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 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
47 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
48 after reception to before a subsequent reception. This should
49 mean slightly faster delivery, and also confirmation of reception
52 JH/02 Move from using the pcre library to pcre2. The former is no longer
53 being developed or supported (by the original developer).
55 JH/03 Constification work in the filters module required a major version
56 bump for the local-scan API. Specifically, the "headers_charset"
57 global which is visible via the API is now const and may therefore
58 not be modified by local-scan code.
60 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
61 sendfile() didi not account for the way the ClamAV driver code called it.
63 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
64 time check was being done for every character; replace that with one
67 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
68 sent was prefixed with a length byte.
70 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
71 RFC 5321. Previously Dovecot (at least) would log errors during
74 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
75 in 4.61 without a reason logged. Binaries will be bigger, which might
76 matter on diskspace-constrained systems, but debug is easier.
78 JH/09 Fix macro-definition during "-be" expansion testing. The move to
79 write-protected store for macros had not accounted for these runtime
80 additions; fix by removing this protection for "-be" mode.
82 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
83 handing out large-numbered file descriptors, violating the usual Unix
84 assumption (and required by Posix) that the lowest possible number will be
85 allocated by the kernel when a new one is needed. In the daemon, and any
86 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
87 are not useable for FD_SET() [and hence select()] and overwrite the stack.
88 Assorted crashes happen.
90 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
91 main-section options or in a connect ACL, the value from the first ever
92 connection was never replaced for subsequent connections. Found by
95 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
96 though only once PCRE2 was introduced: the memory accounting used under
97 debug offset allocations by an int, giving a hard trap in early startup.
98 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
100 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
101 with underbars is given. The write-protection of configuration introduced
102 in 4.95 trapped when normalisation was applied to an option not needing
105 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
107 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
108 failing to close the certificates directory, every hour or any time it
111 JH/16 Debugging initiated by an ACL control now continues through into routing
112 and transport processes. Previously debugging stopped any time Exim
113 re-execs, or for processing a queued message.
115 JH/17 The "expand" debug selector now gives more detail, specifically on the
116 result of expansion operators and items.
118 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
119 bad comparison between the option value and the name of the file to
120 be included was done, and a mismatch was wrongly identified.
121 4.88 to 4.95 are affected.
123 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
125 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
126 supplied as the db file-pair basename. Previously, if a directory
127 path was given, for example via the autoreply "once" option, the DB
128 file.pag and file.dir files would be created in that directory's
131 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
132 "taint" log_selector. These were previously deprecated.
134 JH/22 Fix static address-list lookups to properly return the matched item.
135 Previously only the domain part was returned.
137 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
138 the call into OpenSSL to send a TLS Close was being repeated; this
139 resulted in the library waiting for the peer's Close. If that was never
140 sent we waited forever. Fix by tracking send calls.
142 JH/24 The ${run} expansion item now expands its command string elements after
143 splitting. Previously it was before; the new ordering makes handling
144 zero-length arguments simpler. The old ordering can be obtained by
145 appending a new option "preexpand", after a comma, to the "run".
147 JH/25 Taint-check exec arguments for transport-initiated external processes.
148 Previously, tainted values could be used. This affects "pipe", "lmtp" and
149 "queryprogram" transport, transport-filter, and ETRN commands.
150 The ${run} expansion is also affected: in "preexpand" mode no part of
151 the command line may be tainted, in default mode the executable name
154 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
155 the the facility was not passed across execs, and only the first message
156 passed over a connection could use BDAT; any further ones using DATA.
158 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
159 uses $sending_ip_address and an interface is specified.
160 Previously any use of the local address in the EHLO name disabled
161 PIPECONNECT, the common case being to use the rDNS of it.
163 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
164 resumption. Previously verify failed because no certificate status is
165 passed on the wire for the restarted session. Fix by using the recorded
166 ocsp status of the stored session for the new connection.
168 JH/29 TLS resumption: the key for session lookup in the client now includes
169 more info that a server could potentially use in configuring a TLS
170 session, avoiding oferring mismatching sessions to such a server.
171 Previously only the server IP was used.
173 JH/30 Fix string_copyn() for limit greater than actual string length.
174 Previously the copied amount was the limit, which could result in a
175 overlapping memcpy for newly allocated destination soon after a
176 source string shorter than the limit. Found/investigated by KM.
178 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
179 close; it may be needed for a subsequent connection. This caused a
180 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
182 JH/32 Fix CHUNKING for a second message on a connection when the first was
183 rejected. Previously we did not reset the chunking-offered state, and
184 erroneously rejected the BDAT command. Investigation help from
187 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
188 an empty address. Previously the expansion returned an error.
190 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
191 proxy. Previously these were misparsed, leading to paniclog entries.
197 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
198 related applications. Previously an "H" was used where available info
199 says that "M" should be, so change to match.
201 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
202 as arguments, so an implementation trying to copy these into a local
203 buffer was taking a taint-enforcement trap. Fix by using dynamically
204 created buffers. Similar fix for radius expansion condition.
206 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
207 reasonable, eg. to count headers. Fix by using dynamically created
208 buffers rather than a local. Do similar fixes for ACL actions "dcc",
209 "log_reject_target", "malware" and "spam"; the arguments are expanded
210 so could be handling tainted values.
212 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
213 broken the (no-op) support for this sendmail command. Restore it
214 to doing nothing, silently, and returning good status.
216 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
217 record path was given (or the default used) without a leading directory
218 path, an error occurred on trying to open it. Use the transport's working
221 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
222 transport. Previously it was the name on the DNS A-record; use instead
223 the head of the CNAME chain leading there (if there is one). This seems
224 to align better with RFC 6125.
226 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
227 smtp_accept_max_per_host allocated resources which were not released
228 when the limit was exceeded. This eventually crashed the daemon. Fix
229 by adding a release action in that path.
231 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
232 expanded; previously using tainted values was rejected. Fix by using
233 dynamically-created buffers.
235 JH/09 Relax restrictions on ACL verify condition needing access to message
236 headers. Previously they were only permitted in data and non-smtp ACLs;
237 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
238 not_blind, header_sender and header_names_ascii verification.
240 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
241 Previously a macro used one argument twice; when called with the
242 argument as an expression having side-effects, incorrect operation
243 resulted. Use an inlineable function.
245 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
246 held open for a verify callout. Previously this wan not accounted for
247 and a corrupt onward SMTP conversation resulted.
249 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
250 Previously they were embedded naively in the constructed address; when
251 needed, strip the quoting and quote the entire local_part.
252 Also make the inbound_srs expansion condition handle quoting.
254 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
255 excluded, not matching the documentation.
257 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
258 was given for the sqlite_dbfile a trap resulted.
260 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
261 "name" argument resulted in a trap. There is no reason to disallow such;
262 this was a coding error.
264 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
265 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
266 spent suspended, ignoring the POSIX definition. Previously we assumed
267 it did and a constant offset from real time could be used as a correction.
268 Change to using the same clock source for the start-of-message and the
269 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
270 exists, just to get a clock slightly more aligned to reality.
272 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
273 RFC says it is optional some validators care. The missing char was not
274 intended but triggered by a line-wrap alignment. Discovery and fix by
275 Guillaume Outters, hacked on by JH.
277 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
278 name being quoted was tainted a trap would be taken. Fix by using
279 dynamically created buffers. The routine could have been called by a
280 rewrite with the "h" flag, by using the "-F" command-line option, or
281 by using a "name=" option on a control=submission ACL modifier.
283 JH/19 SPF: change the Authentication-Results expansion component to give
284 smtp.helo when the sender domain is empty. Previously it gave
287 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
288 not in 127.0.0.0/8 to help in spotting list domains taken over by a
289 domain-parking registrar.
291 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
292 Previously when a whitespace character was specified it was not inserted
293 after removing the newline.
295 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
296 the domain part of the recipient address. This overrides any tls_sni
297 option set, which was previously used.
299 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
302 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
303 is_tainted() had an off-by-one error in the overenthusiastic direction.
304 Find and fix by Gavan. Although NetBSD is not a supported platform for
305 4.94 this bug could affect other platforms.
307 PP/01 Fix default prime selection to be consistent.
308 One path used ike23 still, instead of exim.dev.20160529.3; now both
309 execution flows will use the same DH primes (currently
310 exim.dev.20160529.3).
312 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
313 option in smtp transport, to match the documentation. Previously
314 verification was not being done in some cases where it should have been.
316 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
317 than one server was defined and depending on the platform memory layout
318 details, an internal consistency trap could be hit while walking the list
321 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
322 files. The value is used by the authresults expansion item. Previously
323 if this was used in a router or transport, a crash could result.
325 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
326 have a sufficient incidence of EINTR returns from select that an
327 interaction with other operations done by the main daemon loop exposed
328 a bug in the error-handling. This was benign apart from the log
331 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
332 for consistency with delivered "=>" and failed "**" lines. While we're
333 there, handle PRX and TFO.
335 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
336 applied. This resulted, if any header-line rewrite rules were configured,
337 in a panic-log triggerable by sending a message with a long address in
338 a header. Fix by increasing the arbitrary limit to larger than a single
339 (dewrapped) 5322 header line maximum size.
341 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
342 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
343 RFC 6648 which deprecates X- options in protocols as a general practice.
344 Changeover between the implementations is handled by the mechanisms
347 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
348 a remote which uses callout/hold. Previously the local was queued.
350 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
353 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
354 log line. The asterisk is now consistently appended to the remote IP
355 (and port, if given), and will also be provided on defer and fail log
356 lines. Previously it could be placed on the local IP if that was being
357 logged, and was only provided on delivery lines.
359 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
361 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
362 authenticator client_send option. Previously the next char, after a pair
363 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
364 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
365 documentation. There is still no way to get a leading ^ immediately
366 after a NUL (ie. for the password of a PLAIN method authenticator.
368 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
369 files. For bad sizes read, delete the record and whine to paniclog.
371 JH/38 When logging an AUTH failure, as server, do not include sensitive
372 information. Previously, the credentials would be included if given
373 as part of the AUTH command line and an ACL denied authentication.
375 JH/39 Bug 2691: fix $local_part_data. When the matching list element
376 referred to a file, bad data was returned. This likely also affected
379 JH/40 The gsasl authenticator now supports caching of the salted password
380 generated by the client-side implementation. This required the addition
381 of a new variable: $auth4.
383 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
384 left undeleted; the attempt to re-create it then failed - resulting in
385 the usual "SIGHUP tp have daemon reload configuration" to not work.
386 This affected any platform not supporting "abstract" Unix-domain
387 sockets (i.e. not Linux).
389 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
390 recipients" response to RCPT in a later response, with a 250. The
391 previous coding assumed this would not happen, and under PIPELINING
392 would result in both lost and duplicate recipients for a message.
394 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
395 Previously the weighting was incorrectly applied. Similar fix for socks
396 proxies. Found and fixed by Heiko Schlichting.
398 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
399 not handle sub-lists included using the +namedlist syntax. While
400 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
401 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
404 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
405 Previously the permanent pool was used, so the sore could not be freed.
406 This meant a connection with many messages would use continually-growing
409 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
410 per-pool so as not to waste too much space. Previously a constant size
411 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
412 DOS attacks harder. The cost is wasted memory use in the larger blocks.
414 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
415 same space repeatedly, and a smaller process footprint.
417 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
418 for GnuTLS. Previously the name started "%s" which, while not a bug,
419 looked as if if might be one.
421 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
422 "continued connection") make the $proxy_* variables available. Previously
423 the information was not passed across the exec() call for subsequent
424 transport executions. This also mean that the log lines for the
425 messages can show the proxy information.
427 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
428 receive time. With modern systems the difference is significant.
429 The historical behaviour can be restored by disabling (a new) log_selector
430 "queue_time_exclusive".
432 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
433 content) were specifically tested for. Now, also cover expansions
434 resulting in ACL names and inline ACL content.
436 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
437 making it unusable in complex expressions.
439 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
440 fail with a taint-check on the spoolfile name, and leave the message
443 HS/01 Enforce absolute PID file path name.
445 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
447 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
449 PP/02 Bug 2643: Correct TLS DH constants.
450 A missing NUL termination in our code-generation tool had led to some
451 incorrect Diffie-Hellman constants in the Exim source.
452 Reported by kylon94, code-gen tool fix by Simon Arlott.
454 PP/03 Impose security length checks on various command-line options.
455 Fixes CVE-2020-SPRSS reported by Qualys.
457 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
458 better. Reported by Qualys.
460 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
461 providing a particularly obnoxious sender full name.
464 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
466 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
469 PP/08 Change default for recipients_max from unlimited to 50,000.
471 PP/09 Fix security issue with too many recipients on a message (to remove a
472 known security problem if someone does set recipients_max to unlimited,
473 or if local additions add to the recipient list).
474 Fixes CVE-2020-RCPTL reported by Qualys.
476 PP/10 Fix security issue in SMTP verb option parsing
477 Fixes CVE-2020-EXOPT reported by Qualys.
479 PP/11 Fix security issue in BDAT state confusion.
480 Ensure we reset known-good where we know we need to not be reading BDAT
481 data, as a general case fix, and move the places where we switch to BDAT
482 mode until after various protocol state checks.
483 Fixes CVE-2020-BDATA reported by Qualys.
485 HS/03 Die on "/../" in msglog file names
487 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
488 the Exim runtime user are allowed to create files.
490 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
493 QS/03 When reading the output from interpreted forward files we do not
494 pass the pipe between the parent and the interpreting process to
495 executed child processes (if any).
497 QS/04 Always die if requested from internal logging, even is logging is
500 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
501 the API; compilation noo longer completes with DMARC support included.
502 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
503 been reported on other platforms.
505 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
507 JH/56 Make the majority of info read from config files readonly, for defence-in-
508 depth against exploits. Suggestion by Qualys.
509 Not supported on Solaris 10.
511 JH/57 Fix control=fakreject for a custom message containing tainted data.
512 Previously this resulted in a log complaint, due to a re-expansion present
513 since fakereject was originally introduced.
515 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
516 modifier was given, a loop resulted.
518 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
519 pipelined SMTP command followed the BDAT LAST then it would be
520 incorrectly treated as part of the message body, causing a verification
523 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
524 require looser validation rules than those for 821-level addresses,
525 which only permit IP addresses.
531 JH/01 Avoid costly startup code when not strictly needed. This reduces time
532 for some exim process initialisations. It does mean that the logging
533 of TLS configuration problems is only done for the daemon startup.
535 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
537 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
538 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
541 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
542 DKIM signing is being done. Previously a transport_filter would always
543 disable CHUNKING, falling back to traditional DATA.
545 JH/05 Regard command-line recipients as tainted.
547 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
549 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
550 PAM library frees one of the arguments given to it, despite the
551 documentation. Therefore a plain malloc must be used.
553 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
554 on-stack buffers were used, resulting in a taint trap when DSN information
555 copied from a received message was written into the buffer.
557 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
558 the ordering of its ARC headers. This caused a crash.
560 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
561 a new record was being constructed with information from the peer, a trap
564 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
565 installation would get error messages from DMARC verify, when it hit the
566 nonexistent file indicated by the default. Distros wanting DMARC enabled
567 should both provide the file and set the option.
568 Also enforce no DMARC verification for command-line sourced messages.
570 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
571 could, depending on the platform, hang at the STARTTLS response.
573 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
574 message on a connection. Previously if one message had ARC headers and
575 the following one did not, a crash could result when adding an
576 Authentication-Results: header.
578 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
579 Exim main code and Exim-related utities. The introduction of taint
580 tracking also did many adjustments to string handling. Since then, eximon
581 frequently terminated with an assert failure.
583 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
584 check for 452 responses. This slightly helps the inefficieny of doing
585 a large alias-expansion into a recipient-limited target. The max_rcpt
586 transport option still applies (and at the current default, will override
587 the new feature). The check is done for either cause of synch, and forces
588 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
589 connection. The new facility is not tunable at this time.
591 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
592 library live data was being used, so the results became garbage. Make
593 copies while it is still usable.
595 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
596 on delivery deferred (==) and failed (**) lines (if a delivery was
597 attemtped). Previously it was only on completion (=>) lines.
599 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
600 for the expansion of the server_scram_iter and server_scram_salt options.
602 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
603 are now specifically given a NO_DATA response without hitting the system
604 resolver. The library goes on to do the now-standard TXT lookup.
