in host lists is now included, even if
6649 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6650 address was recognized as an IP address, but was then not correctly decoded
6651 into binary, causing unexpected and incorrect effects when compared with
6658 1. Very minor rewording of debugging text in manualroute to say "list of
6659 hosts" instead of "hostlist".
6661 2. If verify=header_syntax was set, and a header line with an unqualified
6662 address (no domain) and a large number of spaces between the end of the
6663 name and the colon was received, the reception process suffered a buffer
6664 overflow, and (when I tested it) crashed. This was caused by some obsolete
6665 code that should have been removed. The fix is to remove it!
6667 3. When running in the test harness, delay a bit after writing a bounce
6668 message to get a bit more predictability in the log output.
6670 4. Added a call to search_tidyup() just before forking a reception process. In
6671 theory, someone could use a lookup in the expansion of smtp_accept_max_
6672 per_host which, without the tidyup, could leave open a database connection.
6674 5. Added the variables $recipient_data and $sender_data which get set from a
6675 lookup success in an ACL "recipients" or "senders" condition, or a router
6676 "senders" option, similar to $domain_data and $local_part_data.
6678 6. Moved the writing of debug_print from before to after the "senders" test
6681 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6682 problems for message scanning, either using a data ACL, or using
6683 local_scan() because the Received: header was not generated till after they
6684 were called (in order to set the time as the time of reception completion).
6685 I have revised the way this works. The header is now generated after the
6686 body is received, but before the ACL or local_scan() are called. After they
6687 are run, the timestamp in the header is updated.
6693 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6694 before starting a queue runner without re-exec. This happened only when
6695 deliver_drop_privilege was set or when the Exim user was set to root. The
6696 effect of the bug was that timeouts during subsequent deliveries caused
6697 crashes instead of being properly handled. The handler is now left at its
6698 default (and expected) setting.
6700 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6701 message, again when deliver_drop_privilege is set or Exim is run as root.
6702 The bug described in (1) was not present in this case, but the tidying up
6703 of the other signals was missing. I have made the two cases consistent.
6705 3. The ignore_target_hosts setting on a manualroute router was being ignored
6706 for hosts that were looked up using the /MX notation.
6708 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6711 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6712 operated on the sender address. After changing the $sender_address to <>
6713 for the sender address verify, Exim was re-instated it as the original
6714 (before rewriting) address, but remembering that it had rewritten it, so it
6715 wasn't rewriting it again. This bug also had the effect of breaking the
6716 sender address verification caching when the sender address was rewritten.
6718 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6719 This has been changed so that if the ip literal address matches
6720 ignore_target_hosts, the router declines.
6722 7. Added expansion conditions match_domain, match_address, and match_local_
6723 part (NOT match_host).
6725 8. The placeholder for the Received: header didn't have a length field set.
6727 9. Added code to Exim itself and to exim_lock to test for a specific race
6728 condition that could lead to file corruption when using MBX delivery. The
6729 issue is with the lockfile that is created in /tmp. If this file is removed
6730 after a process has opened it but before that process has acquired a lock,
6731 there is the potential for a second process to recreate the file and also
6732 acquire a lock. This could lead to two Exim processes writing to the file
6733 at the same time. The added code performs the same test as UW imapd; it
6734 checks after acquiring the lock that its file descriptor still refers to
6735 the same named file.
6737 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6738 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6741 11. Added the smtp_active_hostname option. If used, this will typically be made
6742 to depend on the incoming interface address. Because $interface_address is
6743 not set up until the daemon has forked a reception process, error responses
6744 that can happen earlier (such as "too many connections") no longer contain
6747 12. If an expansion in a condition on a "warn" statement fails because a lookup
6748 defers, the "warn" statement is abandoned, and the next ACL statement is
6749 processed. Previously this caused the whole ACL to be aborted.
6751 13. Added the iplsearch lookup type.
6753 14. Added ident_timeout as a log selector.
6755 15. Added tls_certificate_verified as a log selector.
6757 16. Added a global option tls_require_ciphers (compare the smtp transport
6758 option of the same name). This controls incoming TLS connections.
6760 17. I finally figured out how to make tls_require_ciphers do a similar thing
6761 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6762 before starting the TLS session.
6764 18. Tabs are now shown as \t in -bP output.
6766 19. If the log selector return_path_on_delivery was set, Exim crashed when
6767 bouncing a message because it had too many Received: header lines.
6769 20. If two routers both had headers_remove settings, and the first one included
6770 a superfluous trailing colon, the final name in the first list and the
6771 first name in the second list were incorrectly joined into one item (with a
6772 colon in the middle).
6778 1. Added -C and -D options to the exinext utility, mainly to make it easier
6779 to include in the automated testing, but these could be helpful when
6780 multiple configurations are in use.
6782 2. The exinext utility was not formatting the output nicely when there was
6783 an alternate port involved in the retry record key, nor when there was a
6784 message id as well (for retries that were specific to a specific message
6785 and a specific host). It was also confused by IPv6 addresses, because of
6786 the additional colons they contain. I have fixed the IPv4 problem, and
6787 patched it up to do a reasonable job for IPv6.
6789 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6790 delivery, the log line now contains "pipelined" if PIPELINING was used.
6792 4. An SMTP transport process used to panic and die if the bind() call to set
6793 an explicit outgoing interface failed. This has been changed; it is now
6794 treated in the same way as a connect() failure.
6796 5. A reference to $sender_host_name in the part of a conditional expansion
6797 that was being skipped was still causing a DNS lookup. This no longer
6800 6. The def: expansion condition was not recognizing references to header lines
6801 that used bh_ and bheader_.
6803 7. Added the _cache feature to named lists.
6805 8. The code for checking quota_filecount in the appendfile transport was
6806 allowing one more file than it should have been.
6808 9. For compatibility with Sendmail, the command line option
6816 and sets the incoming protocol and host name (for trusted callers). The
6817 host name and its colon can be omitted when only the protocol is to be set.
6818 Note the Exim already has two private options, -pd and -ps, that refer to
6819 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6820 "s", but I don't think that's a major issue.
6822 10. A number of refactoring changes to the code, none of which should affect
6825 (a) The number of logging options was getting close to filling up the
6826 32-bit word that was used as a bit map. I have split them into two classes:
6827 those that are passed in the argument to log_write(), and those that are
6828 only ever tested independently outside of that function. These are now in
6829 separate 32-bit words, so there is plenty of room for expansion again.
6830 There is no change in the user interface or the logging behaviour.
6832 (b) When building, for example, log lines, the code previously used a
6833 macro that called string_cat() twice, in order to add two strings. This is
6834 not really sufficiently general. Furthermore, there was one instance where
6835 it was actually wrong because one of the argument was used twice, and in
6836 one call a function was used. (As it happened, calling the function twice
6837 did not affect the overall behaviour.) The macro has been replaced by a
6838 function that can join an arbitrary number of extra strings onto a growing
6841 (c) The code for expansion conditions now uses a table and a binary chop
6842 instead of a serial search (which was left over from when there were very
6843 few conditions). Also, it now recognizes conditions like "pam" even when
6844 the relevant support is not compiled in: a suitably worded error message is
6845 given if an attempt is made to use such a condition.