605 Use of dnsdb lookups is not affected.
607 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
608 only retrieve the errormessage once. Previously two calls to dlerror()
609 were used, and the second one (for mainlog/paniclog) retrieved null
612 JH/20 Taint checking: disallow use of tainted data for
613 - the appendfile transport file and directory options
614 - the pipe transport command
615 - the autoreply transport file, log and once options
616 - file names used by the redirect router (including filter files)
618 - paths used by single-key lookups
619 Previously this was permitted.
621 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
622 adjusted the size of a major service buffer; this failed because the
623 buffer was in use at the time. Change to a compile-time increase in the
624 buffer size, when this authenticator is compiled into exim.
626 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
627 previous fast-mode was untenable in the face of glibs using mmap to
628 support larger malloc requests.
630 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
631 New values supported, if defined on system where compiled:
632 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
633 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
635 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
636 running a limited number of proceses in parallel, a benefit is gained. The
637 amount varies with the platform hardware and load. The use of the option
638 queue_run_in_order means we cannot do this, as ordering becomes
641 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
642 had introduced a string-copy (for ensuring NUL-termination) which was not
643 appropriate for that case, which can include embedded NUL bytes in the
644 block of data. Investigation showed the copy to actually be needless, the
645 data being length-specified.
647 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
648 done during a receiving connection, and both used TLS, global info was
649 used rather than per-connection info for tracking the state of data
650 queued for transmission. This could result in a connection hang.
652 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
653 Previously, when delivering serveral messages down a single connection
654 only the first would provide a SIZE. This was due to the size information
655 not being properly tracked.
657 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
658 TAI (at 37 seconds currently), pretend to be in UTC for time-related
659 expansion and logging. Previously, spurious values such as a future
660 minute could be seen.
662 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
663 it could crash from a null-deref. This could also affect the
664 ${addresses: } operator and ${readsock } item.
666 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
667 message following a mime one, the variable was not reset.
669 JH/30 When an pipelined-connect fails at the first response, assume incorrect
670 cached capability (perhaps the peer reneged?) and immediately retry in
673 JH/31 Fix spurious detection of timeout while writing to transport filter.
675 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
676 an attempt to copy the string was made before checking it.
678 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
679 taint of the lookup key was maintained; we now regard the presence in the
680 filesystem as sufficient validation.
682 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
683 argument is supplied.
685 JH/35 The dsearch lookup now requires that the directory is an absolute path.
686 Previously this was not checked, and nonempty relative paths made an
687 access under Exim's current working directory.
689 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
690 Previously no event was raised.
692 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
693 parameter supplied by the sender MAIL FROM command. Previously it was
694 ignored, and only the check_spool_space option value for the required
697 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
698 the size of the signing public-key. Previously it was instead giving
699 the size of the signature hash.
701 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
702 the default. See the (new) dkim_verify_min_keysizes option.
704 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
705 and an ACL use a lookup for checking either the local_part or domain,
706 stale data could be accessed. Ensure that variable references are
707 dropped between messages.
709 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
710 by the client was not checked as pointing within response data before
711 being used. A malicious client could thus cause an out-of-bounds read and
712 possibly gain authentication. Fix by adding the check.
714 JH/42 Internationalisation: change the default for downconversion in the smtp
715 transport to be "if needed". Previously it was "as previously set" for
716 the message, which usually meant "if needed" for message-submission but
717 "no" for everything else. However, MTAs have been seen using SMTPUTF8
718 even when the envelope addresses did not need it, resulting in forwarding
719 failures to non-supporting MTAs. A downconvert in such cases will be
720 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
721 transport. The change does mean that addresses needing conversion will
722 be converted when previously a delivery failure would occur.
724 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
725 response was received it would be used unchecked in a fail-DSN, violating
726 standards on line-length limits. Truncate if needed.
728 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
729 doesn't work. (Additionally add a new main config option to configure the
736 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
737 side, to decode a TLS 1.3 packet capture.
739 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
740 Previously the default library behaviour applied, sending two, each in
743 JH/03 Debug output for ACL now gives the config file name and line number for
746 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
748 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
750 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
751 buffer overrun for (non-chunking) other transports.
753 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
754 TLS1.3, means that a server rejecting a client certificate is not visible
755 to the client until the first read of encrypted data (typically the
756 response to EHLO). Add detection for that case and treat it as a failed
757 TLS connection attempt, so that the normal retry-in-clear can work (if
758 suitably configured).
760 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
761 and/or domain. Found and fixed by Jason Betts.
763 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
764 configuration). If a CNAME target was not a wellformed name pattern, a
767 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
768 the OS reports them interleaved with other addresses.
770 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
771 used both for input and for a verify callout, both encrypted, SMTP
772 responses being sent by the server could be lost. This resulted in
773 dropped connections and sometimes bounces generated by a peer sending
776 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
777 string. Previously it was possible to cause undefined behaviour in a
778 library routine (usually a crash). Found by "zerons".
780 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
783 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
784 API was removed, so update to use the newer ones.
786 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
787 any timeout set, is taking a long time. Previously we would hang on to a
788 rotated logfile "forever" if the input was arriving with long gaps
789 (a previous attempt to fix addressed lack, for a long time, of initial
792 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
793 shared (NFS) environment. The length of the tempfile name is now
794 4 + 16 ("hdr.$message_exim_id") which might break on file
795 systems which restrict the file name length to lower values.
798 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
799 shared (NFS) environment.
801 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
802 did for all versions <4.90). Notably -M, -m, --invert, -I may be
805 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
806 on some platforms for bit 31.
808 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
809 to changes apparently associated with TLS1.3 handling some of the APIs
810 previously used were either nonfunctional or inappropriate. Strings
811 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
812 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
813 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
814 This affects log line X= elements, the $tls_{in,out}_cipher variables,
815 and the use of specific cipher names in the encrypted= ACL condition.
817 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
819 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
820 verification result was not updated unless hosts_require_ocsp applied.
822 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
823 queue_list_requires_admin set to false, non-admin users were denied the
826 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
827 directory-of-certs mode. Previously they were advertised despite the
830 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
831 A single TCP connection by a client will now hold a TLS connection open
832 for multiple message deliveries, by default. Previously the default was to
835 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
836 default. If built with the facility, DANE will be used. The facility
837 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
839 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
840 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
841 must be defined and you must still, unless you define DISABLE_TLS, manage
842 the the include-dir and library-file requirements that go with that
843 choice. Non-TLS builds are still supported.
845 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
848 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
849 default. If the platform supports and has the facility enabled, it will
850 be requested on all coneections.
852 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
853 controlled by the build-time option SUPPORT_PIPE_CONNECT.
855 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
857 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
858 success-DSN messages. Previously the From: header was always the default
859 one for these; the option was ignored.
861 JH/28 Fix the timeout on smtp response to apply to the whole response.
862 Previously it was reset for every read, so a teergrubing peer sending
863 single bytes within the time limit could extend the connection for a
864 long time. Credit to Qualsys Security Advisory Team for the discovery.
866 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
867 delivery address, which leaked information of the results of local
868 forwarding. Change to the original envelope recipient address, per
871 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
872 requested. Previously not bounce was generated and a log entry of
873 error ignored was made.
875 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
877 JH/32 Introduce a general tainting mechanism for values read from the input
878 channel, and values derived from them. Refuse to expand any tainted
879 values, to catch one form of exploit.
881 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
882 was unused and the unexpanded text used for the test. Found and
883 fixed by Ruben Jenster.
885 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
886 an attempt to use a TLS library read routine dereffed a nul pointer,
889 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
890 them in our smtp response.
892 JH/36 Have the general router option retry_use_local_part default to true when
893 any of the restrictive preconditions are set (to anything). Previously it
894 was only for check_local user. The change removes one item of manual
895 configuration which is required for proper retries when a remote router
896 handles a subset of addresses for a domain.
898 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
899 link count into consideration.
901 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
902 caused the extension of big_buffer, the following lines were ignored.
904 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
905 accordance with RFC 2308. Previously there was no expiry, so a longlived
906 receive process (eg. due to ACL delays) versus a short SOA value could
909 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
911 JH/39 Promote DMARC support to mainline.
913 JH/40 Bug 2452: Add a References: header to DSNs.
915 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
916 parameters. The relevant library call is documented as "Deprecated: This
917 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
918 3.6.0, DH parameters are negotiated following RFC7919."
920 HS/06 Change the default of dnssec_request_domains to "*"
922 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
923 carried on and emitted a BDAT command, even when PIPELINING was not
926 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
927 buffer was used for the filename, resulting in a trap when tainted
928 arguments (eg. $domain) were used.
930 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
931 recommended to avoid a possible server-load attack. The feature can be
932 re-enabled via the openssl_options main cofiguration option.
934 JH/45 local_scan API: documented the current smtp_printf() call. This changed
935 for version 4.90 - adding a "more data" boolean to the arguments.
936 Bumped the ABI version number also, this having been missed previously;
937 release versions 4.90 to 4.92.3 inclusive were effectively broken in
938 respect of usage of smtp_printf() by either local_scan code or libraries
939 accessed via the ${dlfunc } expansion item. Both will need coding
940 adjustment for any calls to smtp_printf() to match the new function
941 signature; a FALSE value for the new argument is always safe.
943 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
944 the file-offset (which the Linux syscall does, and exim expects); this
945 resulted in an indefinite loop.
947 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
948 to do ARC verification. The Authentication-Results: header line added
949 by the configuration then had no ARC item.
951 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
952 an incorrect mode was used for reading the result, resulting in it being
959 JH/01 Remove code calling the customisable local_scan function, unless a new
960 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
962 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
963 non-signal-safe functions being used.
965 JH/03 Bug 2269: When presented with a received message having a stupidly large
966 number of DKIM-Signature headers, disable DKIM verification to avoid
967 a resource-consumption attack. The limit is set at twenty.
969 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
970 report of oldest_pass in ${authres } in consequence, and separate out
971 some descriptions of reasons for verification fail.
973 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
974 files in the spool were present and unlocked. A queue-runner could spot
975 them, resulting in a duplicate delivery. Fix that by doing the unlock
976 after the unlink. Investigation by Tim Stewart. Take the opportunity to
977 add more error-checking on spoolfile handling while that code is being
980 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
981 No known attacks, no CVE, this is defensive hardening.
983 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
984 a queue-runner could start a delivery while other operations were ongoing.
985 Cutthrough delivery was a common victim, resulting in duplicate delivery.
986 Found and investigated by Tim Stewart. Fix by using the open message data
987 file handle rather than opening another, and not locally closing it (which
988 releases a lock) for that case, while creating the temporary .eml format
989 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
991 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
992 $sender_verify_failure/$recipient_verify_failure to "random".
994 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
997 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
998 Previously this would segfault.
1000 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1003 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1004 like zero, since the resolver should be doing this for us, But we need one
1005 as a CNAME but no MX presence gets the CNAME returned; we need to check
1006 that doesn't point to an MX to declare it "no MX returned" rather than
1007 "error, loop". A new main option is added so the older capability of
1008 following some limited number of chain links is maintained.
1010 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1012 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1013 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1014 now-deprecated earlier definitions used only specified the range up to TLS
1015 1.2 (in the older-version library docs).
1017 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1019 JH/15 Rework TLS client-side context management. Stop using a global, and
1020 explicitly pass a context around. This enables future use of TLS for
1021 connections to service-daemons (eg. malware scanning) while a client smtp
1022 connection is using TLS; with cutthrough connections this is quite likely.
1024 JH/16 Fix ARC verification to do AS checks in reverse order.
1026 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1028 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1029 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1030 Previously the "utf8" would be re-prepended for every additional message.
1032 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1033 Previously thery were accepted, resulting in issues when attempting to
1034 forward messages to a non-supporting MTA.
1036 PP/02 Let -n work with printing macros too, not just options.
1038 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1039 one parent address was copied, and bogus data was used at delivery-logging
1040 time. Either a crash (after delivery) or bogus log data could result.
1041 Discovery and analysis by Tim Stewart.
1043 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1044 Previously if the string ended mid-character, we did not insert the
1045 promised '?' replacement.
1047 PP/04 Documentation: current string operators work on bytes, not codepoints.
1049 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1050 should pack well giving a smaller memory footprint so better caching and
1051 therefore performance. Group the declarations where this can't be done so
1052 that the byte-sized flag variables are not interspersed among pointer
1053 variables, giving a better chance of good packing by the compiler.
1055 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1056 non-null, to avoid issues with sites running BATV. Previously reports were
1057 sent with an empty envelope sender so looked like bounces.
1059 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1060 The ignore_error flag wasn't being returned from the filter subprocess so
1061 was not set for later routers. Investigation and fix by Matthias Kurz.
1063 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1064 and a msg:complete for the whole, when a message is manually removed using
1065 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1067 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1068 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1069 hope that is portable enough.
1071 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1072 requiring a known-CA anchor certificate; make it now rely entirely on the
1073 TLSA as an anchor. Checking the name on the leaf cert against the name
1074 on the A-record for the host is still done for TA (but not for EE mode).
1076 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1077 included in delivery lines for non-proxied connections, when compiled with
1078 SUPPORT_SOCKS and running with proxy logging enabled.
1080 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1081 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1082 move the existing event to fire before the normal logging of message
1083 failure so that custom logging is bracketed by normal logging.
1085 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1086 msg:fail:internal event. Developement by Matthias Kurz.
1088 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1089 far too small for todays use of crypto signatures stored there. Go all
1090 the way to the max DNS message size of 64kB, even though this might be
1091 overmuch for IOT constrained device use.
1093 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1094 copy a string over itself. The library routine is documented as not
1095 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1097 JH/32 For main options check_spool_space and check_inode_space, where the
1098 platform supports 64b integers, support more than the previous 2^31 kB
1099 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1100 the previous G, M, k.
1102 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1103 $authenticated_fail_id variable on authentication failure. Previously
1106 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1107 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1108 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1109 GNUTLS_SEC_PARAM_MEDIUM.
1111 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1112 side. Previously we would continue as if no SNI had been received.
1114 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1115 "<" character, which should be a list-separator specification, we walked
1116 off past the nul-terimation.
1118 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1119 causes) even when the retry time is not yet met. Previously they were
1120 not, meaning that when (say) an account was over-quota and temp-rejecting,
1121 and multiple senders' messages were queued, only one sender would get
1122 notified on each configured delay_warning cycle.
1124 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1126 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1127 cipher-suites, an error can be left on the stack even for a succeeding
1128 accept; this results in impossible error messages when a later operation
1131 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1132 return error codes indicating retry. Under TLS1.3 this becomes required.
1134 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1135 it only wrote the new authenticators, resulting in a lack of tracking of
1136 peer changes of ESMTP extensions until the next cache flush.
1138 JH/41 Fix the loop reading a message header line to check for integer overflow,
1139 and more-often against header_maxsize. Previously a crafted message could
1140 induce a crash of the recive process; now the message is cleanly rejected.
1142 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1143 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1149 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1150 When redis_servers is set to a list of > 1 element, and the Redis servers
1151 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1152 case of MOVED into a DEFER case instead, thus moving the query onto the
1153 next server in the list. For a cluster of N elements, all N servers must
1154 be defined in redis_servers.
1156 GF/02 Catch and remove uninitialized value warning in exiqsumm
1157 Check for existence of @ARGV before looking at $ARGV[0]
1159 JH/01 Replace the store_release() internal interface with store_newblock(),
1160 which internalises the check required to safely use the old one, plus
1161 the allocate and data copy operations duplicated in both (!) of the
1162 extant use locations.
1164 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1165 modifier. This matches the restriction on the commandline.
1167 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1168 Previously only the last row was returned.
1170 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1171 we assumed that tags in the header were well-formed, and parsed the
1172 element content after inspecting only the first char of the tag.
1173 Assumptions at that stage could crash the receive process on malformed
1176 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1177 While running the DKIM ACL we operate on the Permanent memory pool so that
1178 variables created with "set" persist to the DATA ACL. Also (at any time)
1179 DNS lookups that fail create cache records using the Permanent pool. But
1180 expansions release any allocations made on the current pool - so a dnsdb
1181 lookup expansion done in the DKIM ACL releases the memory used for the
1182 DNS negative-cache, and bad things result. Solution is to switch to the
1183 Main pool for expansions.
1184 While we're in that code, add checks on the DNS cache during store_reset,
1185 active in the testsuite.