6847 11. Added ${time_interval:xxxxx}.
6849 12. A bug was causing one of the ddress fields not to be passed back correctly
6850 from remote delivery subprocesses. The field in question was not being
6851 subsequently used, so this caused to problems in practice.
6853 13. Added new log selectors queue_time and deliver_time.
6855 14. Might have fixed a bug in maildirsizefile handling that threw up
6856 "unexpected character" debug warnings, and recalculated the data
6857 unnecessarily. In any case, I expanded the warning message to give more
6860 15. Added the message "Restricted characters in address" to the statements in
6861 the default ACL that block characters like @ and % in local parts.
6863 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6864 Three changes have been made:
6866 (a) There was a serious bug; a negative response to MAIL caused the whole
6867 recipient domain to be cached as invalid, thereby blocking all messages
6868 to all local parts at the same domain, from all senders. This bug has
6869 been fixed. The domain is no longer cached after a negative response to
6870 MAIL if the sender used is not empty.
6872 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6875 (c) A new callout option, "use_sender" has been added for people who want
6876 the modified behaviour.
6882 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6885 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6886 indeed breaks things for older releases.
6888 3. Added additional logging to the case where there is a problem reading data
6889 from a filter that is running in a subprocess using a pipe, in order to
6890 try to track down a specific problem.
6892 4. Testing facility fudge: when running in the test harness and attempting
6893 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6894 getting "No route to host". Convert this to a timeout.
6896 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6899 6. Some OS don't have socklen_t but use size_t instead. This affects the
6900 fifth argument of getsockopt() amongst other things. This is now
6901 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6902 can be set for individual OS. I have set it for SunOS5, OSF1, and
6903 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6904 some earlier ones do not.
6906 7. Change 4.30/15 was not doing the test caselessly.
6908 8. The standard form for an IPv6 address literal was being rejected by address
6909 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6910 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6911 this, as well as the form without the "IPv6" on the front (but only when
6912 address literals are enabled, of course).
6914 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6916 10. Exim crashed if a message with an empty sender address specified by -f
6917 encountered a router with an errors_to setting. This could be provoked only
6918 by a command such as
6922 where an empty string was supplied; "<>" did not hit this bug.
6924 11. Installed PCRE release 4.5.
6926 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6927 remained set. It is now erased.
6929 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6930 times from message ids (which are base 36 rather than the normal 62).
6932 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6933 were being counted as actual protocol errors, and logged if the log
6934 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6935 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6936 and DATA following a set of rejected RCPTs do not count as protocol errors.
6937 In other words, Exim assumes they were pipelined, though this may not
6938 actually be the case. Of course, in all cases the client gets an
6939 appropriate error code.
6941 15. If a lookup fails in an ACL condition, a message about the failure may
6942 be available; it is used if testing the ACL cannot continue, because most
6943 such messages specify what the cause of the deferral is. However, some
6944 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6945 that caused an old message to be retained and used if a later statement
6946 caused a defer, replacing the real cause of the deferral.
6948 16. If an IP address had so many PTR records that the DNS lookup buffer
6949 was not large enough to hold them, Exim could crash while trying to process
6950 the truncated data. It now detects and logs this case.
6952 17. Further to 4.21/58, another change has been made: if (and only if) the
6953 first line of a message (the first header line) ends with CRLF, a bare LF
6954 in a subsequent header line has a space inserted after it, so as not to
6955 terminate the header.
6957 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6958 unnecessarily, used atoi() instead of strtol(), and didn't check the
6959 termination when getting file sizes from file names by regex.
6961 19. Completely re-implemented the support for maildirsize files, in the light
6962 of a number of problems with the previous contributed implementation
6963 (4.30/29). In particular:
6965 . If the quota is zero, the maildirsize file is maintained, but no quota is
6968 . If the maildir directory does not exist, it is created before any attempt
6969 to write a maildirsize file.
6971 . The quota value in the file is just a cache; if the quota is changed in
6972 the transport, the new value overrides.
6974 . A regular expression is available for excluding directories from the
6977 20. The autoreply transport checks the characters in options that define the
6978 message's headers; it allows continued headers, but it was checking with
6979 isspace() after an embedded newline instead of explicitly looking for a
6982 21. If all the "regular" hosts to which an address was routed had passed their
6983 expiry times, and had not reached their retry times, the address was
6984 bounced, even if fallback hosts were defined. Now Exim should go on to try
6987 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6988 equivalent code in the SMTP transport. Some hosts send humungous responses
6989 to HELO/EHLO, more than 1024 it seems.
6991 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6992 gives compiler warnings in some environments. I've now done it "properly",
6995 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6996 (because of problems with the built-in one) was declared to return uschar *
6997 instead of char *, causing compiler failure.
6999 25. Fixed a file descriptor leak when processing alias/forward files.
7001 26. Fixed a minor format string issue in dbfn.c.
7003 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7005 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7006 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7009 29. When a sender address is verified, it is cached, to save repeating the test
7010 when there is more than one recipient in a message. However, when the
7011 verification involves a callout, it is possible for different callout
7012 options to be set for different recipients. It is too complicated to keep
7013 track of this in the cache, so now Exim always runs a verification when a
7014 callout is required, relying on the callout cache for the optimization.
7015 The overhead is duplication of the address routing, but this should not be
7018 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7019 to be verified with callout=postmaster, and the main callout worked but the
7020 postmaster check failed, the verification correctly failed. However, if a
7021 subsequent RCPT command asked for sender verification *without* the
7022 postmaster check, incorrect caching caused this verification also to fail,
7025 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7026 it was not caching the DNS options (qualify_single, search_parents) that
7027 were used when the lookup failed. A subsequent lookup with different
7028 options therefore always gave the same answer, though there were cases
7029 where it should not have. (Example: a "domains = !$mx_any" option on a
7030 dnslookup router: the "domains" option is always processed without any
7031 widening, but the router might have qualify_single set.) Now Exim uses the
7032 cached value only when the same options are set.
7034 32. Added John Jetmore's "exipick" utility to the distribution.
7036 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7037 than a timeout (e.g. a certificate is required, and is not provided), an
7038 Exim server now closes the connection immediately. Previously it waited for
7039 the client to close - but if the client is SSL, it seems that they each
7040 wait for each other, leading to a delay before one of them times out.
7042 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7043 maintained 0.8.x compatibility because I don't think many are using it, and
7044 it is clearly obsolete.
7046 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7049 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7050 client certificate was expired. A simple patch fixes this, though I don't
7051 understand the full logic of why the verify callback is called multiple
7054 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7055 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7056 which causes problems with some clients (such as the Certicom SSL Plus
7057 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7058 disables the coutermeasure allowing Eudora to connect."
7060 38. Exim was not checking that a write() to a log file succeeded. This could
7061 lead to Bad Things if a log got too big, in particular if it hit a file
7062 size limit. Exim now panics and dies if it cannot write to a log file, just
7063 as it does if it cannot open a log file.
7065 39. Modified OS/Makefile-Linux so that it now contains
7067 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7069 The two -D definitions ensure that Exim is compiled with large file
7070 support, which makes it possible to handle log files that are bigger than
7073 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7074 instance) a domain was checked against a named list that involved a lookup,
7075 causing $domain_data to be set, then another domain was checked against the
7076 same list, then the first domain was re-checked, the value of $domain_data
7077 after the final check could be wrong. In particular, if the second check
7078 failed, it could be set empty. This bug probably also applied to
7081 41. The strip_trailing_dot option was not being applied to the address given
7082 with the -f command-line option.
7084 42. The code for reading a message's header from the spool was incrementing
7085 $received_count, but never initializing it. This meant that the value was
7086 incorrect (doubled) while delivering a message in the same process in which
7087 it was received. In the most common configuration of Exim, this never
7088 happens - a fresh exec is done - but it can happen when
7089 deliver_drop_privilege is set.
7091 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7092 it now includes up to 150 characters of the unexpected data in the log
7095 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7096 and building data strings. The size of both of these buffers was 10 000
7097 bytes - far larger than anybody would *ever* want, thought I. Needless to
7098 say, somebody hit the limit. I have increased the maximum line length to
7099 20 000 and the maximum data length of concatenated lines to 100 000. I have
7100 also fixed two bugs, because there was no checking on these buffers. Tsk,
7101 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7102 buffer is too small.
7104 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7105 lsearch lookups. Now it does.
7107 46. When parsing a route_list item in a manualroute router, a fixed-length
7108 buffer was used for the list of hosts. I made this 1024 bytes long,
7109 thinking that nobody would ever have a list of hosts that long. Wrong.
7110 Somebody had a whole pile of complicated expansion conditions, and the
7111 string was silently truncated, leading to an expansion error. It turns out
7112 that it is easier to change to an unlimited length (owing to other changes
7113 that have happened since this code was originally written) than to build
7114 structure for giving a limitation error. The length of the item that
7115 expands into the list of hosts is now unlimited.
7117 47. The lsearch lookup could not handle data where the length of text line was
7118 more than 4095 characters. Such lines were truncated, leading to shortened
7119 data being returned. It should now handle lines of any length.
7121 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7122 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7125 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7126 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7127 no longer happens in comment lines. A list of replacements is now placed
7128 at the head of all of the source files, except those whose only change is
7129 to replace PERL_COMMAND in the very first #! line.
7131 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7132 messages on the queue, with a bottom-up merge sort, using code contributed
7133 by Michael Haardt. This should make operations like -bp somewhat faster on
7134 large queues. It won't affect queue runners, except when queue_run_in_order
7137 51. Installed eximstats 1.31 in the distribution.
7139 52. Added support for SRV lookups to the dnslookup router.
7141 53. If an ACL referred to $message_body or $message_body_end, the value was not
7142 reset for any messages that followed in the same SMTP session.
7144 54. The store-handling optimization for building very long strings was not
7145 differentiating between the different store pools. I don't think this
7146 actually made any difference in practice, but I've tidied it.
7148 55. While running the routers to verify a sender address, $sender_address
7149 was still set to the sender address. This is wrong, because when routing to
7150 send a bounce to the sender, it would be empty. Therefore, I have changed
7151 it so that, while verifying a sender address, $sender_address is set to <>.
7152 (There is no change to what happens when verifying a recipient address.)
7154 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7155 target A or AAAA records (if not already returned) without resetting the
7156 qualify_single or search_parents options of the DNS resolver. These are
7157 inappropriate in this case because the targets of MX and SRV records must
7158 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7159 target that, when qualified, matched something in the local domain. These
7160 two options are now turned off when doing these lookups.
7162 57. It seems that at least some releases of Reiserfs (which does not have the
7163 concept of a fixed number of inodes) returns zero and not -1 for the
7164 number of available inodes. This interacted badly with check_spool_inodes,
7165 which assumed that -1 was the "no such thing" setting. What I have done is
7166 to check that the total number of inodes is greater than zero before doing
7167 the test of how many are available.
7169 58. When a "warn" ACL statement has a log_message modifier, the message is
7170 remembered, and not repeated. This is to avoid a lot of repetition when a
7171 message has many recipients that cause the same warning to be written.
7172 However, Exim was preserving the list of already written lines for an
7173 entire SMTP session, which doesn't seem right. The memory is now reset if a
7174 new message is started.
7176 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7177 debugging output unless log_rewrite was also set.
7179 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7180 of (int)(handle) when we know that handle contains (void *)(-1).
7182 61. The Exim daemon panic-logs an error return when it closes the incoming
7183 connection. However "connection reset by peer" seems to be common, and
7184 isn't really an error worthy of noting specially, so that particular error
7187 62. When Exim is trying to find all the local interfaces, it used to panic and
7188 die if the ioctl to get the interface flags failed. However, it seems that
7189 on at least one OS (Solaris 9) it is possible to have an interface that is
7190 included in the list of interfaces, but for which you get a failure error
7191 for this call. This happens when the interface is not "plumbed" into a
7192 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7193 failure of the "get flags" call assumes that the interface is down.
7195 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7196 makes life easier for people who are doing arithmetic on fields extracted
7197 from dates, where you often get leading zeros that should not be
7198 interpreted as octal.
7200 64. Added qualify_domain to the redirect router, to override the global
7203 65. If a pathologically long header line contained very many addresses (the
7204 report of this problem mentioned 10 000) and each of them was rewritten,
7205 Exim could use up a very large amount of memory. (It kept on making new
7206 copies of the header line as it rewrote, and never released the old ones.)
7207 At the expense of a bit more processing, the header rewriting function has
7208 been changed so that it no longer eats memory in this way.
7210 66. The generation of the Received: header has been moved from the time that a
7211 message starts to be received, to the time that it finishes. The timestamp
7212 in the Received: header should now be very close to that of the <= log
7213 line. There are two side-effects of this change:
7215 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7216 logged header lines no longer include the local Received: line, because
7217 it has not yet been created. The same applies to a copy of the message
7218 that is returned to a non-SMTP sender when a message is rejected.
7220 (b) When a filter file is tested using -bf, no additional Received: header
7221 is added to the test message. After some thought, I decided that this
7224 This change does not affect the value of $received_for. It is still set
7225 after address rewriting, but before local_scan() is called.
7227 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7229 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7230 gave an unhelpful panic error message, and a defer error. I have managed to
7231 change this behaviour so that it now rejects any supplied certificate,
7232 which seems right, as the list of acceptable certificates is empty.
7234 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7235 gave an unhelpful defer error. I have not managed to make this reject any
7236 supplied certificates, but the error message it gives is "no certificate
7237 supplied", which is not helpful.