1186 Problem spotted, and debugging aided, by Wolfgang Breyha.
1188 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1189 When none of the hosts presented to a transport match an already-open
1190 connection, close it and proceed with the list. Previously we would
1191 queue the message. Spotted by Lena with Yahoo, probably involving
1194 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1195 Previously a spurious "250 OK id=" response was appended to the proper
1198 JH/08 The "support for" informational output now, which built with Content
1199 Scanning support, has a line for the malware scanner interfaces compiled
1200 in. Interface can be individually included or not at build time.
1202 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1203 by the template makefile "src/EDITME". The "STREAM" support for an older
1204 ClamAV interface method is removed.
1206 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1207 rows affected is given instead).
1209 JH/11 The runtime Berkeley DB library version is now additionally output by
1210 "exim -d -bV". Previously only the compile-time version was shown.
1212 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1213 SMTP connection. Previously, when one had more recipients than the
1214 first, an abortive onward connection was made. Move to full support for
1215 multiple onward connections in sequence, handling cutthrough connection
1216 for all multi-message initiating connections.
1218 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1219 routers. Previously, a multi-recipient message would fail to match the
1220 onward-connection opened for the first recipient, and cause its closure.
1222 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1223 a timeout on read on a GnuTLS initiating connection, resulting in the
1224 initiating connection being dropped. This mattered most when the callout
1225 was marked defer_ok. Fix to keep the two timeout-detection methods
1228 JH/15 Relax results from ACL control request to enable cutthrough, in
1229 unsupported situations, from error to silently (except under debug)
1230 ignoring. This covers use with PRDR, frozen messages, queue-only and
1233 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1235 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1236 metadata, resulting in a crash in free().
1238 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1239 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1240 Broken also in d185889f4, with init system revamp.
1242 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1243 Previously we abruptly closed the connection after reading a malware-
1244 found indication; now we go on to read the "scan ok" response line,
1247 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1248 ACL. Previously, a crash would result.
1250 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1251 macro text after a replacement (previously it was only once per line) and
1252 by skipping builtin macros when searching for an uppercase lead character.
1254 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1255 for the build is renamed.
1257 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1258 was allocated for every new TLS startup, meaning one per message. Fix
1259 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1261 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1262 reported the original. Fix to report (as far as possible) the ACL
1263 result replacing the original.
1265 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1266 OpenSSL. Certificate information is loaded for every new TLS startup,
1267 and the resources needed to be freed.
1269 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1271 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1274 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1275 DATA response info to the (existing) per-recipient response info for
1276 the "C=" log element. It can have useful tracking info from the
1277 destination system. Patch from Simon Arlott.
1279 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1280 length value. Previously this would segfault.
1282 HS/02 Support Avast multiline protoocol, this allows passing flags to
1283 newer versions of the scanner.
1285 JH/28 Ensure that variables possibly set during message acceptance are marked
1286 dead before release of memory in the daemon loop. This stops complaints
1287 about them when the debug_store option is enabled. Discovered specifically
1288 for sender_rate_period, but applies to a whole set of variables.
1289 Do the same for the queue-runner and queue-list loops, for variables set
1290 from spool message files. Do the same for the SMTP per-message loop, for
1291 certain variables indirectly set in ACL operations.
1293 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1294 as a multi-recipient message from a mailinglist manager). The coding had
1295 an arbitrary cutoff number of characters while checking for more input;
1296 enforced by writing a NUL into the buffer. This corrupted long / fast
1297 input. The problem was exposed more widely when more pipelineing of SMTP
1298 responses was introduced, and one Exim system was feeding another.
1299 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1300 receiving system, and refused recipients seen by the sending system
1301 (propating to people being dropped from mailing lists).
1302 Discovered and pinpointed by David Carter.
1304 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1305 replaced by the ${authresults } expansion.
1307 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1309 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1310 allows proper process termination in container environments.
1312 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1313 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1315 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1316 and "err_perm", deprecated since 4.83 when the RFC-defined words
1317 "temperror" and "permerror" were introduced.
1319 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1320 transport-filters or DKIM-signing. The restriction was lost in the
1321 consolidation of verify-callout and delivery SMTP handling.
1322 Extend the restriction to also cover ARC-signing.
1324 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1325 in defer=pass mode supply a 450 to the initiator. Previously the message
1328 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1329 tls_require_ciphers is used as before.
1331 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1332 "pass_unscanned". Only tmpfails from the scanner are written to
1333 the paniclog, as they may require admin intervention (permission
1334 denied, license issues). Other scanner errors (like decompression
1335 bombs) do not cause a paniclog entry.
1337 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1338 Previously it was possible to log spurious information in receive log
1341 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1342 triggered odd behaviour from Outlook Express clients.
1344 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1347 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1348 since the IETF WG has not yet settled on that versus the original
1349 "bare" representation.
1351 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1352 Previously the millisecond value corrupted the output.
1353 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1354 corrupted the output.
1360 JH/01 Rework error string handling in TLS interface so that the caller in
1361 more cases is responsible for logging. This permits library-sourced
1362 string to be attached to addresses during delivery, and collapses
1363 pairs of long lines into single ones.
1365 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1366 during configuration. Wildcards are allowed and expanded.
1368 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1369 This permits better logging.
1371 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1372 do not close it down and have the child transport start it up again on
1373 the passed-on TCP connection. Instead, proxy the child (and any
1374 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1375 affected: the continued delivery log lines do not have any DNSSEC, TLS
1376 Certificate or OCSP information. TLS cipher information is still logged.
1378 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1379 identical IP addresses on different listening ports. Will also affect
1382 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1383 add noisy ifdef guards to special-case this sillyness.
1384 Patch from Bernd Kuhls.
1386 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1387 than 255 are no longer allowed.
1389 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1390 Disable the session-cache too, which might reduce our load. Since we
1391 currrectly use a new context for every connection, both as server and
1392 client, there is no benefit for these.
1393 GnuTLS appears to not support tickets server-side by default (we don't
1394 call gnutls_session_ticket_enable_server()) but client side is enabled
1395 by default on recent versions (3.1.3 +) unless the PFS priority string
1398 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1399 <https://reproducible-builds.org/specs/source-date-epoch/>.
1401 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1402 the check for any unsuccessful recipients did not notice the limit, and
1403 erroneously found still-pending ones.
1405 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1406 MSG_MORE. Only in-clear (not on TLS connections).
1408 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1409 transport-filter is involved we can buffer the headers in memory for
1410 creating the signature, and read the spool data file once for the
1411 signature and again for transmission.
1413 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1414 4.77 as the kernel support then wasn't solid, having issues in 64bit
1415 mode. Now, it's been long enough. Add support for FreeBSD also.
1417 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1418 case where the routing stage had gathered several addresses to send to
1419 a host before calling the transport for the first, we previously failed
1420 to close down TLS in the old transport process before passing the TCP
1421 connection to the new process. The new one sent a STARTTLS command
1422 which naturally failed, giving a failed delivery and bloating the retry
1423 database. Investigation and fix prototype from Wolfgang Breyha.
1425 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1426 false-negatives in the check that the sender had not preempted a response
1427 or prompt from Exim (running as a server), due to that code's lack of
1428 awareness of the SMTP input buffering.
1430 PP/04 Add commandline_checks_require_admin option.
1431 Exim drops privileges sanely, various checks such as -be aren't a
1432 security problem, as long as you trust local users with access to their
1433 own account. When invoked by services which pass untrusted data to
1434 Exim, this might be an issue. Set this option in main configuration
1435 AND make fixes to the calling application, such as using `--` to stop
1438 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1439 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1440 responses to those, into a single TLS record each way (this usually means
1441 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1444 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1445 affects you only if you're dancing at the edge of the param size limits.
1446 If you are, and this message makes sense to you, then: raise the
1447 configured limit or use OpenSSL 1.1. Nothing we can do for older
1450 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1451 cmdline element to get the documented default one. Previously it was
1454 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1455 are unsuable for non-message-specific reasons - eg. connection timeout,
1456 banner-time rejection.
1458 JH/16 Fix logging of delivery remote port, when specified by router, under
1461 PP/06 Repair manualroute's ability to take options in any order, even if one
1462 is the name of a transport.
1465 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1467 JH/17 Change the list-building routines interface to use the expanding-string
1468 triplet model, for better allocation and copying behaviour.
1470 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1471 Previously it was constructed the first time a possibly-matching string
1472 was met in the configuration file input during startup; now it is done
1475 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1476 compatible one, to avoid the (poorly documented) possibility of a config
1477 file in the working directory redirecting the DB files, possibly correpting
1478 some existing file. CVE-2017-10140 assigned for BDB.
1480 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1481 cache-hot. Previously, although the result was properly cached, the
1482 initial verify call returned a defer.
1484 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1485 the main verify for receipient in uncached-mode.
1487 JH/22 Retire historical build files to an "unsupported" subdir. These are
1488 defined as "ones for which we have no current evidence of testing".
1490 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1491 if present. Previously it was ignored.
1493 JH/24 Start using specified-initialisers in C structure init coding. This is
1494 a C99 feature (it's 2017, so now considered safe).
1496 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1497 if was a fixed-sized field and bitmask ops via macros; it is now more
1500 PP/07 GitHub PR 56: Apply MariaDB build fix.
1501 Patch provided by Jaroslav Å karvada.
1503 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1504 during Coverity cleanups [4.87 JH/47]
1505 Diagnosis and fix provided by Michael Fischer v. Mollard.
1507 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1508 the right size to place the terminating semicolon on its own folded
1509 line, the header hash was calculated to an incorrect value thanks to
1510 the (relaxed) space the fold became.
1512 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1513 and confused the parent.
1515 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1516 which could crash as a result. This could lead to undeliverable messages.
1518 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1521 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1522 matters on fast-turnover and PID-randomising systems, which were getting
1523 out-of-order delivery.
1525 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1526 a possibly-overlapping copy. The symptom was that "Remote host closed
1527 connection in response to HELO" was logged instead of the actual 4xx
1530 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1531 Previously only that bufferd was discarded, resulting in SYMTP command
1534 JH/32 DKIM: when a message has multiple signatures matching an identity given
1535 in dkim_verify_signers, run the dkim acl once for each. Previously only
1536 one run was done. Bug 2189.
1538 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1539 "panic the current process" to "deliberately defer". The panic log is
1540 still written with the problem list name; the mail and reject logs now
1541 get a temp-reject line for the message that was being handled, saying
1542 something like "domains check lookup or other defer". The SMTP 451
1543 message is still "Temporary local problem".
1545 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1546 A crafted sequence of BDAT commands could result in in-use memory beeing
1547 freed. CVE-2017-16943.
1549 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1550 from SMTP input. Previously it was always done; now only done for DATA
1551 and not BDAT commands. CVE-2017-16944.
1553 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1554 to the message (such as an overlong header line). Previously this was
1555 not done and we did not exit BDAT mode. Followon from the previous item
1556 though a different problem.
1562 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1563 than -2003 did; needs libidn2 in addition to libidn.
1565 JH/02 The path option on a pipe transport is now expanded before use.
1567 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1568 Patch provided by "Björn", documentation fix added too.
1570 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1571 missing a wire-to-host endian conversion.
1573 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1574 close after a BDAT command line could be taken as a following command,
1575 giving a synch failure. Fix by only checking for synch immediately
1576 before acknowledging the chunk.
1578 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1579 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1580 macro. Patches provided by Josh Soref.
1582 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1583 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1584 by IANA as a service extension. However, John Klensin suggests that we
1587 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1588 the dkim code may be unix-mode line endings rather than smtp wire-format
1589 CRLF, so prepend a CR to any bare LF.
1591 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1592 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1594 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1595 the first were themselves being wrongly included in the feed into dkim
1596 processing; with most chunk sizes in use this resulted in an incorrect
1597 body hash calculated value.
1599 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1600 DKIM signature block, for verification. Although advised against by
1601 standards it is specifically not ruled illegal.
1603 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1605 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1606 missing a body hash (the bh= tag).
1608 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1609 It seems that HAProxy sends the Proxy Protocol information in clear and
1610 only then does a TLS startup, so do the same.
1612 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1613 TCP connections (such as for Spamd) unless the daemon successfully set
1614 Fast Open mode on its listening sockets. This fixes breakage seen on
1615 too-old kernels or those not configured for Fast Open, at the cost of
1616 requiring both directions being enabled for TFO, and TFO never being used
1617 by non-daemon-related Exim processes.
1619 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1620 endings, at least on the first header line. Try to canonify any that get
1621 past that check, despite the cost.
1623 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1624 now limited to an arbitrary five deep, while parsing addresses with the
1625 strip_excess_angle_brackets option enabled.
1627 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1628 instead leave the unprompted TLS handshake in socket buffer for the
1629 TLS library to consume.
1631 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1633 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1635 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1636 region in one operation, for speed, and this leaves assigned pointers
1637 dangling. Add checks run only under the testsuite which checks all
1638 variables at a store-reset and panics on a dangling pointer; add code
1639 explicitly nulling out all the variables discovered. Fixes one known
1640 bug: a transport crash, where a dangling pointer for $sending_ip_address
1641 originally assigned in a verify callout, is re-used.
1643 PP/06 Drop '.' from @INC in various Perl scripts.
1645 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1647 PP/08 Reduce a number of compilation warnings under clang; building with
1648 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1649 should be warning-free.
1651 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1653 HS/01 Fix portability problems introduced by PP/08 for platforms where
1654 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1656 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1657 chunk. This allows us to accept broken chunked messages. We need a more
1658 general solution here.
1660 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1661 already-broken messages in the queue.
1663 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1665 JH/19 Fix reference counting bug in routing-generated-address tracking.
1671 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1672 supports it and a size is available (ie. the sending peer gave us one).
1674 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1675 years of being deprecated). The replacements are the ACLs
1676 acl_smtp_mime and acl_not_smtp_mime.
1678 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1679 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1680 or even an in-clear connection were permitted. Now, if the host lookup
1681 was dnssec and dane was requested then the host is only used if the
1682 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1683 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1684 if one fails this test.
1685 This means that a poorly-configured remote DNS will make it incommunicado;
1686 but it protects against a DNS-interception attack on it.
1688 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1689 non-noisy when a race steals the message being considered.
1691 JH/05 If main configuration option tls_certificate is unset, generate a
1692 self-signed certificate for inbound TLS connections.
1694 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1695 in rewrites and routers.
1697 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1698 and logged a warning sing 4.83; now they are a configuration file error.
1700 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1701 (lacking @domain). Apply the same qualification processing as RCPT.
1703 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1705 JH/10 Support ${sha256:} applied to a string (as well as the previous
1708 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1709 a cutthrough deliver is pending, as we always want to make a connection.
1710 This also avoids re-routing the message when later placing the cutthrough
1711 connection after a verify cache hit.
1712 Do not update it with the verify result either.
1714 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1715 when routing results in more than one destination address.
1717 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1718 signing (which inhibits the cutthrough capability). Previously only
1719 the presence of an option was tested; now an expansion evaluating as
1720 empty is permissible (obviously it should depend only on data available
1721 when the cutthrough connection is made).
1723 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1724 the relevant preceding SMTP command did not note the pipelining mode.
1726 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1727 Previously they were not counted.
1729 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1730 as one having no matching records. Previously we deferred the message
1731 that needed the lookup.
1733 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1734 distinguished as "(=".
1736 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1737 for missing MX records. Previously it only worked for missing A records.
1739 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1741 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1742 after the data-go-ahead and data-ack. Patch from Jason Betts.
1744 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1745 even for a "none" policy. Patch from Tony Meyer.
1747 JH/22 Fix continued use of a connection for further deliveries. If a port was
1748 specified by a router, it must also match for the delivery to be
1751 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1752 When one of the recipients of a message was unsuitable for the connection
1753 (has no matching addresses), we lost track of needing to mark it
1754 deferred. As a result mail would be lost.
1756 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1758 JH/25 Decoding ACL controls is now done using a binary search; the source code
1759 takes up less space and should be simpler to maintain. Merge the ACL
1760 condition decode tables also, with similar effect.
1762 JH/26 Fix problem with one_time used on a redirect router which returned the
1763 parent address unchanged. A retry would see the parent address marked as
1764 delivered, so not attempt the (identical) child. As a result mail would
1767 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1768 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1769 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1772 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1773 Main config options check_{log,spool}_{inodes,space} are now
1774 100 inodes, 10MB unless set otherwise in the configuration.