7239 70. exigrep's output now also includes lines that are not associated with any
7240 message, but which match the given pattern. Implemented by a patch from
7241 Martin Sluka, which also tidied up the Perl a bit.
7243 71. Recipient callout verification, like sender verification, was using <> in
7244 the MAIL FROM command. This isn't really the right thing, since the actual
7245 sender may affect whether the remote host accepts the recipient or not. I
7246 have changed it to use the actual sender in the callout; this means that
7247 the cache record is now keyed on a recipient/sender pair, not just the
7248 recipient address. There doesn't seem to be a real danger of callout loops,
7249 since a callout by the remote host to check the sender would use <>.
7250 [SEE ABOVE: changed after hitting problems.]
7252 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7253 temporary errors. However, in the case of such a code being given after
7254 the end of a data transmission (i.e. after ".") Exim was failing to write
7255 a retry record for the message. (Yes, there was some broken host that was
7256 actually sending 8xx at this point.)
7258 73. An unknown lookup type in a host list could cause Exim to panic-die when
7259 the list was checked. (An example that provoked this was putting <; in the
7260 middle of a list instead of at the start.) If this happened during a DATA
7261 ACL check, a -D file could be left lying around. This kind of configuration
7262 error no longer causes Exim to die; instead it causes a defer error. The
7263 incident is still logged to the main and panic logs.
7265 74. Buglet left over from Exim 3 conversion. The message "too many messages
7266 in one connection" was written to the rejectlog but not the mainlog, except
7267 when address rewriting (yes!) was being logged.
7269 75. Added write_rejectlog option.
7271 76. When a system filter was run not as root (that is, when system_filter_user
7272 was set), the values of the $n variables were not being returned to the
7273 main process; thus, they were not subsequently available in the $sn
7276 77. Added +return_path_on_delivery log selector.
7278 78. A connection timeout was being treated differently from recipients deferred
7279 when testing hosts_max_try with a message that was older than the host's
7280 retry timeout. (The host should not be counted, thus allowing all hosts to
7281 be tried at least once before bouncing.) This may have been the cause of an
7282 occasionally reported bug whereby a message would remain on the queue
7283 longer than the retry timeout, but would be bounced if a delivery was
7284 forced. I say "may" because I never totally pinned down the problem;
7285 setting up timeout/retry tests is difficult. See also the next item.
7287 79. The ultimate address timeout was not being applied to errors that involved
7288 a combination of host plus message (for example, a timeout on a MAIL
7289 command). When an address resolved to a number of possible hosts, and they
7290 were not all tried for each delivery (e.g. because of hosts_max_try), a
7291 message could remain on the queue longer than the retry timeout.
7293 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7296 81. Fixed an obscure SMTP outgoing bug which required at least the following
7297 conditions: (a) there was another message waiting for the same server;
7298 (b) the server returned 5xx to all RCPT commands in the first message so
7299 that the message was not completed; (c) the server dropped the connection
7300 or gave a negative response to the RSET that Exim sends to abort the
7301 transaction. The observed case was a dropped connection after DATA that had
7302 been sent in pipelining mode. That is, the server had advertised PIPELINING
7303 but was not implementing it correctly. The effect of the bug was incorrect
7304 behaviour, such as trying another host, and this could lead to a crash.
7310 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7311 and daemon.c were passed as pointers to ints; they should have been
7312 pointers to socklen_t variables (which are typically unsigned ints).
7314 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7317 3. Fixed a really odd bug that affected only the testing scheme; patching a
7318 certain fixed string in the binary changed the value of another string that
7319 happened to be identical to the end of the original first string.
7321 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7322 name", it returns that address as the IP address. On some operating
7323 systems (e.g. Solaris), it also passes back the IP address string as the
7324 "host name". However, on others (e.g. Linux), it passes back an empty
7325 string. Exim wasn't checking for this, and was changing the host name to an
7326 empty string, assuming it had been canonicalized.
7328 5. Although rare, it is permitted to have more than one PTR record for a given
7329 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7330 all the names associated with an address, because they do in Solaris.
7331 However, it seems that they do not in Linux for data that comes from the
7332 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7333 I found this out when I moved to a new Linux workstation and tried to run
7334 the Exim test suite.
7336 To get round this problem I have changed the code so that it now does its
7337 own call to the DNS to look up PTR records when searching for a host name.
7338 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7339 addresses that are only in /etc/hosts are still found.
7341 This behaviour is, however, controlled by an option called host_lookup_
7342 order, which defaults to "bydns:byaddr". If people want to use the other
7343 order, or indeed, just use one or the other means of lookup, they can
7344 specify it in this variable.
7346 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7347 some operating systems, this comes back from gethostbyaddr() as an empty
7348 string, and this is what Exim used to test for. However, it seems that in
7349 other systems, "." is yielded. Exim now tests for this case too.
7351 7. The values of check_spool_space and check_log_space are now held internally
7352 as a number of kilobytes instead of an absolute number of bytes. If a
7353 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7354 kilobyte. This means that much larger values can be stored.
7356 8. Exim monitor: an attempt to get the action menu when not actually pointing
7357 at a message produces an empty menu entitled "No message selected". This
7358 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7359 no entries in it ("Shell widget menu has zero width and/or height"). So I
7360 have added a single, blank menu entry in this case.
7362 9. Added ${quote_local_part.
7364 10. MIME decoding is now applied to the contents of Subject: header lines when
7367 11. Now that a reference to $sender_host_address automatically causes a reverse
7368 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7369 host lookup before query-style lookups in lists that might use this
7370 variable. This has therefore been abolished, and the "net-" prefix is no
7371 longer necessary for query-style lookups.
7373 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7374 have been a typo for LFLAGS, so it has been changed.
7376 13. The install script calls Exim with "-C /dev/null" in order to find the
7377 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7378 to be output. However, since Exim outputs its version number before the
7379 error, it didn't break the script. It just looked ugly. I fixed this by
7380 always allowing "-C /dev/null" if the caller is root.
7382 14. Ignore overlarge ACL variable number when reading spool file - insurance
7383 against a later release with more variables having written the file.
7385 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7386 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7387 this, as well as the form without the "IPv6" on the front.
7389 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7390 OS/Makefile-Darwin file.
7392 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7393 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7395 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7396 variable. However, some OS use ints for inodes. I've added cast to long int
7397 to get rid of the compiler warning.
7399 19. I had forgotten to lock out "/../" in configuration file names when
7400 ALT_CONFIG_PREFIX was set.
7402 20. Routers used for verification do not need to specify transports. However,
7403 if such a router generated a host list, and callout was configured, Exim
7404 crashed, because it could not find a port number from the (non-existent)
7405 transport. It now assumes port 25 in this circumstance.
7407 21. Added the -t option to exigrep.
7409 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7410 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7411 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7412 them is set, LOOKUP_LSEARCH is forced.
7414 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7416 24. Added sender and host information to the "rejected by local_scan()" log
7417 line; previously there was no indication of these.
7419 25. Added .include_if_exists.
7421 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7422 Linux. It turns out that not all file systems support this. Apparently some
7423 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7424 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7425 not supported on the file descriptor, is now ignored when Exim is trying to
7426 sync a directory. This applies only to Linux.
7428 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7430 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7431 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7434 29. Added support for maildirsize files from supplied patch (modified a bit).
7436 30. The use of :fail: followed by an empty string could lead Exim to respond to
7437 sender verification failures with (e.g.):
7439 550 Verification failed for <xxx>
7440 550 Sender verify failed
7442 where the first response line was missing the '-' that indicates it is not
7443 the final line of the response.
7445 31. The loop for finding the name of the user that called Exim had a hardwired
7446 limit of 10; it now uses the value of finduser_retries, which is used for
7447 all other user lookups.
7449 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7452 33. Exim was neglecting to zero errno before one call of strtol() when
7453 expanding a string and expecting an integer value. On some systems this
7454 resulted in spurious "integer overflow" errors. Also, it was casting the
7455 result into an int without checking.
7457 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7458 did not work. The code looks as if it has *never* worked, though it appears
7459 to have been documented since at least release 1.62. I have made it work.
7461 35. The "timeout_DNS" error in retry rules, also documented since at least
7462 1.62, also never worked. As it isn't clear exactly what this means, and
7463 clearly it isn't a major issue, I have abolished the feature by treating it
7464 as "timeout", and writing a warning to the main and panic logs.
7466 36. The display of retry rules for -brt wasn't always showing the error code
7469 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7470 timeout_connect_A, timeout_connect_MX.
7472 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7473 to the empty sender.
7475 39. The daemon was not analysing the content of -oX till after it had closed
7476 stderr and disconnected from the controlling terminal. This meant that any
7477 syntax errors were only noted on the panic log, and the return code from
7478 the command was 0. By re-arranging the code a little, I've made the
7479 decoding happen first, so such errors now appear on stderr, and the return
7480 code is 1. However, the actual setting up of the sockets still happens in
7481 the disconnected process, so errors there are still only recorded on the
7484 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7485 connections (as happens on some IP stacks) was logged at start up time as
7486 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7487 it from "IPv6 and IPv4", which means that two separate sockets are being
7490 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7491 says whether AF_INET or AF_INET6 was passed as an argument.
7493 42. Exiwhat output was messed up when time zones were included in log
7496 43. Exiwhat now gives more information about the daemon's listening ports,
7497 and whether -tls-on-connect was used.
7499 44. The "port" option of the smtp transport is now expanded.
7501 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7502 silently ignored. Now an error message is written to the main and panic
7505 46. There's a new ACL modifier called "logwrite" which writes to a log file
7506 as soon as it is encountered.
7508 47. Added $local_user_uid and $local_user_gid at routing time.
7510 48. Exim crashed when trying to verify a sender address that was being
7513 49. Exim was recognizing only a space character after ".include". It now also
7514 recognizes a tab character.
7516 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7517 extracting the relevant information from the specification. The man page no
7518 longer contains scrambled data for the -d option, and I've added a section
7519 at the front about calling Exim under different names.
7521 51. Added "extra_headers" argument to the "mail" command in filter files.
7523 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7526 53. Installed eximstats 1.29.
7528 54. Added transport_filter_timeout as a generic transport option.
7530 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7531 Cc: header lines. This was required by RFC 822, but it not required by RFC
7534 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7535 incoming messages that did not have them. Now it does so only if the
7536 message originates locally, that is, if there is no associated remote host
7537 address. When Resent- header lines are present, this applies to the Resent-
7538 lines rather than the non-Resent- lines.
7540 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7541 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7543 58. Messages for configuration errors now include the name of the main
7544 configuration file - useful now that there may be more than one file in a
7545 list (.included file names were always shown).
7547 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7548 for those rare installations that do not start the daemon as root or run it
7549 setuid root. I've cut out the call to initgroups() if the daemon is not
7552 60. The Exim user and group can now be bound into the binary as text strings
7553 that are looked up at the start of Exim's processing.
7555 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7557 62. Added $mailstore_basename variable.
7559 63. Installed patch to sieve.c from Michael Haardt.
7561 64. When Exim failed to open the panic log after failing to open the main log,
7562 the original message it was trying to log was written to stderr and debug
7563 output, but if they were not available (the usual case in production), it
7564 was lost. Now it is written to syslog before the two lines that record the
7565 failures to open the logs.
7567 65. Users' Exim filters run in subprocesses under the user's uid. It is
7568 possible for a "deliver" command or an alias in a "personal" command to
7569 provoke an address rewrite. If logging of address rewriting is configured,
7570 this fails because the process is not running as root or exim. There may be
7571 a better way of dealing with this, but for the moment (because 4.30 needs
7572 to be released), I have disabled address rewrite logging when running a
7573 filter in a non-root, non-exim process.
7579 1. The buildconfig auxiliary program wasn't quoting the value set for
7580 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7581 not defined. This bug was masked in 4.22 by the effect that was fixed in
7584 2. Some messages that were rejected after a message id was allocated were
7585 shown as "incomplete" by exigrep. It no longer does this for messages that
7586 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7588 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7589 have allow_domain_literals set, the ID did not get logged in the <= line.
7590 Domain literals are now always recognized in Message-ID: header lines.
7592 4. The first argument for a ${extract expansion item is the key name or field
7593 number. Leading and trailing spaces in this item were not being ignored,
7594 causing some misleading effects.
7596 5. When deliver_drop_privilege was set, single queue runner processes started
7597 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7598 same command in the process it spins off) were not dropping privilege.
7600 6. When the daemon running as "exim" started a queue runner, it always
7601 re-executed Exim in the spun-off process. This is a waste of effort when
7602 deliver_drop_privilege is set. The new process now just calls the
7603 queue-runner function directly.
7609 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7612 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7613 ignored. Though the use of -oP was forcing the writing of a pid file, it
7614 was always written to the default place.
7616 3. If the message "no IP address found for host xxxx" is generated during
7617 incoming verification, it is now followed by identification of the incoming
7618 connection (so you can more easily find what provoked it).
7620 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7622 5. Added some features to "harden" Exim a bit more against certain attacks:
7624 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7625 be put in Local/Makefile. This is like the never_users runtime option,
7626 but it cannot be overridden. The default setting is "root".
7628 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7629 prefix string with which any file named in a -C command line option
7632 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7633 is retained for -C and -D only if the caller of Exim is root. Without
7634 it, the exim user may also use -C and -D and retain privilege.
7636 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7637 command line option is disabled.
7639 6. Macro names set by the -D option must start with an upper case letter, just
7640 like macro names defined in the configuration file.
7642 7. Added "dereference=" facility to LDAP.
7644 8. Two instances of the typo "uknown" in the source files are fixed.
7646 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7647 the Configure-Makefile script screwed up while processing it.
7649 10. Incorporated PCRE 4.4.
7651 11. The SMTP synchronization check was not operating right at the start of an
7652 SMTP session. For example, it could not catch a HELO sent before the client
7653 waited for the greeting. There is now a check for outstanding input at the
7654 point when the greeting is written. Because of the duplex, asynchronous
7655 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7656 way, but not yet received, when the check is performed.
7658 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7659 on TCP/IP sockets, because this apparently causes some broken clients to
7662 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7663 unchanged) from the Cygwin maintainer.