1776 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1777 Previously it only applied to the main-section connection policy
1780 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1782 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1783 by me. Added RFC7919 DH primes as an alternative.
1785 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1786 are not in the system include path.
1788 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1789 GnuTLS, when a session startup failed (eg because the client disconnected)
1790 Exim did stdio operations after fclose. This was exposed by a recent
1791 change which nulled out the file handle after the fclose.
1793 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1794 signed directly by the cert-signing cert, rather than an intermediate
1795 OCSP-signing cert. This is the model used by LetsEncrypt.
1797 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1799 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1800 an incoming connection.
1802 HS/02 Bug 1802: Do not half-close the connection after sending a request
1805 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1806 fallback to "prime256v1".
1808 JH/34 SECURITY: Use proper copy of DATA command in error message.
1809 Could leak key material. Remotely exploitable. CVE-2016-9963.
1815 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1816 and 3.4.4 - once the server is enabled to respond to an OCSP request
1817 it does even when not requested, resulting in a stapling non-aware
1818 client dropping the TLS connection.
1820 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1821 support variable-length bit vectors. No functional change.
1823 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1824 The I= interface field on outgoing lines is now after the H= remote
1825 host field, same as incoming lines. There is a separate
1826 outgoing_interface log selector which allows you to disable the
1829 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1830 If not running log_selector +smtp_connection the mainlog would be held
1831 open indefinitely after a "too many connections" event, including to a
1832 deleted file after a log rotate. Leave the per net connection logging
1833 leaving it open for efficiency as that will be quickly detected by the
1834 check on the next write.
1836 HS/01 Bug 1671: Fix post transport crash.
1837 Processing the wait-<transport> messages could crash the delivery
1838 process if the message IDs didn't exist for some reason. When
1839 using 'split_spool_directory=yes' the construction of the spool
1840 file name failed already, exposing the same netto behaviour.
1842 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1843 mime_regex ACL conditions.
1845 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1846 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1847 to HELO, local diagnostic string.
1849 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1850 log line to debug. Even when configured with a tls authenticator many
1851 client connections are expected to not authenticate in this way, so
1852 an authenticate fail is not an error.
1854 HS/02 Add the Exim version string to the process info. This way exiwhat
1855 gives some more detail about the running daemon.
1857 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1858 matter for fast-change records such as DNSBLs.
1860 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1861 retry record, even if constant. There may be multiple transports with
1862 different interface settings and the retry behaviour needs to be kept
1865 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1867 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1869 JH/10 Bug 840: fix log_defer_output option of pipe transport
1871 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1872 hosts_randomize. This matters a lot when combined with mua_wrapper.
1874 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1875 ${quote_pgsql:<string>} operator.
1877 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1879 JH/14 Change header file name for freeradius-client. Relevant if compiling
1880 with Radius support; from the Gentoo tree and checked under Fedora.
1882 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1884 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1885 option transport_filter, meaning no filtering.
1887 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1889 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1890 defaults to "*" (all hosts). The variable is now available when not built
1891 with TLS, default unset, mainly to enable keeping the testsuite sane.
1892 If a server certificate is not supplied (via tls_certificate) an error is
1893 logged, and clients will find TLS connections fail on startup. Presumably
1894 they will retry in-clear.
1895 Packagers of Exim are strongly encouraged to create a server certificate
1896 at installation time.
1898 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1899 with the $config_file variable.
1901 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1902 in transport context, after the attempt, and per-recipient. The latter type
1903 is per host attempted. The event data is the error message, and the errno
1904 information encodes the lookup type (A vs. MX) used for the (first) host,
1905 and the trailing two digits of the smtp 4xx response.
1907 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1908 to write to mainlog (or rejectlog, paniclog) in the window between file
1909 creation and permissions/ownership being changed. Particularly affects
1910 installations where exicyclog is run as root, rather than exim user;
1911 result is that the running daemon panics and dies.
1913 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1915 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1916 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1917 "pri" and "weight". Note that the previous implicit priority given by the
1918 list order is no longer honoured.
1920 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1921 for DKIM processing.
1923 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1924 by defining SUPPORT_SOCKS.
1926 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1927 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1928 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1929 variables are renamed to proxy_{local,external}_{address,port}.
1931 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1932 for a build by defining SUPPORT_I18N
1934 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1935 of the query string, and make ${quote_redis:} do that quoting.
1937 JH/29 Move Events support from Experimental to mainline, enabled by default
1938 and removable for a build by defining DISABLE_EVENT.
1940 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1942 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1943 cached by the daemon.
1945 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1946 by defining LOOKUP_REDIS. The libhiredis library is required.
1948 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1949 keys are given for lookup.
1951 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1952 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1953 only supported when built with TLS support. The PolarSSL SHA routines
1954 are still used when the TLS library is too old for convenient support.
1956 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1957 openssl_options), for security. OpenSSL forces this from version 1.1.0
1958 server-side so match that on older versions.
1960 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1961 allocation for $value could be released as the expansion processing
1962 concluded, but leaving the global pointer active for it.
1964 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1965 and to use the domains and local_parts ACL conditions.
1967 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1968 incorrectly not doubled on cutthrough transmission, hence seen as a
1969 body-termination at the receiving system - resulting in truncated mails.
1970 Commonly the sender saw a TCP-level error, and retransmitted the message
1971 via the normal store-and-forward channel. This could result in duplicates
1972 received - but deduplicating mailstores were liable to retain only the
1973 initial truncated version.
1975 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1977 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1979 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1980 we're in there, support oversigning also; bug 1309.
1982 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1984 HS/04 Add support for keep_environment and add_environment options.
1986 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1987 either intentional arithmetic overflow during PRNG, or testing config-
1990 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1991 delivery resulted in actual delivery. Cancel cutthrough before DATA
1994 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1995 rejects a recipient: pass the reject to the originator.
1997 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1998 Many were false-positives and ignorable, but it's worth fixing the
2001 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2002 for the new environment-manipulation done at startup. Move the routines
2003 from being local to tls.c to being global via the os.c file.
2005 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2006 an extract embedded as result-arg for a map, the first arg for extract
2007 is unavailable so we cannot tell if this is a numbered or keyed
2008 extraction. Accept either.
2014 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2017 JH/02 The smtp transport option "multi_domain" is now expanded.
2019 JH/03 The smtp transport now requests PRDR by default, if the server offers
2022 JH/04 Certificate name checking on server certificates, when exim is a client,
2023 is now done by default. The transport option tls_verify_cert_hostnames
2024 can be used to disable this per-host. The build option
2025 EXPERIMENTAL_CERTNAMES is withdrawn.
2027 JH/05 The value of the tls_verify_certificates smtp transport and main options
2028 default to the word "system" to access the system default CA bundle.
2029 For GnuTLS, only version 3.0.20 or later.
2031 JH/06 Verification of the server certificate for a TLS connection is now tried
2032 (but not required) by default. The verification status is now logged by
2033 default, for both outbound TLS and client-certificate supplying inbound
2036 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2039 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2040 Status Notification (bounce) messages are now MIME format per RFC 3464.
2041 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2042 under the control of the dsn_advertise_hosts option, and routers may
2043 have a dsn_lasthop option.
2045 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2046 default, modifiable by a malware= option. The list separator for
2047 the options can now be changed in the usual way. Bug 68.
2049 JH/10 The smtp_receive_timeout main option is now expanded before use.
2051 JH/11 The incoming_interface log option now also enables logging of the
2052 local interface on delivery outgoing connections.
2054 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2055 if the interface and destination host and port all match.
2057 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2060 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2061 Patch from Andrew Lewis.
2063 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2064 now supports optional time-restrictions, weighting, and priority
2065 modifiers per server. Patch originally by <rommer@active.by>.
2067 JH/16 The spamd_address main option now supports a mixed list of local
2068 and remote servers. Remote servers can be IPv6 addresses, and
2069 specify a port-range.
2071 JH/17 Bug 68: The spamd_address main option now supports an optional
2072 timeout value per server.
2074 JH/18 Bug 1581: Router and transport options headers_add/remove can
2075 now have the list separator specified.
2077 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2080 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2083 JH/21 Support for the A6 type of dns record is withdrawn.
2085 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2086 rather than the verbs used.
2088 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2089 from 255 to 1024 chars.
2091 JH/24 Verification callouts now attempt to use TLS by default.
2093 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2094 are generic router options now. The defaults didn't change.
2096 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2097 Original patch from Alexander Shikoff, worked over by JH.
2099 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2100 files if ZCAT_COMMAND is not executable.
2102 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2104 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2106 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2107 Normally benign, it bites when the pair was led to by a CNAME;
2108 modern usage is to not canonicalize the domain to a CNAME target
2109 (and we were inconsistent anyway for A-only vs AAAA+A).
2111 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2113 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2114 when evaluating $sender_host_dnssec.
2116 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2117 $sender_helo_dnssec variable.
2119 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2121 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2123 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2125 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2126 documented as working, but never had. Support all but $spam_report.
2128 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2129 added for tls authenticator.
2131 HS/03 Add perl_taintmode main config option
2137 TL/01 When running the test suite, the README says that variables such as
2138 no_msglog_check are global and can be placed anywhere in a specific
2139 test's script, however it was observed that placement needed to be near
2140 the beginning for it to behave that way. Changed the runtest perl
2141 script to read through the entire script once to detect and set these
2142 variables, reset to the beginning of the script, and then run through
2143 the script parsing/test process like normal.
2145 TL/02 The BSD's have an arc4random API. One of the functions to induce
2146 adding randomness was arc4random_stir(), but it has been removed in
2147 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2148 function when detected.
2150 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2151 cause callback expansion.
2153 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2154 syntax errors in an expansion can be treated as a string instead of
2155 logging or causing an error, due to the internal use of bool_lax
2156 instead of bool when processing it.
2158 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2159 server certificates when making smtp deliveries.
2161 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2163 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2165 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2167 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2168 Merged patch from Sebastian Wiedenroth.
2170 JH/05 Fix results-pipe from transport process. Several recipients, combined
2171 with certificate use, exposed issues where response data items split
2172 over buffer boundaries were not parsed properly. This eventually
2173 resulted in duplicates being sent. This issue only became common enough
2174 to notice due to the introduction of connection certificate information,
2175 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2177 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2178 size buffer was used, resulting in syntax errors when an expansion
2181 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2182 version 3.3.6 or later.
2184 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2185 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2186 both become "event_action", the variables become $event_name, $event_data
2187 and $event_defer_errno. There is a new variable $verify_mode, usable in
2188 routers, transports and related events. The tls:cert event is now also
2189 raised for inbound connections, if the main configuration event_action
2192 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2193 early OCSP support, but no stapling (appears to be less than 1.0.0).
2195 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2196 server certificate names available under the smtp transport option
2197 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2200 JH/10 Time-related extraction expansions from certificates now use the main
2201 option "timezone" setting for output formatting, and are consistent
2202 between OpenSSL and GnuTLS compilations. Bug 1541.
2204 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2205 encoded parameter in the incoming message. Bug 1558.
2207 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2208 include certificate info, eximon was claiming there were spoolfile
2211 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2213 JH/14 Log delivery-related information more consistently, using the sequence
2214 "H=<name> [<ip>]" wherever possible.
2216 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2217 are problematic for Debian distribution, omit them from the release
2220 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2222 JH/16 Fix string representation of time values on 64bit time_t architectures.
2225 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2226 output list separator was used.
2231 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2232 checkers that were complaining about end of non-void function with no
2235 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2236 This was a regression introduced in 4.83 by another bugfix.
2238 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2240 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2241 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2247 TF/01 Correctly close the server side of TLS when forking for delivery.
2249 When a message was received over SMTP with TLS, Exim failed to clear up
2250 the incoming connection properly after forking off the child process to
2251 deliver the message. In some situations the subsequent outgoing
2252 delivery connection happened to have the same fd number as the incoming
2253 connection previously had. Exim would try to use TLS and fail, logging
2254 a "Bad file descriptor" error.
2256 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2257 utilities have not been installed.
2259 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2260 temporary space as the ACL may create new global variables.
2262 TL/01 LDAP support uses per connection or global context settings, depending
2263 upon the detected version of the libraries at build time.
2265 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2266 to extract and use the src ip:port in logging and expansions as if it
2267 were a direct connection from the outside internet. PPv2 support was
2268 updated based on HAProxy spec change in May 2014.
2270 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2272 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2273 Properly escape header and check for NULL return.
2275 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2278 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2280 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2281 characters in header names, implemented as a verify condition.
2282 Contributed by Michael Fischer v. Mollard.
2284 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2285 results permerror and temperror. Previous values are deprecated but
2286 still accepted. In a future release, err_perm and err_temp will be
2287 completely removed, which will be a backward incompatibility if the
2288 ACL tests for either of these two old results. Patch contributed by
2289 user bes-internal on the mailing list.
2291 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2293 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2294 selectors, in both main and reject logs.
2296 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2299 JH/07 Add malware type "sock" for talking to simple daemon.
2301 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2303 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2304 routers/transports under cutthrough routing.
2306 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2307 numbers. Touch up "bool" conditional to keep the same definition.
2309 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2311 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2313 JH/12 Expand items in router/transport headers_add or headers_remove lists
2314 individually rather than the list as a whole. Bug 1452.
2316 Required for reasonable handling of multiple headers_ options when
2317 they may be empty; requires that headers_remove items with embedded
2318 colons must have them doubled (or the list-separator changed).
2320 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2321 view the policy declared in the DMARC record. Currently, $dmarc_status
2322 is a combined value of both the record presence and the result of the
2325 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2327 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2328 dnslookup router and the smtp transport (applying to the forward
2331 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2332 of ldap servers used for a specific lookup. Patch provided by Heiko
2335 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2336 New variable $lookup_dnssec_authenticated for observability.
2338 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2339 Patch submitted by Lars Timman.
2341 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2343 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2344 Requires trusted mode and valid format message id, aborts otherwise.
2345 Patch contributed by Heiko Schlichting.
2347 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2348 certextract with support for various fields. Bug 1358.
2350 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2351 is requested by default, modifiable by smtp transport option
2354 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2355 operate on certificate variables to give certificate fingerprints
2356 Also new ${sha256:cert_variable}.
2358 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2360 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2361 Christian Aistleitner.
2363 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2365 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2366 file. Patch from Wolfgang Breyha.
2368 JH/25 Expand the coverage of the delivery $host and $host_address to
2369 client authenticators run in verify callout. Bug 1476.
2371 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2372 align with daemon_smtp_ports. Bug 72.
2374 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2375 support and error reporting did not work properly.
2377 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2378 and is readable. Patch from Andrew Colin Kissa.
2380 TL/14 Enhance documentation of ${run expansion and how it parses the
2381 commandline after expansion, particularly in the case when an
2382 unquoted variable expansion results in an empty value.
2384 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2386 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2387 by renaming away from it. Addresses GH issue 15, reported by
2390 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2392 TL/15 SECURITY: prevent double expansion in math comparison functions
2393 (can expand unsanitized data). Not remotely exploitable.
2400 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2402 PP/02 Make -n do something, by making it not do something.
2403 When combined with -bP, the name of an option is not output.
2405 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2408 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2409 $sender_host_name and config options to manage this, and basic check
2412 PP/05 DSCP support for outbound connections and control modifier for inbound.
2414 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2415 (Only plugin which currently uses this is kerberos4, which nobody should
2416 be using, but we should make it available and other future plugins might
2417 conceivably use it, even though it would break NAT; stuff *should* be
2418 using channel bindings instead).
2420 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2421 name; added for Sendmail compatibility; requires admin caller.
2422 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2423 just ignore it); requires trusted caller.
2424 Also parse but ignore: -Ac -Am -X<logfile>
2427 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2429 TL/02 Add +smtp_confirmation as a default logging option.
2431 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2432 Patch by Magnus Holmgren from 2007-02-20.
2434 TL/04 Bugzilla 1281 - Spec typo.
2435 Bugzilla 1283 - Spec typo.
2436 Bugzilla 1290 - Spec grammar fixes.
2438 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2440 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2442 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2443 mailing list by by Dmitry Isaikin.
2445 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2447 JH/02 Support "G" suffix to numbers in ${if comparisons.