7665 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7666 was supported instead of testing for mbx. Effectively a typo.
7668 15. The spa authenticator server code was not checking that the input it
7669 received was valid base64.
7671 16. The debug output line for the "set" modifier in ACLs was not showing the
7672 name of the variable that was being set.
7674 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7676 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7677 lookup on its own. Something else was needed to trigger the lookup. For
7678 example, a match in host_lookup or the need for a host name in a host list.
7679 Now, if $sender_host_name is referenced and the host name has not yet been
7680 looked up, a lookup is performed. If the lookup fails, the variable remains
7681 empty, and $host_lookup_failed is set to "1".
7683 19. Added "eqi" as a case-independent comparison operator.
7685 20. The saslauthd authentication condition could segfault if neither service
7686 nor realm was specified.
7688 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7689 error message that was logged was misleading, and incoming SMTP
7690 connections were dropped. The message is now more accurate, and temporary
7691 errors are given to SMTP connections.
7693 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7694 (except RSET) with a 421 error, until QUIT is received. However, it was
7695 failing to send a response to QUIT.
7697 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7698 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7699 ACL is now run after the helo_try_verify_hosts code.
7701 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7702 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7703 case-independent, so other case variants are also recognized). Apparently
7704 some systems use these upper case variants.
7706 25. If more than two messages were waiting for the same host, and a transport
7707 filter was specified for the transport, Exim sent two messages over the
7708 same TCP/IP connection, and then failed with "socket operation on non-
7709 socket" when it tried to send the third.
7711 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7713 27. The extern definition of crypt16() in expand.c was not being excluded when
7714 the OS had its own crypt16() function.
7716 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7717 as a preferred synonym for return_size_limit, both as an option and as an
7720 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7722 30. Changed the default configuration ACL to relax the local part checking rule
7723 for addresses that are not in any local domains. For these addresses,
7724 slashes and pipe symbols are allowed within local parts, but the sequence
7725 /../ is explicitly forbidden.
7727 31. SPA server authentication was not clearing the challenge buffer before
7730 32. log_message in a "warn" ACL statement was writing to the reject log as
7731 well as to the main log, which contradicts the documentation and doesn't
7732 seem right (because no rejection is happening). So I have stopped it.
7734 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7735 However, I am unable to do any testing of this.
7737 34. Fixed an infelicity in the appendfile transport. When checking directories
7738 for a mailbox, to see if any needed to be created, it was accidentally
7739 using path names with one or more superfluous leading slashes; tracing
7740 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7742 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7743 contents are added to the log line that is written for every discarded
7744 recipient. (Previously a log_message setting was ignored.)
7746 36. The ${quote: operator now quotes the string if it is empty.
7748 37. The install script runs exim in order to find its version number. If for
7749 some reason other than non-existence or emptiness, which it checks, it
7750 could not run './exim', it was installing it with an empty version number,
7751 i.e. as "exim-". This error state is now caught, and the installation is
7754 38. An argument was missing from the function that creates an error message
7755 when Exim fails to connect to the socket for saslauthd authentication.
7756 This could cause Exim to crash, or give a corrupted message.
7758 39. Added isip, isip4, and isip6 to ${if conditions.
7760 40. The ACL variables $acl_xx are now saved with the message, and can be
7761 accessed later in routers, transports, and filters.
7763 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7764 strings in the file are not string-expanded.
7766 42. If a MAIL command specified a SIZE value that was too large to fit into an
7767 int variable, the check against message_size_limit failed. Such values are
7768 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7769 day this will have to be increased, but I don't think I want to be around
7770 when emails are that large.
7777 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7778 iconv() is not standard in FreeBSD.
7780 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7781 IPv6 enabled. The observed symptom was a segmentation fault on return from
7782 the function os_common_find_running_interfaces() in src/os.c.
7784 3. In the check_special_case() function in daemon.c I had used "errno" as an
7785 argument name, which causes warnings on some systems. This was basically a
7786 typo, since it was named "eno" in the comments!
7788 4. The code that waits for the clock to tick (at a resolution of some fraction
7789 of a second) so as to ensure message-id uniqueness was always waiting for
7790 at least one whole tick, when it could have waited for less. [This is
7791 almost certainly not relevant at current processor speeds, where it is
7792 unlikely to ever wait at all. But we try to future-proof.]
7794 5. The function that sleeps for a time interval that includes fractions of a
7795 second contained a race. It did not block SIGALRM between setting the
7796 timer, and suspending (a couple of lines later). If the interval was short
7797 and the sigsuspend() was delayed until after it had expired, the suspension
7798 never ended. On busy systems this could lead to processes getting stuck for
7801 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7802 process, before it forks any delivery processes. The open lookup caching
7803 mechanism meant that the open file or database connection was passed into
7804 the delivery process. The problem was that delivery processes always tidy
7805 up cached lookup data. This could cause a problem for the next delivery
7806 process started by the queue runner, because the external queue runner
7807 process does not know about the closure. So the next delivery process
7808 still has data in the lookup cache. In the case of a file lookup, there was
7809 no problem because closing a file descriptor in a subprocess doesn't affect
7810 the parent. However, if the lookup was caching a connection to a database,
7811 the connection was closed, and the second delivery process was likely to
7812 see errors such as "PGSQL: query failed: server closed the connection
7813 unexpectedly". The problem has been fixed by closing all cached lookups
7814 in a queue runner before running a delivery process.
7816 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7817 seem to have the "const" qualifier which it has on other OS. I've
7820 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7821 *of the same type* (client or server) with the same public name that an
7822 error should be diagnosed.
7824 9. When Exim looked up a host name for an IP address, but failed to find the
7825 original IP address when looking up the host name (a safety check), it
7826 output the message "<ip address> does not match any IP for NULL", which was
7827 confusing, to say the least. The bug was that the host name should have
7828 appeared instead of "NULL".
7830 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7831 user that is built into the binary, and the -C or -D options is used, root
7832 privilege is dropped before the configuration file is read. In addition,
7833 logging is switched to stderr instead of the normal log files. If the
7834 configuration then re-defines the Exim user, the unprivileged environment
7835 is probably not what is expected, so Exim logs a panic warning message (but
7838 However, if deliver_drop_privilege is set, the unprivileged state may well
7839 be exactly what is intended, so the warning has been cut out in that case,
7840 and Exim is allowed to try to write to its normal log files.
7846 1. smtp_return_error_details was not giving details for temporary sender
7847 or receiver verification errors.
7849 2. Diagnose a configuration error if two authenticators have the same public
7852 3. Exim used not to create the message log file for a message until the first
7853 delivery attempt. This could be confusing when incoming messages were held
7854 for policy or load reasons. The message log file is now created at the time
7855 the message is received, and an initial "Received" line is written to it.
7857 4. The automatically generated man page for command line options had a minor
7858 bug that caused no ill effects; however, a more serious problem was that
7859 the procedure for building the man page automatically didn't always
7860 operate. Consequently, release 4.20 contains an out-of-date version. This
7861 shouldn't happen again.