2449 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2451 NM/01 Bugzilla 1197 - Spec typo
2452 Bugzilla 1196 - Spec examples corrections
2454 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2456 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2457 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2460 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2461 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2463 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2464 "acl {{name}{arg}...}", and optional args on acl condition
2467 JH/05 Permit multiple router/transport headers_add/remove lines.
2469 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2471 JH/07 Avoid using a waiting database for a single-message-only transport.
2472 Performance patch from Paul Fisher. Bugzilla 1262.
2474 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2477 JH/09 Add $headers_added variable, with content from use of ACL modifier
2478 add_header (but not yet added to the message). Bugzilla 199.
2480 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2481 Pulled from Bugzilla 817 by Wolfgang Breyha.
2483 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2485 (nb: this is the same fix as in Exim 4.80.1)
2487 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2490 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2493 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2495 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2496 advertise SMTP AUTH mechanism to us, instead of a generic
2497 protocol violation error. Also, make Exim more robust to bad
2498 data from the Dovecot auth socket.
2500 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2502 When a queue runner is handling a message, Exim first routes the
2503 recipient addresses, during which it prunes them based on the retry
2504 hints database. After that it attempts to deliver the message to
2505 any remaining recipients. It then updates the hints database using
2508 So if a recipient address works intermittently, it can get repeatedly
2509 deferred at routing time. The retry hints record remains fresh so the
2510 address never reaches the final cutoff time.
2512 This is a fairly common occurrence when a user is bumping up against
2513 their storage quota. Exim had some logic in its local delivery code
2514 to deal with this. However it did not apply to per-recipient defers
2515 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2517 This change adds a proper retry rule check during routing so that the
2518 final cutoff time is checked against the message's age. We only do
2519 this check if there is an address retry record and there is not a
2520 domain retry record; this implies that previous attempts to handle
2521 the address had the retry_use_local_parts option turned on. We use
2522 this as an approximation for the destination being like a local
2523 delivery, as in LMTP.
2525 I suspect this new check makes the old local delivery cutoff check
2526 redundant, but I have not verified this so I left the code in place.
2528 TF/02 Correct gecos expansion when From: is a prefix of the username.
2530 Test 0254 submits a message to Exim with the header
2534 When I ran the test suite under the user fanf2, Exim expanded
2535 the header to contain my full name, whereas it should have added
2536 a Resent-Sender: header. It erroneously treats any prefix of the
2537 username as equal to the username.
2539 This change corrects that bug.
2541 GF/01 DCC debug and logging tidyup
2542 Error conditions log to paniclog rather than rejectlog.
2543 Debug lines prefixed by "DCC: " to remove any ambiguity.
2545 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2547 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2548 Bug spotted by Jeremy Harris; was flawed since initial commit.
2549 Would have resulted in OCSP responses post-SNI triggering an Exim
2550 NULL dereference and crash.
2552 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2554 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2555 Bug detection, analysis and fix by Samuel Thibault.
2556 Bugzilla 1331, Debian bug #698092.
2558 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2560 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2561 Server implementation by Todd Lyons, client by JH.
2562 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2563 config variable "prdr_enable" controls whether the server
2564 advertises the facility. If the client requests PRDR a new
2565 acl_data_smtp_prdr ACL is called once for each recipient, after
2566 the body content is received and before the acl_smtp_data ACL.
2567 The client is controlled by both of: a hosts_try_prdr option
2568 on the smtp transport, and the server advertisement.
2569 Default client logging of deliveries and rejections involving
2570 PRDR are flagged with the string "PRDR".
2572 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2573 fclose(). Diagnosis by Todd Lyons.
2575 PP/17 Update configure.default to handle IPv6 localhost better.
2576 Patch by Alain Williams (plus minor tweaks).
2579 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2580 This is now consistent with GnuTLS, and is now documented: the
2581 previous undocumented portable approach to treating the option as
2582 unset was to force an expansion failure. That still works, and
2583 an empty string is now equivalent.
2585 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2586 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2587 not performing validation itself.
2589 PP/20 Added force_command boolean option to pipe transport.
2590 Patch from Nick Koston, of cPanel Inc.
2592 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2595 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2597 PP/21 Fix eximon continuous updating with timestamped log-files.
2598 Broken in a format-string cleanup in 4.80, missed when I repaired the
2599 other false fix of the same issue.
2600 Report and fix from Heiko Schlichting.
2603 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2604 Report from Prashanth Katuri.
2606 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2607 It's SecureTransport, so affects any MacOS clients which use the
2608 system-integrated TLS libraries, including email clients.
2610 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2611 using a MIME ACL for non-SMTP local injection.
2612 Report and assistance in diagnosis by Warren Baker.
2614 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2616 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2618 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2619 last id that failed so it may be referenced in subsequent ACL's.
2621 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2624 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2625 ldap library initialization, allowing self-signed CA's to be
2626 used. Also properly sets require_cert option later in code by
2627 using NULL (global ldap config) instead of ldap handle (per
2628 session). Bug diagnosis and testing by alxgomz.
2630 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2631 the src/util/ subdirectory.
2633 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2634 renamed to Transport Post Delivery Action by Jeremy Harris, as
2637 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2638 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2639 redis_servers = needs to be configured which will be used by the redis
2640 lookup. Patch from Warren Baker, of The Packet Hub.
2642 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2644 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2645 hostname or reverse DNS when processing a host list. Used suggestions
2646 from multiple comments on this bug.
2648 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2650 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2651 Missed a few lines, added it to make the runtest require no keyboard
2654 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2655 contains upper case chars. Make router use caseful_local_part.
2657 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2658 support when GnuTLS has been built with p11-kit.
2664 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2666 This, or similar/improved, will also be change PP/11 of 4.82.
2672 PP/01 Handle short writes when writing local log-files.
2673 In practice, only affects FreeBSD (8 onwards).
2674 Bugzilla 1053, with thanks to Dmitry Isaikin.
2676 NM/01 Bugzilla 949 - Documentation tweak
2678 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2681 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2683 PP/02 Implemented gsasl authenticator.
2685 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2687 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2688 `pkg-config foo` for cflags/libs.
2690 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2691 with rest of GSASL and with heimdal_gssapi.
2693 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2694 `pkg-config foo` for cflags/libs for the TLS implementation.
2696 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2697 properties get this fed in as external SSF. A number of robustness
2698 and debugging improvements to the cyrus_sasl authenticator.
2700 PP/08 cyrus_sasl server now expands the server_realm option.
2702 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2703 Patch by Jeremy Harris.
2705 PP/10 Added dbmjz lookup type.
2707 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2709 PP/12 MAIL args handles TAB as well as SP, for better interop with
2710 non-compliant senders.
2711 Analysis and variant patch by Todd Lyons.
2713 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2714 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2715 Patch from Dirk Mueller <dmueller@suse.com>
2717 PP/13 tls_peerdn now print-escaped for spool files.
2718 Observed some $tls_peerdn in wild which contained \n, which resulted
2719 in spool file corruption.
2721 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2722 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2723 or write after TLS renegotiation, which otherwise led to messages
2726 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2727 as a tracking header (ie: a signed header comes before the signature).
2728 Patch from Wolfgang Breyha.
2730 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2731 comma-sep list; embedded commas doubled.
2733 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2735 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2737 Report and patch from Dmitry Banschikov.
2739 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2740 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2741 failures. We appear to now support TLS1.1+ with Exim.
2743 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2744 lets Exim select keys and certificates based upon TLS SNI from client.
2745 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2746 before an outbound SMTP session. New log_selector, +tls_sni.
2748 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2749 NULL dereference. Report and patch from Alun Jones.
2751 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2752 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2753 Not seeing resolver debug output on NetBSD, but suspect this is a
2754 resolver implementation change.
2756 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2757 Left warnings. Added "eximon gdb" invocation mode.
2759 PP/21 Defaulting "accept_8bitmime" to true, not false.
2761 PP/22 Added -bw for inetd wait mode support.
2763 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2764 locate the relevant includes and libraries. Made this the default.
2766 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2767 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2769 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2770 This may cause build issues on older platforms.
2772 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2773 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2774 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2775 Added SNI support via GnuTLS too.
2776 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2778 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2780 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2781 Applied second patch from Janne, implementing suggestion to default
2782 multiple-strings-in-record handling to match SPF spec.
2784 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2786 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2787 read-only, out of scope).
2788 Patch from Wolfgang Breyha, report from Stuart Northfield.
2790 PP/29 Fix three issues highlighted by clang analyser static analysis.
2791 Only crash-plausible issue would require the Cambridge-specific
2792 iplookup router and a misconfiguration.
2793 Report from Marcin Mirosław.
2795 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2797 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2798 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2799 As part of this, removing so much warning spew let me fix some minor
2800 real issues in debug logging.
2802 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2803 assignment on my part. Fixed.
2805 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2806 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2807 Janne Snabb (who went above and beyond: thank you).
2809 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2810 string otherwise requires a connection and a bunch more work and it's
2811 relatively easy to get wrong. Should also expose TLS library linkage
2814 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2815 64-bit ${eval} (JH/03).
2817 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2818 GNU libc to support some of the 64-bit stuff, should not lead to
2819 conflicts. Defined before os.h is pulled in, so if a given platform
2820 needs to override this, it can.
2822 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2823 protection layer was required, which is not implemented.
2824 Bugzilla 1254, patch from Wolfgang Breyha.
2826 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2827 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2828 tls_dhparam take prime identifiers. Also unbreak combination of
2829 OpenSSL+DH_params+TLSSNI.
2831 PP/39 Disable SSLv2 by default in OpenSSL support.
2837 PP/01 Solaris build fix for Oracle's LDAP libraries.
2838 Bugzilla 1109, patch from Stephen Usher.
2840 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2842 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2845 TF/02 Fix a couple more cases where we did not log the error message
2846 when unlink() failed. See also change 4.74-TF/03.
2848 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2849 lock up or crash if it happened to be inside a call to libc when it
2850 got a SIGUSR1 from exiwhat.
2852 The SIGUSR1 handler appends the current process status to the process
2853 log which is later printed by exiwhat. It used to use the general
2854 purpose logging code to do this, but several functions it calls are
2855 not safe for signals.
2857 The new output code in the SIGUSR1 handler is specific to the process
2858 log, and simple enough that it's easy to inspect for signal safety.
2859 Removing some special cases also simplifies the general logging code.
2860 Removing the spurious timestamps from the process log simplifies
2863 TF/04 Improved ratelimit ACL condition.
2865 The /noupdate option has been deprecated in favour of /readonly which
2866 has clearer semantics. The /leaky, /strict, and /readonly update modes
2867 are mutually exclusive. The update mode is no longer included in the
2868 database key; it just determines when the database is updated. (This
2869 means that when you upgrade Exim will forget old rate measurements.)
2871 Exim now checks that the per_* options are used with an update mode that
2872 makes sense for the current ACL. For example, when Exim is processing a
2873 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2874 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2875 must specify per_mail/readonly. If you omit the update mode it defaults to
2876 /leaky where that makes sense (as before) or /readonly where required.
2878 The /noupdate option is now undocumented but still supported for
2879 backwards compatibility. It is equivalent to /readonly except that in
2880 ACLs where /readonly is required you may specify /leaky/noupdate or
2881 /strict/noupdate which are treated the same as /readonly.
2883 A useful new feature is the /count= option. This is a generalization
2884 of the per_byte option, so that you can measure the throughput of other
2885 aggregate values. For example, the per_byte option is now equivalent
2886 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2888 The per_rcpt option has been generalized using the /count= mechanism
2889 (though it's more complicated than the per_byte equivalence). When it is
2890 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2891 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2892 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2893 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2894 Note that using per_rcpt with a non-readonly update mode in more than
2895 one ACL will cause the recipients to be double-counted. (The per_mail
2896 and per_byte options don't have this problem.)
2898 The handling of very low rates has changed slightly. If the computed rate
2899 is less than the event's count (usually one) then this event is the first
2900 after a long gap. In this case the rate is set to the same as this event's
2901 count, so that the first message of a spam run is counted properly.
2903 The major new feature is a mechanism for counting the rate of unique
2904 events. The new per_addr option counts the number of different
2905 recipients that someone has sent messages to in the last time period. It
2906 behaves like per_rcpt if all the recipient addresses are different, but
2907 duplicate recipient addresses do not increase the measured rate. Like
2908 the /count= option this is a general mechanism, so the per_addr option
2909 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2910 example, measure the rate that a client uses different sender addresses
2911 with the options per_mail/unique=$sender_address. There are further
2912 details in the main documentation.
2914 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2916 TF/06 Removed a few PCRE remnants.
2918 TF/07 Automatically extract Exim's version number from tags in the git
2919 repository when doing development or release builds.
2921 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2922 Bugzilla 879. Patch from Paul Fisher.
2924 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2925 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2928 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2930 PP/05 Variable $av_failed, true if the AV scanner deferred.
2931 Bugzilla 1078. Patch from John Horne.
2933 PP/06 Stop make process more reliably on build failure.
2934 Bugzilla 1087. Patch from Heiko Schlittermann.
2936 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2937 Bugzilla 1089. Patch from Heiko Schlittermann.
2939 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2940 Bugzilla 1131. Patch from Holger Weiß.
2942 PP/09 Handle IPv6 addresses with SPF.
2943 Bugzilla 860. Patch from Wolfgang Breyha.
2945 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2947 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2950 PP/11 match_* no longer expand right-hand-side by default.
2951 New compile-time build option, EXPAND_LISTMATCH_RHS.
2952 New expansion conditions, "inlist", "inlisti".
2954 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2956 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2958 PP/14 fix log_write() format string regression from TF/03.
2959 Bugzilla 1152. Patch from Dmitry Isaikin.
2965 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2967 PP/02 Harmonised TLS library version reporting; only show if debugging.
2968 Layout now matches that introduced for other libraries in 4.74 PP/03.
2970 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2972 PP/04 New "dns_use_edns0" global option.
2974 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2977 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2978 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2980 TK/01 Updated PolarSSL code to 0.14.2.
2981 Bugzilla 1097. Patch from Andreas Metzler.
2983 PP/07 Catch divide-by-zero in ${eval:...}.
2984 Fixes bugzilla 1102.
2986 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2989 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2990 format-string attack -- SECURITY: remote arbitrary code execution.
2992 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2993 time unintentionally subject to list matching rules, letting the header
2994 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2995 arbitrary string expansion). This allowed for information disclosure.
2997 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2998 INT_MIN/-1 -- value coerced to INT_MAX.
3004 NM/01 Workaround for PCRE version dependency in version reporting
3007 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3008 This fixes portability to compilers other than gcc, notably
3009 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3011 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3012 makefiles for portability to HP-UX and POSIX correctness.
3014 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3015 Also via indented variable definition in the Makefile.
3016 (Debugging by Oliver Heesakkers).
3018 PP/02 Restore caching of spamd results with expanded spamd_address.
3019 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3021 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3022 Improves build reliability. Fix from: Frank Elsner
3024 NM/02 Fix wide character breakage in the rfc2047 coding
3025 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3027 NM/03 Allow underscore in dnslist lookups
3028 Fixes bug 1026. Patch from Graeme Fowler
3030 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3031 Code patches from Adam Ciarcinski of NetBSD.
3033 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3036 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3037 is logged, to avoid truncation. Patch from John Horne.
3039 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3040 Patch from Jakob Hirsch.
3042 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3043 SQL string expansion failure details.
3044 Patch from Andrey Oktyabrski.
3046 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3047 Patch from Simon Arlott.
3049 PP/09 New lookups functionality failed to compile on old gcc which rejects
3050 extern declarations in function scope.
3051 Patch from Oliver Fleischmann
3053 PP/10 Use sig_atomic_t for flags set from signal handlers.
3054 Check getgroups() return and improve debugging.
3055 Fixed developed for diagnosis in bug 927 (which turned out to be
3058 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3059 Patch from Mark Zealey.
3061 PP/12 Bugzilla 1056: Improved spamd server selection.
3062 Patch from Mark Zealey.
3064 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3065 Based on patch from Heiko Schlittermann.
3067 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3068 Patch from Uwe Doering, sign-off by Michael Haardt.
3070 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3071 variable declaration deep within a block. Bug and patch from
3074 PP/15 lookups-Makefile IRIX compatibility coercion.
3076 PP/16 Make DISABLE_DKIM build knob functional.
3078 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3079 Patch by Simon Arlott
3081 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3082 variable argument macros. Our copy now differs from upstream.
3088 TF/01 Failure to get a lock on a hints database can have serious
3089 consequences so log it to the panic log.
3091 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3092 controlled using the smtp_confirmation log selector.
3094 TF/03 Include the error message when we fail to unlink a spool file.
3096 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3097 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3098 for maintaining out-of-tree patches for some time.
3100 PP/01 Bugzilla 139: Documentation and portability issues.
3101 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3102 Handle per-OS dynamic-module compilation flags.
3104 PP/02 Let /dev/null have normal permissions.
3105 The 4.73 fixes were a little too stringent and complained about the
3106 permissions on /dev/null. Exempt it from some checks.
3107 Reported by Andreas M. Kirchwitz.
3109 PP/03 Report version information for many libraries, including
3110 Exim version information for dynamically loaded libraries. Created
3111 version.h, now support a version extension string for distributors
3112 who patch heavily. Dynamic module ABI change.
3114 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3115 privilege escalation vulnerability whereby the Exim run-time user
3116 can cause root to append content of the attacker's choosing to
3119 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3122 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3123 If dropping privileges for untrusted macros, we disabled normal logging
3124 on the basis that it would fail; for the Exim run-time user, this is not
3125 the case, and it resulted in successful deliveries going unlogged.
3126 Fixed. Reported by Andreas Metzler.
3132 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3133 only prepend for the Resent-* case. Fixes regression introduced in
3134 Exim 4.70 by NM/22 for Bugzilla 607.
3136 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3137 increasing numbers of administrators be bitten by this.
3139 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3141 PP/03 Bugzilla 994: added openssl_options main configuration option.
3143 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3145 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3147 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3149 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3150 without a peer certificate, leading to a segfault because of an
3151 assumption that peers always have certificates. Be a little more
3152 paranoid. Problem reported by Martin Tscholak.
3154 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3155 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3156 NB: ClamAV planning to remove STREAM in "middle of 2010".
3157 CL also introduces -bmalware, various -d+acl logging additions and
3158 more caution in buffer sizes.
3160 PP/09 Implemented reverse_ip expansion operator.
3162 PP/10 Bugzilla 937: provide a "debug" ACL control.
3164 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3166 PP/12 Bugzilla 973: Implement --version.
3168 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3170 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3172 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3174 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3175 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3176 ignore trailing whitespace.
3178 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3180 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3183 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3184 Notification from Dr Andrew Aitchison.
3186 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3187 ExtendedDetectionInfo response format.
3188 Notification from John Horne.
3190 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3193 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3194 XSL and documented dependency on system catalogs, with examples of how
3197 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3200 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3201 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3202 configuration file which is writeable by the Exim user or group.
3204 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3205 of configuration files to cover files specified with the -C option if
3206 they are going to be used with root privileges, not just the default
3209 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3210 option (effectively making it always true).
3212 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3213 files to be used while preserving root privileges.
3215 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3216 that rogue child processes cannot use them.
3218 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3219 run-time user, instead of root.
3221 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3222 Exim run-time user without dropping privileges.
3224 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3225 result string, instead of calling string_vformat() twice with the same
3228 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3229 for other users. Others should always drop root privileges if they use
3230 -C on the command line, even for a whitelisted configure file.
3232 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3234 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3240 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3241 $data_path, and $header_path variables; fixed documentation bugs and
3244 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3245 exipick to access non-standard spools, including the "frozen" queue
3248 NM/01 Bugzilla 965: Support mysql stored procedures.
3249 Patch from Alain Williams
3251 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3253 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3254 Patch from Andreas Metzler
3256 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3257 Patch from Kirill Miazine
3259 NM/05 Bugzilla 671: Added umask to procmail example.
3261 JJ/03 installed exipick 20100323.0, fixing doc bug
3263 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3264 directory. Notification and patch from Dan Rosenberg.
3266 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3268 TK/02 Improve log output when DKIM signing operation fails.
3270 MH/01 Treat the transport option dkim_domain as a colon separated
3271 list, not as a single string, and sign the message with each element,
3272 omitting multiple occurences of the same signer.
3274 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3275 Bugzilla 985, 986. Patch by Simon Arlott
3277 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3278 Patch by Simon Arlott
3280 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3281 MBX locking. Notification from Dan Rosenberg.
3287 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3289 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3291 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3293 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3295 NM/04 Bugzilla 915: Flag broken perl installation during build.
3301 TK/01 Added patch by Johannes Berg that expands the main option
3302 "spamd_address" if it starts with a dollar sign.
3304 TK/02 Write list of recipients to X-Envelope-Sender header when building
3305 the mbox-format spool file for content scanning (suggested by Jakob
3308 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3309 (http://www.dcc-servers.net/) support via dccifd. Activated by
3310 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3312 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3313 by Mark Daniel Reidel <mr@df.eu>.
3315 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3316 When building exim an external PCRE library is now needed -
3317 PCRE is a system library on the majority of modern systems.
3318 See entry on PCRE_LIBS in EDITME file.
3320 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3321 conversation. Added nologin parameter to request.
3322 Patch contributed by Kirill Miazine.
3324 TF/01 Do not log submission mode rewrites if they do not change the address.
3326 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3328 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3329 log files in place. Contributed by Roberto Lima.
3331 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3333 TF/03 Bugzilla 615: When checking the local_parts router precondition
3334 after a local_part_suffix or local_part_prefix option, Exim now
3335 does not use the address's named list lookup cache, since this
3336 contains cached lookups for the whole local part.
3338 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3339 Robert Millan. Documentation is in experimental-spec.txt.
3341 TF/04 Bugzilla 668: Fix parallel build (make -j).
3343 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3345 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3346 Patch provided by Jan Srzednicki.
3348 TF/05 Leading white space used to be stripped from $spam_report which
3349 wrecked the formatting. Now it is preserved.
3351 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3352 that they are available at delivery time.
3354 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3356 TF/08 TLS error reporting now respects the incoming_interface and
3357 incoming_port log selectors.
3359 TF/09 Produce a more useful error message if an SMTP transport's hosts
3360 setting expands to an empty string.
3362 NM/06 Bugzilla 744: EXPN did not work under TLS.
3363 Patch provided by Phil Pennock.
3365 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3366 Patch provided by Richard Godbee.
3368 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3369 acl_smtp_notquit, added index entry.
3371 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3372 Patch provided by Eugene Bujak.
3374 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3375 accept(). Patch provided by Maxim Dounin.
3377 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3378 Patch provided by Phil Pennock.
3380 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3382 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3383 Patch provided by Brad "anomie" Jorsch.
3385 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3386 Patch provided by Dean Brooks.
3388 TK/05 Add native DKIM support (does not depend on external libraries).
3390 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3391 Patch provided by Graeme Fowler.
3393 NM/16 Bugzilla 851: Documentation example syntax fix.
3395 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3397 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3400 NM/19 Bugzilla 745: TLS version reporting.
3401 Patch provided by Phil Pennock.
3403 NM/20 Bugzilla 167: bool: condition support.
3404 Patch provided by Phil Pennock.
3406 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3407 clients. Patch provided by Phil Pennock.
3409 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3410 Patch provided by Brad "anomie" Jorsch.
3412 NM/23 Bugzilla 687: Fix misparses in eximstats.
3413 Patch provided by Heiko Schlittermann.
3415 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3416 Patch provided by Heiko Schlittermann.
3418 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3419 plus update to original patch.
3421 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3423 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3424 Patch provided by David Brownlee.
3426 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3428 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3430 NM/30 Bugzilla 888: TLS documentation bugfixes.
3432 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3434 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3435 Unlike the original bugzilla I have changed all shell scripts in src tree.
3437 NM/33 Bugzilla 898: Transport filter timeout fix.
3438 Patch by Todd Rinaldo.
3440 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3441 Patch by Serge Demonchaux.
3443 NM/35 Bugzilla 39: Base64 decode bug fixes.
3444 Patch by Jakob Hirsch.
3446 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3448 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3450 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3452 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3458 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3459 ALT-N's libdkim that I have put here:
3460 http://duncanthrax.net/exim-experimental/
3462 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3463 were called 'true' and it seems that C99 defines that as a reserved
3464 keyword to be used with 'bool' variable types. That means you could
3465 not include C99-style headers which use bools without triggering
3466 build errors in sieve.c.
3468 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3469 as mailq or other aliases. Changed the --help handling significantly
3470 to do whats expected. exim_usage() emits usage/help information.
3472 SC/01 Added the -bylocaldomain option to eximstats.
3474 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3476 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3478 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3484 PH/01 Another patch from the Sieve maintainer.
3486 PH/02 When an IPv6 address is converted to a string for single-key lookup
3487 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3488 dots are used instead of colons so that keys in lsearch files need not
3489 contain colons. This was done some time before quoting was made available
3490 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3491 (notated using the quote facility) so as to distinguish them from IPv4
3492 keys. This meant that lookups for IP addresses in host lists did not work
3493 for iplsearch lookups.
3495 This has been fixed by arranging for IPv6 addresses to be expressed with
3496 colons if the lookup type is iplsearch. This is not incompatible, because
3497 previously such lookups could never work.
3499 The situation is now rather anomalous, since one *can* have colons in
3500 ordinary lsearch keys. However, making the change in all cases is
3501 incompatible and would probably break a number of configurations.
3503 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3506 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3507 conversion specification without a maximum field width, thereby enabling
3508 a rogue spamd server to cause a buffer overflow. While nobody in their
3509 right mind would setup Exim to query an untrusted spamd server, an
3510 attacker that gains access to a server running spamd could potentially
3511 exploit this vulnerability to run arbitrary code as the Exim user.
3513 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3514 $primary_hostname instead of what libspf2 thinks the hosts name is.
3516 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3517 a directory entry by the name of the lookup key. Previously, if a
3518 symlink pointed to a non-existing file or a file in a directory that
3519 Exim lacked permissions to read, a lookup for a key matching that
3520 symlink would fail. Now it is enough that a matching directory entry
3521 exists, symlink or not. (Bugzilla 503.)
3523 PH/03 The body_linecount and body_zerocount variables are now exported in the
3526 PH/04 Added the $dnslist_matched variable.
3528 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3529 This means they are set thereafter only if the connection becomes
3532 PH/06 Added the client_condition to authenticators so that some can be skipped
3533 by clients under certain conditions.
3535 PH/07 The error message for a badly-placed control=no_multiline_responses left
3536 "_responses" off the end of the name.
3538 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3540 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3541 (without spaces) instead of just copying the configuration text.
3543 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3545 PH/11 Added $max_received_linelength.
3547 PH/12 Added +ignore_defer and +include_defer to host lists.
3549 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3550 way in which PCRE > 7.0 is built.
3552 PH/14 Implemented queue_only_load_latch.
3554 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3555 MAIL command. The effect was to mangle the value on 64-bit systems.
3557 PH/16 Another patch from the Sieve maintainer.
3559 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3561 PH/18 If a system quota error occurred while trying to create the file for
3562 a maildir delivery, the message "Mailbox is full" was not appended to the
3563 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3564 only to a quota excession during the actual writing of the file.
3566 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3567 characters?) which causes problems in log lines. The DN values are now
3568 passed through string_printing() before being added to log lines.
3570 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3571 and InterBase are left for another time.)
3573 PH/21 Added message_body_newlines option.
3575 PH/22 Guard against possible overflow in moan_check_errorcopy().
3577 PH/23 POSIX allows open() to be a macro; guard against that.
3579 PH/24 If the recipient of an error message contained an @ in the local part
3580 (suitably quoted, of course), incorrect values were put in $domain and
3581 $local_part during the evaluation of errors_copy.
3587 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3588 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3591 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3592 issue a MAIL command.
3594 PH/02 In an ACL statement such as
3596 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3598 if a client was not listed at all, or was listed with a value other than
3599 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3600 the condition was not true (as it should be), so access was not denied.
3601 The bug was that the ! inversion was incorrectly passed on to the second
3602 item. This has been fixed.
3604 PH/03 Added additional dnslists conditions == and =& which are different from
3605 = and & when the dns lookup returns more than one IP address.
3607 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3608 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3610 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3611 FSYNC, which compiles an option called disable_fsync that allows for
3612 bypassing fsync(). The documentation is heavily laced with warnings.
3614 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3616 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3617 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3618 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3619 including adding "make clean"; (3) Added -fPIC when compiling the test
3620 dynamically loaded module, to get rid of a warning.
3622 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3623 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3624 = 0s. The bug is otherwise harmless.
3626 PH/07 There was a bug in the dovecot authenticator such that the value of
3627 $auth1 could be overwritten, and so not correctly preserved, after a
3628 successful authentication. This usually meant that the value preserved by
3629 the server_setid option was incorrect.
3631 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3633 PH/09 Installed PCRE release 7.0.
3635 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3636 run for batched SMTP input. It is now run at the start of every message
3637 in the batch. While fixing this I discovered that the process information
3638 (output by running exiwhat) was not always getting set for -bs and -bS
3639 input. This is fixed, and it now also says "batched" for BSMTP.
3641 PH/11 Added control=no_pipelining.
3643 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3644 patch, slightly modified), and move the expansion of helo_data till after
3645 the connection is made in the smtp transport (so it can use these
3648 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3650 PH/14 Added log_selector = +pid.
3652 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3654 PH/16 Add ${if forany and ${if forall.
3656 PH/17 Added dsn_from option to vary the From: line in DSNs.
3658 PH/18 Flush SMTP output before performing a callout, unless control =
3659 no_callout_flush is set.
3661 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3662 was true (the default) a successful delivery failed to delete the retry
3663 item, thus causing premature timeout of the address. The bug is now
3666 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3668 PH/21 Long custom messages for fakedefer and fakereject are now split up
3669 into multiline responses in the same way that messages for "deny" and
3670 other ACL rejections are.
3672 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3673 with slight modification.
3675 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3676 draft, changing the syntax and factoring some duplicate code".
3678 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3679 for deliveries of the second and subsequent messages over the same SMTP
3682 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3683 ${reduce, with only minor "tidies".
3685 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3687 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3688 expansion side effects.
3690 PH/27 When a message times out after an over-quota error from an Exim-imposed
3691 quota, the bounce message says "mailbox is full". This message was not
3692 being given when it was a system quota that was exceeded. It now should
3695 MH/03 Made $recipients available in local_scan(). local_scan() already has
3696 better access to the recipient list through recipients_list[], but
3697 $recipients can be useful in postmaster-provided expansion strings.
3699 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3700 in the case of a MAIL command with additional options following the
3701 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3702 were accidentally chopped off.
3704 PH/29 SMTP synchronization checks are implemented when a command is read -
3705 there is a check that no more input is waiting when there shouldn't be
3706 any. However, for some commands, a delay in an ACL can mean that it is
3707 some time before the response is written. In this time, more input might
3708 arrive, invalidly. So now there are extra checks after an ACL has run for
3709 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3710 pipelining has not been advertised.
3712 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3714 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3715 returned in the SMTP response when smtp_return_error_details was false.
3716 This has been fixed.
3718 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3719 instead of the C I/O that was originally supplied, because problems were
3720 reported on Solaris.
3722 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3723 Exim which did not show up earlier: it was assuming that a call to
3724 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3725 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3726 was a macro that became an assignment, so it seemed to work. This has
3727 changed to a proper function call with a void return, hence the compile
3728 error. Exim's code has been fixed.
3730 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3733 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3735 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3737 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3738 because it turns out to be too big or there is a timeout) there is no
3739 "Completed" line in the log. When some messages of this type were
3740 selected by exigrep, they were listed as "not completed". Others were
3741 picked up by some special patterns. I have improved the selection
3742 criteria to be more general.
3744 PH/38 The host_find_failed option in the manualroute router can now be set
3745 to "ignore", to completely ignore a host whose IP address cannot be
3746 found. If all hosts are ignored, the behaviour is controlled by the new
3747 host_all_ignored option.
3749 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3750 homing or because of multiple MX records with /mx) generated more than
3751 one IP address, and the following item turned out to be the local host,
3752 all the secondary addresses of the first item were incorrectly removed
3753 from the list, along with the local host and any following hosts (which
3754 is what is supposed to happen).
3756 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3757 whoever called Exim into the -H file. In the case of the daemon it was
3758 behaving confusingly. When first started, it used values for whoever
3759 started the daemon, but after a SIGHUP it used the Exim user (because it
3760 calls itself on a restart). I have changed the code so that it now always
3763 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3764 message are rejected with the same error (e.g. no authentication or bad
3765 sender address), and a DATA command is nevertheless sent (as can happen
3766 with PIPELINING or a stupid MUA), the error message that was given to the
3767 RCPT commands is included in the rejection of the DATA command. This is
3768 intended to be helpful for MUAs that show only the final error to their
3771 PH/42 Another patch from the Sieve maintainer.
3773 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3774 Eximstats - Fixed some broken HTML links and added missing column headers
3776 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3777 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3779 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3781 PH/43 Yet another patch from the Sieve maintainer.
3783 PH/44 I found a way to check for a TCP/IP connection going away before sending
3784 the response to the final '.' that terminates a message, but only in the
3785 case where the client has not sent further data following the '.'
3786 (unfortunately, this is allowed). However, in many cases there won't be
3787 any further data because there won't be any more messages to send. A call
3788 to select() can be used: if it shows that the input is "ready", there is
3789 either input waiting, or the socket has been closed. An attempt to read
3790 the next input character can distinguish the two cases. Previously, Exim
3791 would have sent an OK response which the client would never have see.
3792 This could lead to message repetition. This fix should cure that, at
3793 least in a lot of common cases.
3795 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3796 advertised in response to EHLO.
3802 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3803 fixed by 4.65/MH/01 (is this a record?) are fixed:
3805 (i) An empty string was always treated as zero by the numeric comparison
3806 operators. This behaviour has been restored.
3808 (ii) It is documented that the numeric comparison operators always treat
3809 their arguments as decimal numbers. This was broken in that numbers
3810 starting with 0 were being interpreted as octal.
3812 While fixing these problems I realized that there was another issue that
3813 hadn't been noticed. Values of message_size_limit (both the global option
3814 and the transport option) were treated as octal if they started with 0.
3815 The documentation was vague. These values are now always treated as
3816 decimal, and I will make that clear in the documentation.
3822 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3823 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3826 MH/01 Don't check that the operands of numeric comparison operators are
3827 integers when their expansion is in "skipping" mode (fixes bug
3828 introduced by 4.64-PH/07).
3830 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3831 child addresses, Exim now panics and dies. Previously, because the count
3832 is held in a short int, deliveries were likely to be lost. As such a
3833 large number of recipients for a single message is ridiculous
3834 (performance will be very, very poor), I have chosen to impose a limit
3835 rather than extend the field.
3841 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3842 leftover -K file (the existence of which was triggered by #402).
3843 While we were at it, introduced process PID as part of the -K
3844 filename. This should rule out race conditions when creating
3847 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3848 processing considerably. Previous code took too long for large mails,
3849 triggering a timeout which in turn triggers #401.
3851 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3852 in the DK code in transports.c. sendfile() is not really portable,
3853 hence the _LINUX specificness.
3855 TF/01 In the add_headers option to the mail command in an Exim filter,
3856 there was a bug that Exim would claim a syntax error in any
3857 header after the first one which had an odd number of characters
3860 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3861 callout verification, Exim cached a "reject" for the entire domain. This
3862 is correct for most verifications, but it is not correct for a recipient
3863 verification with use_sender or use_postmaster set, because in that case
3864 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3865 case of MAIL FROM:<> rejection from other early rejections (e.g.
3866 rejection of HELO). When verifying a recipient using a non-null MAIL
3867 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3868 Whatever the result of the callout, the value of the domain cache is
3869 left unchanged (for any other kind of callout, getting as far as trying
3870 RCPT means that the domain itself is ok).
3872 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3875 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3876 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3877 session, the daemon ignores ECONNECT errors and logs others; it now
3878 ignores EPIPE as well.
3880 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3881 (quoted-printable decoding).
3883 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3884 later the small subsequent patch to fix an introduced bug.
3886 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3888 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3890 PH/08 An error is now given if message_size_limit is specified negative.
3892 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3893 to be given (somewhat) arbitrary names.
3895 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3898 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3899 miscellaneous code fixes
3901 PH/10 Added the log_reject_target ACL modifier to specify where to log
3904 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3905 hostname. This is wrong, because it relates to the incoming message (and
3906 probably the interface on which it is arriving) and not to the outgoing
3907 callout (which could be using a different interface). This has been
3908 changed to use the value of the helo_data option from the smtp transport
3909 instead - this is what is used when a message is actually being sent. If
3910 there is no remote transport (possible with a router that sets up host
3911 addresses), $smtp_active_hostname is used.
3913 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3914 tweaks were necessary in order to get it to work (see also 21 below):
3915 (a) The code assumed that strncpy() returns a negative number on buffer
3916 overflow, which isn't the case. Replaced with Exim's string_format()
3918 (b) There were several signed/unsigned issues. I just did the minimum
3919 hacking in of casts. There is scope for a larger refactoring.
3920 (c) The code used strcasecmp() which is not a standard C function.
3921 Replaced with Exim's strcmpic() function.
3922 (d) The code set only $1; it now sets $auth1 as well.
3923 (e) A simple test gave the error "authentication client didn't specify
3924 service in request". It would seem that Dovecot has changed its
3925 interface. Fortunately there's a specification; I followed it and
3926 changed what the client sends and it appears to be working now.
3928 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3931 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3932 address A is aliased to B and C, where B exists and C does not. Without
3933 -v the output is "A verified" because verification stops after a
3934 successful redirection if more than one address is generated. However,
3935 with -v the child addresses are also verified. Exim was outputting "A
3936 failed to verify" and then showing the successful verification for C,
3937 with its parentage. It now outputs "B failed to verify", showing B's
3938 parentage before showing the successful verification of C.
3940 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3941 look up a TXT record in a specific list after matching in a combined
3944 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3945 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3946 they consult the DNS. I had assumed they would set it the way they
3947 wanted; and indeed my experiments on Linux seem to show that in some
3948 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3949 To be on the safe side, however, I have now made the interface to
3950 host_find_byname() similar to host_find_bydns(), with an argument
3951 containing the DNS resolver options. The host_find_byname() function now
3952 sets these options at its start, just as host_find_bydns() does. The smtp
3953 transport options dns_qualify_single and dns_search_parents are passed to
3954 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3955 of host_find_byname() use the default settings of RES_DEFNAMES
3956 (qualify_single) but not RES_DNSRCH (search_parents).
3958 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3959 spool_read_header() do less string testing, by means of a preliminary
3960 switch on the second character of optional "-foo" lines. (This is
3961 overdue, caused by the large number of possibilities that now exist.
3962 Originally there were few.) While I was there, I also converted the
3963 str(n)cmp tests so they don't re-test the leading "-" and the first
3964 character, in the hope this might squeeze out yet more improvement.
3966 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3967 flag allowing group syntax was set by the header_syntax check but not
3968 turned off, possible causing trouble later; (2) The flag was not being
3969 set at all for the header_verify test, causing "group"-style headers to
3970 be rejected. I have now set it in this case, and also caused header_
3971 verify to ignore an empty address taken from a group. While doing this, I
3972 came across some other cases where the code for allowing group syntax
3973 while scanning a header line wasn't quite right (mostly, not resetting
3974 the flag correctly in the right place). These bugs could have caused
3975 trouble for malformed header lines. I hope it is now all correct.
3977 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3978 with the "reply" argument non-NULL. The code, however (which originally
3979 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3980 but it didn't always do it. This confused somebody who was copying the
3981 code for some other use. I have removed all the tests.
3983 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3984 feature that was used to support insecure browsers during the U.S. crypto
3985 embargo. It requires special client support, and Exim is probably the
3986 only MTA that supported it -- and would never use it because real RSA is
3987 always available. This code has been removed, because it had the bad
3988 effect of slowing Exim down by computing (never used) parameters for the
3989 RSA_EXPORT functionality.
3991 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3992 authenticator to fail if there's a tab character in the incoming data
3993 (there should never be unless someone is messing about, as it's supposed
3994 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3995 if the connection is using TLS or if the remote IP is the same as the
3996 local IP, and the "valid-client-cert option" if a client certificate has
3999 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4000 authenticators. This can be used for authorization after authentication
4001 succeeds. (In the case of plaintext, it servers for both authentication
4004 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4005 if any retry times were supplied.
4007 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4008 connection, where there is no client IP address to check. In this
4009 situation, the verify now always succeeds.
4011 PH/25 Applied John Jetmore's -Mset patch.
4013 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4015 PH/27 In a string expansion for a processed (not raw) header when multiple
4016 headers of the same name were present, leading whitespace was being
4017 removed from all of them, but trailing whitespace was being removed only
4018 from the last one. Now trailing whitespace is removed from each header
4019 before concatenation. Completely empty headers in a concatenation (as
4020 before) are ignored.
4022 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4023 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4025 PH/29 [Removed. This was a change that I later backed out, and forgot to
4026 correct the ChangeLog entry (that I had efficiently created) before
4027 committing the later change.]
4029 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4030 address errors (4xx response to RCPT) over the same connection as other
4031 messages routed to the same hosts. Such deliveries are always "forced",
4032 so retry times are not inspected. This resulted in far too many retries
4033 for the affected addresses. The effect occurred only when there were more
4034 hosts than the hosts_max_try setting in the smtp transport when it had
4035 the 4xx errors. Those hosts that it had tried were not added to the list
4036 of hosts for which the message was waiting, so if all were tried, there
4037 was no problem. Two fixes have been applied:
4039 (i) If there are any address or message errors in an SMTP delivery, none
4040 of the hosts (tried or untried) are now added to the list of hosts
4041 for which the message is waiting, so the message should not be a
4042 candidate for sending over the same connection that was used for a
4043 successful delivery of some other message. This seems entirely
4044 reasonable: after all the message is NOT "waiting for some host".
4045 This is so "obvious" that I'm not sure why it wasn't done
4046 previously. Hope I haven't missed anything, but it can't do any
4047 harm, as the worst effect is to miss an optimization.
4049 (ii) If, despite (i), such a delivery is accidentally attempted, the
4050 routing retry time is respected, so at least it doesn't keep
4051 hammering the server.
4053 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4054 in ${readsocket because some servers need this prod.
4056 PH/32 Added some extra debug output when updating a wait-xxx database.
4058 PH/33 The hint "could be header name not terminated by colon", which has been
4059 given for certain expansion errors for a long time, was not being given
4060 for the ${if def:h_colon_omitted{... case.
4062 PH/34 The spec says: "With one important exception, whenever a domain list is
4063 being scanned, $domain contains the subject domain." There was at least
4064 one case where this was not true.
4066 PH/35 The error "getsockname() failed: connection reset by peer" was being
4067 written to the panic log as well as the main log, but it isn't really
4068 panic-worthy as it just means the connection died rather early on. I have
4069 removed the panic log writing for the ECONNRESET error when getsockname()
4072 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4073 runs only) independently of the message's sender address. This meant
4074 that, if the 4xx error was in fact related to the sender, a different
4075 message to the same recipient with a different sender could confuse
4076 things. In particular, this can happen when sending to a greylisting
4077 server, but other circumstances could also provoke similar problems.
4078 I have changed the default so that the retry time for these errors is now
4079 based a combination of the sender and recipient addresses. This change
4080 can be overridden by setting address_retry_include_sender=false in the
4083 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4084 remote server are returned as part of bounce messages. This was not
4085 happening for LMTP over a pipe (the lmtp transport), but now it is the
4086 same for both kinds of LMTP.
4088 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4089 in header lines that were in fact CNAMEs. This is no longer the case.
4091 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4092 and queue runs started by the daemon processed all messages. This has
4093 been fixed so that -R and -S can now usefully be given with -q<time>.
4095 PH/40 Import PCRE release 6.7 (fixes some bugs).
4097 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4099 PH/42 Give an error if -q is specified more than once.
4101 PH/43 Renamed the variables $interface_address and $interface_port as
4102 $received_ip_address and $received_port, to make it clear that these
4103 values apply to message reception, and not to the outgoing interface when
4104 a message is delivered. (The old names remain recognized, of course.)
4106 PH/44 There was no timeout on the connect() call when using a Unix domain
4107 socket in the ${readsocket expansion. There now is.
4109 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4110 be meaningful with "accept".
4112 SC/01 Eximstats V1.43
4113 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4115 SC/02 Eximstats V1.44
4116 Use a glob alias rather than an array ref in the generated
4117 parser. This improves both readability and performance.
4119 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4120 Collect SpamAssassin and rejection statistics.
4121 Don't display local sender or destination tables unless
4122 there is data to show.
4123 Added average volumes into the top table text output.
4125 SC/04 Eximstats V1.46
4126 Collect data on the number of addresses (recipients)
4127 as well as the number of messages.
4129 SC/05 Eximstats V1.47
4130 Added 'Message too big' to the list of mail rejection
4131 reasons (thanks to Marco Gaiarin).
4133 SC/06 Eximstats V1.48
4134 Mainlog lines which have GMT offsets and are too short to
4135 have a flag are now skipped.
4137 SC/07 Eximstats V1.49 (Alain Williams)
4138 Added the -emptyok flag.
4140 SC/08 Eximstats V1.50
4141 Fixes for obtaining the IP address from reject messages.
4143 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4144 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4145 whitespace changes from 4.64-PH/27
4147 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4150 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4151 are found, allow negative numbers in numeric criteria)
4153 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4155 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4156 to match changes made in 4.64-PH/43
4158 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4160 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4161 "domain" is actually an IP address, and force a failure. This locks out
4162 those revolvers/nameservers that support "A-for-A" lookups, in
4163 contravention of the specifications.
4165 PH/48 When a host name was looked up from an IP address, and the subsequent
4166 forward lookup of the name timed out, the host name was left in
4167 $sender_host_name, contrary to the specification.
4169 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4170 restricted to single-key lookups, Exim was not diagnosing an error if
4171 * or *@ was used with a query-style lookup.
4173 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4175 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4176 long ago, but noone interested enough thought of it. Let's just say that
4177 the "1.1" means that there are some new functions that weren't there at
4178 some point in the past.
4180 PH/51 Error processing for expansion failure of helo_data from an smtp
4181 transport during callout processing was broken.
4183 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4184 tested/used via the -bh/-bhc/-bs options.
4186 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4187 bug, fixed in subsequent PCRE releases).
4189 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4190 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4192 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4198 SC/01 Use a glob alias rather than an array ref in eximstats generated
4199 parser. This improves both readability and performance.
4201 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4202 Don't display local sender or destination tables in eximstats unless
4203 there is data to show.
4204 Added average volumes into the eximstats top table text output.
4206 SC/03 Collect data on the number of addresses (recipients) as well
4207 as the number of messages in eximstats.
4209 TF/01 Correct an error in the documentation for the redirect router. Exim
4210 does (usually) call initgroups() when daemonizing.
4212 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4213 with consistent privilege compared to when running as a daemon.
4215 TF/03 Note in the spec that $authenticated_id is not set for local
4216 submissions from trusted users.
4218 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4219 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4221 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4222 by adding some example configuration directives to the default
4223 configuration file. A little bit of work is required to uncomment the
4224 directives and define how usernames and passwords are checked, but
4225 there is now a framework to start from.
4227 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4228 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4229 without this. I don't know how relevant this is to other LDAP libraries.
4231 PH/02 Add the verb name to the "unknown ACL verb" error.
4233 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4235 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4237 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4238 directory not expanded when it should be if an expanded home directory
4239 was set for the address (which is overridden by the transport).
4241 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4244 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4245 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4246 because it is too late at that time, and has no effect.
4248 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4249 security issue with \' (bugzilla #107). I could not use the
4250 PQescapeStringConn() function, because it needs a PGconn value as one of
4253 PH/08 When testing addresses using -bt, indicate those final addresses that
4254 are duplicates that would not cause an additional delivery. At least one
4255 person was confused, thinking that -bt output corresponded to deliveries.
4256 (Suppressing duplicates isn't a good idea as you lose the information
4257 about possibly different redirections that led to the duplicates.)
4259 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4260 systems where poll() doesn't work, in particular OS X.
4262 PH/10 Added more information to debugging output for retry time not reached.
4264 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4265 operations in malware.c.
4267 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4270 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4271 syslog_duplication set false, log lines that would normally be written
4272 both the the main log and to the reject log were not written to syslog at
4275 PH/14 In the default configuration, change the use of "message" in ACL warn
4276 statements to "add_header".
4278 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4279 not followed by a command (e.g. "seen endif").