7863 5. When building Exim with embedded Perl support, the script that builds the
7864 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7865 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7867 6. The freeze_tell option was not being used for messages that were frozen on
7868 arrival, either by an ACL or by local_scan().
7870 7. Added the smtp_incomplete_transaction log selector.
7872 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7873 was accepting AUTH without a new EHLO.
7875 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7876 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7879 10. Logging of TCP/IP connections (when configured) now happens in the main
7880 daemon process instead of the child process, so that the TCP/IP connection
7881 count is more accurate (but it can never be perfect).
7883 11. The use of "drop" in a nested ACL was not being handled correctly in the
7884 outer ACL. Now, if condition failure induced by the nested "drop" causes
7885 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7886 or "require"), the connection is dropped.
7888 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7889 that yield "discard" can now be used with an "accept" or a "discard" verb,
7890 but an error is generated for any others (because I can't see a useful way
7891 to define what should happen).
7893 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7894 are now processed in the same way as lines in the Exim configuration file.
7895 In particular, continuation lines are supported.
7897 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7899 15. Added -ti meaning -t -i.
7901 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7902 domains, and warn by logging if others are found.
7904 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7905 returned from GIFCONF when Exim is trying to find the list of interfaces on
7906 a host. The code in os.c has been modified to copy each ifreq to an aligned
7907 structure in all cases.
7909 Also, in some cases, the returned ifreq's were being copied to a 'struct
7910 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7911 means the last couple of bytes of an IPv6 address could be chopped if the
7912 ifreq contained only a normal sockaddr (14 bytes storage).
7914 18. Named domain lists were not supported in the hosts_treat_as_local option.
7915 An entry such as +xxxx was not recognized, and was treated as a literal
7918 19. Ensure that header lines added by a DATA ACL are included in the reject log
7919 if the ACL subsequently rejects the message.
7921 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7922 MD5 (which is deprecated).
7924 21. When testing a filter file using -bf, Exim was writing a message when it
7925 took the sender from a "From " line in the message, but it was not doing so
7926 when it took $return_path from a Return-Path: header line. It now does.
7928 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7929 with a valid header line field name (a series of printing characters
7930 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7932 23. Changed "disc" in the source to "disk" to conform to the documentation and
7933 the book and for uniformity.
7935 24. Ignore Sendmail's -Ooption=value command line item.
7937 25. When execve() failed while trying to run a command in a pipe transport,
7938 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7939 could be confused with a return value of 69 from the command itself. This
7940 has been changed to 127, the value the shell returns if it is asked to run
7941 a non-existent command. The wording for the related log line suggests a
7942 non-existent command as the problem.
7944 26. If received_header_text expands to an empty string, do not add a Received:
7945 header line to the message. (Well, it adds a token one on the spool, but
7946 marks it "old" so that it doesn't get used or transmitted.)
7948 27. Installed eximstats 1.28 (addition of -nt option).
7950 28. There was no check for failure on the call to getsockname() in the daemon
7951 code. This can fail if there is a shortage of resources on the system, with
7952 ENOMEM, for example. A temporary error is now given on failure.
7954 29. Contrary to the C standard, it seems that in some environments, the
7955 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7956 program. Exim now does this explicitly; it affects the formatting of
7957 timestamps using strftime().
7959 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7960 complaints. I've now initialized all the variables, to avoid this.
7962 32. Header lines added by a system filter were not being "seen" during
7963 transport-time rewrites.
7965 33. The info_callback() function passed to OpenSSL is set up with type void
7966 (*)(SSL *, int, int), as described somewhere. However, when calling the
7967 function (actually a macro) that sets it up, the type void(*)() is
7968 expected. I've put in a cast to prevent warnings from picky compilers.
7970 34. If a DNS black list lookup found a CNAME record, but there were no A
7971 records associated with the domain it pointed at, Exim crashed.
7973 35. If a DNS black list lookup returned more than one A record, Exim ignored
7974 all but the first. It now scans all returned addresses if a particular IP
7975 value is being sought. In this situation, the contents of the
7976 $dnslist_value variable are a list of all the addresses, separated by a
7979 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7980 to accept a host name and all its aliases if the forward lookup for any of
7981 them yielded the IP address of the incoming connection. Now it accepts only
7982 those names whose forward lookup yields the correct IP address. Any other
7983 names are discarded. This closes a loophole whereby a rogue DNS
7984 administrator could create reverse DNS records to break through a
7985 wildcarded host restriction in an ACL.
7987 37. If a user filter or a system filter that ran in a subprocess used any of
7988 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7989 the wrong values were passed to the pipe command ($thisaddress had the
7990 value of $0, $0 had the value of $1, etc). This bug was introduced by
7991 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7993 38. Improved the line breaking for long SMTP error messages from ACLs.
7994 Previously, if there was no break point between 40 and 75 characters, Exim
7995 left the rest of the message alone. Two changes have been made: (a) I've
7996 reduced the minimum length to 35 characters; (b) if it can't find a break
7997 point between 35 and 75 characters, it looks ahead and uses the first one
7998 that it finds. This may give the occasional overlong line, but at least the
7999 remaining text gets split now.
8001 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8002 file descriptors might be low, and that setting 1000 would always raise it.
8003 It turns out that in some environments, the limit is already over 1000 and
8004 that lowering it causes trouble. So now Exim takes care not to decrease it.
8006 40. When delivering a message, the value of $return_path is set to $sender_
8007 address at the start of routing (routers may change the value). By an
8008 oversight, this default was not being set up when an address was tested by
8009 -bt or -bv, which affected the outcome if any router or filter referred to
8012 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8013 log only when "message" or "log_message" are set. However, if one of the
8014 conditions was an address verification, or a call to a nested ACL, the
8015 messages generated by the underlying test were being passed through. This
8016 no longer happens. The underlying message is available in $acl_verify_
8017 message for both "message" and "log_message" expansions, so it can be
8018 passed through if needed.
8020 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8021 new expansion $bh_ to give the encoded byte string without charset
8022 translation. Translation happens only if iconv() is available; HAVE_ICONV
8023 indicates this at build time. HEADERS_CHARSET gives the charset to
8024 translate to; headers_charset can change it in the configuration, and
8025 "headers charset" can change it in an individual filter file.
8027 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8028 that creates RFC 2047 encoded "words" labels them as that charset instead
8029 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8030 expansion operator; (ii) when Exim creates a From: line for a local
8031 message; (iii) when a header line is rewritten to include a "phrase" part.
8033 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8034 buggy, causing it to skip the first lines of messages whose message ID
8035 ended in 'D'. This would not have bitten before Exim release 4.14, because
8036 message IDs were unlikely to end in 'D' before then. The effect was to have
8037 incorrect size information for certain domains.
8039 45. #include "config.h" was missing at the start of the crypt16.c module. This
8040 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8043 46. If there was a timeout during a "random" callout check, Exim treated it as
8044 a failure of the random address, and carried on sending RSET and the real
8045 address. If the delay was just some slowness somewhere, the response to the
8046 original RCPT would be taken as a response to RSET and so on, causing
8047 mayhem of various kinds.