4281 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4282 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4285 PH/17 Added extra conditions to the default value of delay_warning_condition
4289 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4290 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4291 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4294 The Auto-Submitted: and various List- headers are standardised, whereas I
4295 don't think Precedence: ever was.
4297 PH/18 Refactored debugging code in route_finduser() to show more information,
4298 in particular, the error code if getpwnam() issues one.
4300 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4301 This is apparently needed in addition to the PH/07 change above to avoid
4302 any possible encoding problems.
4304 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4305 but not after initializing Perl.
4307 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4308 output them only if debugging. By default they are written stderr,
4309 apparently, which is not desirable.
4311 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4314 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4317 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4319 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4320 authenticated or an ident call has been made. Suppress the default
4321 values for $authenticated_id and $authenticated_sender (but permit -oMai
4322 and -oMas) when testing with -bh.
4324 PH/24 Re-jigged the order of the tests in the default configuration so that the
4325 tests for valid domains and recipients precede the DNS black list and CSA
4326 tests, on the grounds that those ones are more expensive.
4328 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4329 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4330 This bug exists in every version of Exim that I still have, right back to
4333 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4334 However, an attempt to turn on case-sensitivity in a regex key by
4335 including (?-i) didn't work because the subject string was already
4336 lowercased, and the effects were non-intuitive. It turns out that a
4337 one-line patch can be used to allow (?-i) to work as expected.
4343 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4344 other effects) broke the use of negated acl sub-conditions.
4346 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4349 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4350 "Deny" causes Exim to reject the incoming connection with a 554 error.
4351 Unfortunately, if there is a major crisis, such as a disk failure,
4352 tcp-wrappers gives "deny", whereas what one would like would be some
4353 kind of temporary error. A kludge has been added to help with this.
4354 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4355 554 error is used if errno is still zero or contains ENOENT (which occurs
4356 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4359 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4361 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4362 errors. Otherwise a message that provokes a temporary error (when other
4363 messages do not) can cause a whole host to time out.
4365 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4366 addresses were routed directly to files or pipes from a redirect router.
4367 File deliveries just didn't batch; pipe deliveries might have suffered
4370 PH/06 A failure to get a lock for a hints database would erroneously always say
4371 "Failed to get write lock", even when it was really a read lock.
4373 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4374 of 0600. This has been changed to use the value of the lockfile_mode
4375 option (which defaults to 0600).
4377 PH/08 Applied small patch from the Sieve maintainer.
4379 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4380 folder from quota calculations, a direct delivery into this folder messed
4381 up the contents of the maildirsize file. This was because the regex was
4382 used only to exclude .Trash (or whatever) when the size of the mailbox
4383 was calculated. There was no check that a delivery was happening into an
4384 excluded directory. This bug has been fixed by ignoring all quota
4385 processing for deliveries into excluded directories.
4387 PH/10 Added the maildirfolder_create_regex option to appendfile.
4393 PH/01 The code for finding all the local interface addresses on a FreeBSD
4394 system running IPv6 was broken. This may well have applied to all BSD
4395 systems, as well as to others that have similar system calls. The broken
4396 code found IPv4 interfaces correctly, but gave incorrect values for the
4397 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4398 that it would not match correctly against @[] and not recognize the IPv6
4401 PH/02 The ipliteral router was not recognizing addresses of the form user@
4402 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4404 PH/03 Added disable_ipv6.
4406 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4407 decoded form, because it is most often used to construct To: headers
4408 lines in autoreplies, and the decoded form may well be syntactically
4409 invalid. However, $reply_address has leading white space removed, and all
4410 newlines turned into spaces so that the autoreply transport does not
4413 PH/05 If group was specified without a user on a router, and no group or user
4414 was specified on a transport, the group from the router was ignored.
4416 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4417 for visible compile-time settings that can be used to change these
4418 numbers, for those that want even more. Backwards compatibility with old
4419 spool files has been maintained. However, going back to a previous Exim
4420 release will lost any variables that are in spool files.
4422 PH/07 Two small changes when running in the test harness: increase delay when
4423 passing a TCP/IP connection to a new process, in case the original
4424 process has to generate a bounce, and remove special handling of
4425 127.0.0.2 (sic), which is no longer necessary.
4427 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4428 be the same on different OS.
4430 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4433 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4434 whether --show-vars was specified or not
4436 JJ/02 exipick: Added support for new ACL variable spool format introduced
4439 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4440 syntactically invalid From: or Reply-to: line, and a filter used this to
4441 generate an autoreply, and therefore failed to obtain an address for the
4442 autoreply, Exim could try to deliver to a non-existent relative file
4443 name, causing unrelated and misleading errors. What now happens is that
4444 it logs this as a hard delivery error, but does not attempt to create a
4447 PH/11 The exinext utility has a -C option for testing purposes, but although
4448 the given file was scanned by exinext itself; it wasn't being passed on
4449 when Exim was called.
4451 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4452 an end-of-file indication when reading a command response.
4454 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4455 compiled. In many other places in Exim, IPv6 addresses are always
4456 recognized, so I have changed this. It also means that IPv4 domain
4457 literals of the form [IPV4:n.n.n.n] are now always recognized.
4459 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4460 used if the router is not running as root, for example, when verifying at
4461 ACL time, or when using -bh. The debugging output from this situation was
4462 non-existent - all you got was a failure to exec. I have made two
4465 (a) Failures to set uid/gid, the current directory, or a process leader
4466 in a subprocess such as that created by queryprogram now generate
4467 suitable debugging output when -d is set.
4469 (b) The queryprogram router detects when it is not running as root,
4470 outputs suitable debugging information if -d is set, and then runs
4471 the subprocess without attempting to change uid/gid.
4473 PH/15 Minor change to Makefile for building test_host (undocumented testing
4476 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4477 additional section of a DNS packet that returns MX or SRV records.
4478 Instead, it always explicitly searches for A/AAAA records. This avoids
4479 major problems that occur when a DNS server includes only records of one
4480 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4481 fixed another bug: if SRV records were looked up and the corresponding
4482 address records were *not* found in the additional section, the port
4483 values from the SRV records were lost.
4485 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4486 using the correct key (the original address) when searching the retry
4487 rules in order to find which one to use for generating the retry hint.
4489 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4490 adding the default one. Similarly, if it contains a Reply-To: header, the
4491 errors_reply_to option, if set, is not used.
4493 PH/19 When calculating a retry time, Exim used to measure the "time since
4494 failure" by looking at the "first failed" field in the retry record. Now
4495 it does not use this if it is later than than the arrival time of the
4496 message. Instead it uses the arrival time. This makes for better
4497 behaviour in cases where some deliveries succeed, thus re-setting the
4498 "first failed" field. An example is a quota failure for a huge message
4499 when small messages continue to be delivered. Without this change, the
4500 "time since failure" will always be short, possible causing more frequent
4501 delivery attempts for the huge message than are intended.
4502 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4504 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4505 $1, $2, $3) because the numerical variables can be reset during some
4506 expansion items (e.g. "match"), thereby losing the authentication data.
4508 PH/21 Make -bV show the size of off_t variables so that the test suite can
4509 decide whether to run tests for quotas > 2G.
4511 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4512 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4513 filecount value is greater than 2G or if a quota value is greater than 2G
4514 on a system where the size of off_t is not greater than 4, a panic error
4517 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4518 never match. The debug and -bh output now contains an explicit error
4519 message indicating a malformed IPv4 address or mask.
4521 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4522 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4523 PH/23 above applies.
4525 PH/25 Do not write to syslog when running in the test harness. The only
4526 occasion when this arises is a failure to open the main or panic logs
4527 (for which there is an explicit test).
4529 PH/26 Added the /no_tell option to "control=freeze".
4531 PH/27 If a host name lookup failed very early in a connection, for example, if
4532 the IP address matched host_lookup and the reverse lookup yielded a name
4533 that did not have a forward lookup, an error message of the form "no IP
4534 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4535 could be logged. Now it outputs the IP address instead of "NULL".
4537 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4538 allows the sender and the authenticated sender to be set when
4539 submitting a message from within Exim. Since child_open_exim() is
4540 documented for local_scan(), the new function should be too.
4542 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4543 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4544 results in an empty string is now treated as unset.
4546 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4548 PH/31 Added sender_verify_fail logging option.
4550 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4551 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4552 tidied the source and removed it altogether.
4554 PH/33 When a queue run was abandoned because the load average was too high, a
4555 log line was always written; now it is written only if the queue_run log
4556 selector is set. In addition, the log line for abandonment now contains
4557 information about the queue run such as the pid. This is always present
4558 in "start" and "stop" lines but was omitted from the "abandon" line.
4560 PH/34 Omit spaces between a header name and the colon in the error message that
4561 is given when verify = headers_syntax fails (if there are lots of them,
4562 the message gets confusing).
4564 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4565 names, as there are now some PTR records that contain slashes. This check
4566 is only to protect against broken name servers that fall over on strange
4567 characters, so the fact that it applies to all lookups doesn't matter.
4569 PH/36 Now that the new test suite is complete, we can remove some of the
4570 special code in Exim that was needed for the old test suite. For example,
4571 sorting DNS records because real resolvers return them in an arbitrary
4572 order. The new test suite's fake resolver always returns records in the
4575 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4576 bounces) except when queue_only is set, to avoid logging races between
4577 the different processes.
4579 PH/38 Panic-die if .include specifies a non-absolute path.
4581 PH/39 A tweak to the "H" retry rule from its user.
4583 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4584 a label. They prevented compilation on older perls.
4586 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4587 a warning to be raised on newish perls.
4589 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4590 on queue. Changes to match documented behaviour of showing count of
4591 messages matching specified criteria.
4593 PH/40 Changed the default ident timeout from 30s to 5s.
4595 PH/41 Added support for the use of login_cap features, on those BSD systems
4596 that have them, for controlling the resources used by pipe deliveries.
4598 PH/42 The content-scanning code uses fopen() to create files in which to put
4599 message data. Previously it was not paying any attention to the mode of
4600 the files. Exim runs with umask(0) because the rest of the code creates
4601 files with open(), and sets the required mode explicitly. Thus, these
4602 files were ending up world-writeable. This was not a big issue, because,
4603 being within the spool directory, they were not world-accessible. I have
4604 created a function called modefopen, which takes an additional mode
4605 argument. It sets umask(777), creates the file, chmods it to the required
4606 mode, then resets the umask. All the relevant calls to fopen() in the
4607 content scanning code have been changed to use this function.
4609 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4610 to 24 hours. This avoids potential overflow problems when processing G
4611 and H retry rules. I suspect nobody ever tinkers with this value.
4613 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4615 PH/45 When the plaintext authenticator is running as a client, the server's
4616 challenges are checked to ensure they are valid base64 strings. By
4617 default, the authentication attempt is cancelled if an invalid string is
4618 received. Setting client_ignore_invalid_base64 true ignores these errors.
4619 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4620 they are received. Thus, the responses can be made to depend on the
4621 challenges. If an invalid string is ignored, an empty string is placed in
4624 PH/46 Messages that are created by the autoreply transport now contains a
4625 References: header, in accordance with RFCs 2822 and 3834.
4627 PH/47 Added authenticated_sender_force to the smtp transport.
4629 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4631 PH/49 Installed latest patch from the Sieve maintainer.
4633 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4634 was also set, the appendfile transport was unnecessarily scanning a
4635 directory of message files (e.g. for maildir delivery) to find the count
4636 of files (along with the size), even though it did not need this
4637 information. It now does the scan only if it needs to find either the
4638 size of the count of files.
4640 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4642 PH/52 Two bugs concerned with error handling when the smtp transport is
4645 (i) Exim was not creating retry information for temporary errors given
4646 for individual recipients after the DATA command when the smtp transport
4647 was used in LMTP mode. This meant that they could be retried too
4648 frequently, and not timed out correctly.
4650 (ii) Exim was setting the flag that allows error details to be returned
4651 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4652 recipients that were returned after the DATA command.
4654 PH/53 This is related to PH/52, but is more general: for any failing address,
4655 when detailed error information was permitted to be returned to the
4656 sender, but the error was temporary, then after the final timeout, only
4657 "retry timeout exceeded" was returned. Now it returns the full error as
4658 well as "retry timeout exceeded".
4660 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4661 do this, and (what is worse) MTAs that accept it.
4663 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4664 will now be deprecated.
4666 PH/56 New os.c-cygwin from the Cygwin maintainer.
4668 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4669 formats (previously only available in exim formats via -bpr, -bpru,
4670 and -bpra. Now also available in native and exiqgrep formats)
4672 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4673 with very large, slow to parse queues
4675 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4677 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4679 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4680 responses to authentication challenges, though it was showing the
4681 challenges; (ii) I've removed the CR characters from the debug output for
4684 PH/58 Allow for the insertion of a newline as well as a space when a string
4685 is turned into more than one encoded-word during RFC 2047 encoding. The
4686 Sieve code now uses this.
4688 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4689 data_4xx, lost_connection, tls_required.
4691 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4692 message was being sent as an SMTP response.
4694 PH/61 Add -l and -k options to exicyclog.
4696 PH/62 When verifying, if an address was redirected to one new address, so that
4697 verification continued, and the new address failed or deferred after
4698 having set something in $address_data, the value of $address_data was not
4699 passed back to the ACL. This was different to the case when no
4700 redirection occurred. The value is now passed back in both cases.
4702 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4703 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4704 use login_cap.h, so on its own it isn't the distinguishing feature. The
4705 new name refers directly to the setclassresources() function.
4707 PH/65 Added configuration files for NetBSD3.
4709 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4711 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4712 is preferred over IPv4.
4714 PH/68 The bounce_return_message and bounce_return_body options were not being
4715 honoured for bounces generated during the reception of non-SMTP messages.
4716 In particular, this applied to messages rejected by the ACL. This bug has
4717 been fixed. However, if bounce_return_message is true and bounce_return_
4718 body is false, the headers that are returned for a non-SMTP message
4719 include only those that have been read before the error was detected.
4720 (In the case of an ACL rejection, they have all been read.)
4722 PH/69 The HTML version of the specification is now built in a directory called
4723 spec_html instead of spec.html, because the latter looks like a path with
4724 a MIME-type, and this confuses some software.
4726 PH/70 Catch two compiler warnings in sieve.c.
4728 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4729 function verify_get_ident() calls ip_connect() to connect a socket, but
4730 if the "connect()" function timed out, ip_connect() used to close the
4731 socket. However, verify_get_ident() also closes the socket later, and in
4732 between Exim writes to the log, which may get opened at this point. When
4733 the socket was closed in ip_connect(), the log could get the same file
4734 descriptor number as the socket. This naturally causes chaos. The fix is
4735 not to close the socket in ip_connect(); the socket should be closed by
4736 the function that creates it. There was only one place in the code where
4737 this was missing, in the iplookup router, which I don't think anybody now
4738 uses, but I've fixed it anyway.
4740 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4741 well as to direct DNS lookups. Otherwise the handling of names in host
4742 lists is inconsistent and therefore confusing.
4748 PH/01 Two changes to the default runtime configuration:
4750 (1) Move the checks for relay_from_hosts and authenticated clients from
4751 after to before the (commented out) DNS black list checks.
4753 (2) Add control=submission to the relay_from_hosts and authenticated
4754 clients checks, on the grounds that messages accepted by these
4755 statements are most likely to be submissions.
4757 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4759 (1) Generate an error if the third argument for the ${prvs expansion is
4762 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4765 (3) Reset the variables that are obtained from the first argument of
4766 ${prvscheck and used in the second argument before leaving the code,
4767 because their memory is reclaimed, so using them afterwards may do
4770 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4771 one (it's much easier than Tom thought :-).
4773 (5) Because of (4), we can now allow for the use of $prvscheck_result
4774 inside the third argument.
4776 PH/03 For some reason, the default setting of PATH when running a command from
4777 a pipe transport was just "/usr/bin". I have changed it to
4780 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4781 anything to be listed in the output from -bV.
4783 PH/05 When a filter generated an autoreply, the entire To: header line was
4784 quoted in the delivery log line, like this:
4786 => >A.N.Other <ano@some.domain> <original@ddress> ...
4788 This has been changed so that it extracts the operative address. There
4789 may be more than one such address. If so, they are comma-separated, like
4792 => >ano@some.domain,ona@other.domain <original@ddress> ...
4794 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4795 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4796 not being looked up in the rDNS to get a host name, Exim was showing