8049 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8050 when I implemented it. It didn't allow for the fact that some option values
8051 may legitimately be negative (e.g. size_addition), and it didn't even do
8052 the right test for positive values.
8054 48. Domain names in DNS records are case-independent. Exim always looks them up
8055 in lower case. Some resolvers return domain names in exactly the case they
8056 appear in the zone file, that is, they may contain uppercase letters. Not
8057 all resolvers do this - some return always lower case. Exim was treating a
8058 change of case by a resolver as a change of domain, similar to a widening
8059 of a domain abbreviation. This triggered its re-routing code and so it was
8060 trying to route what was effectively the same domain again. This normally
8061 caused routing to fail (because the router wouldn't handle the domain
8062 twice). Now Exim checks for this case specially, and just changes the
8063 casing of the domain that it ultimately uses when it transmits the message
8066 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8069 50. If a filter generated a file delivery with a non-absolute name (possible if
8070 no home directory exists for the router), the forbid_file option was not
8073 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8074 the existing equality matching.
8076 52. Exim was using ints instead of ino_t variables in some places where it was
8077 dealing with inode numbers.
8079 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8080 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8081 called TMPDIR, and if it finds it is different, it changes its value.
8083 54. The smtp_printf() function is now made available to local_scan() so
8084 additional output lines can be written before returning. There is also an
8085 smtp_fflush() function to enable the detection of a dropped connection.
8086 The variables smtp_input and smtp_batched_input are exported to
8089 55. Changed the default runtime configuration: the message "Unknown user"
8090 has been removed from the ACL, and instead placed on the localuser router,
8091 using the cannot_route_message feature. This means that any verification
8092 failures that generate their own messages won't get overridden. Similarly,
8093 the "Unrouteable address" message that was in the ACL for unverifiable
8094 relay addresses has also been removed.
8096 56. Added hosts_avoid_esmtp to the smtp transport.
8098 57. The exicyclog script was not checking for the esoteric option
8099 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8100 will work only if exicyclog is run under the appropriate euid.
8102 58. Following a discussion on the list, the rules by which Exim recognises line
8103 endings on incoming messages have been changed. The -dropcr and drop_cr
8104 options are now no-ops, retained only for backwards compatibility. The
8105 following line terminators are recognized: LF CRLF CR. However, special
8106 processing applies to CR:
8108 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8109 nor a local message in the state where . is a terminator.
8111 (ii) If a bare CR is encountered in a header line, an extra space is added
8112 after the line terminator so as not to end the header. The reasoning
8113 behind this is that bare CRs in header lines are most likely either
8114 to be mistakes, or people trying to play silly games.
8116 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8117 was being incorrectly given as 18 bytes larger than it should have been.
8118 This is a VOB (very old bug).
8120 60. This may never have affected anything current, but just in case it has:
8121 When the local host is found other than at the start of a list of hosts,
8122 the local host, those with the same MX, and any that follow, are discarded.
8123 When the list in question was part of a longer list of hosts, the following
8124 hosts (not currently being processed) were also being discarded. This no
8125 longer happens. I'm not sure if this situation could ever has previously
8128 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8131 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8132 header. This is something that is recommended in a new Internet Draft, and
8133 is something that is documented as being done by Sendmail. There are two
8134 possible values. For messages generated by the autoreply transport, Exim
8137 Auto-Submitted: auto-replied
8139 whereas for all other generated messages (e.g. bounces) it adds
8141 Auto-Submitted: auto-generated
8143 63. The "personal" condition in filters now includes a test for the
8144 Auto-Submitted: header. If it contains the string "auto-" the message it
8145 not considered personal.
8147 64. Added rcpt_include_affixes as a generic transport option.
8149 65. Added queue_only_override (default true).
8151 66. Added the syslog_duplication option.
8153 67. If what should have been the first header line of a message consisted of
8154 a space followed by a colon, Exim was mis-interpreting it as a header line.
8155 It isn't of course - it is syntactically invalid and should therefore be
8156 treated as the start of the message body. The misbehaviour could have
8157 caused a number of strange effects, including loss of data in subsequent
8158 header lines, and spool format errors.
8160 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8161 client host had authenticated. This control can now be exercised by an ACL
8162 for more flexibility.
8164 69. By default, callouts do not happen when testing with -bh. There is now a
8165 variant, -bhc, which does actually run the callout code, including
8166 consulting and updating the callout cache.
8168 70. Added support for saslauthd authentication, courtesy of Alexander
8171 71. If statvfs() failed on the spool or log directories while checking their
8172 size for availability, Exim confusingly gave the error "space shortage".
8173 Furthermore, in debugging mode it crashed with a floating point exception.
8174 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8175 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8176 serious problem, Exim now writes to the main and panic logs when this
8177 happens, with details of the failure. It then refuses to accept the
8178 incoming message, giving the message "spool directory problem" or "log
8179 directory problem" with a 421 code for SMTP messages.
8181 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8182 0, 1, and 2 exist, because some OS complain for execs without them (see
8183 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8184 descriptors. However, the code omitted to check that the open succeeded,
8185 causing mysterious errors if for some reason the permissions on /dev/null
8186 got screwed. Now Exim writes a message to the main and panic logs, and
8187 bombs out if it can't open /dev/null.
8189 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8190 interact so that it is all more flexible. It is supposed to remain
8191 backwards compatible. Also added extra_local_interfaces.
8193 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8194 to bomb out with an assertion failure - to the client this appears as a
8195 connection drop. This problem occurs in the part of the code that was taken
8196 from the Samba project. Fortunately, the assertion is in a very simple
8197 function, so I have fixed this by reproducing the function inline in the
8198 one place where it is called, and arranging for authentication to fail
8199 instead of killing the process with assert().
8201 75. The SPA client code was not working when the server requested OEM rather
8202 than Unicode encoding.
8204 76. Added code to make require_files with a specific uid setting more usable in
8205 the case where statting the file as root fails - usually a non-root-mounted
8206 NFS file system. When this happens and the failure is EACCES, Exim now
8207 forks a subprocess and does the per-uid checking as the relevant uid.
8209 77. Added process_log_path.
8211 78. If log_file_path was not explicitly set, a setting of check_log_space or
8212 check_log_inodes was ignored.
8214 79. If a space check for the spool or log partitions fails, the incident is now
8215 logged. Of course, in the latter case the data may get lost...
8217 80. Added the %p formatting code to string_format() so that it can be used to
8218 print addresses in debug_print(). Adjusted all the address printing in the
8219 debugging in store.c to use %p rather than %d.
8221 81. There was a concern that a line of code in smtp_in.c could overflow a
8222 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8223 initially expressed, the concern was not well-founded, because trailing
8224 spaces are removed early. However, if the trailing spaces were followed by
8225 a NULL, they did not get removed, so the overflow was possible. Two fixes
8228 (a) I re-wrote the offending code in a cleaner fashion.
8229 (b) If an incoming SMTP command contains a NULL character, it is rejected
8232 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8233 runs initgroups(), so that if the Exim user is in any additional groups,
8234 they will be used during message reception.
8240 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob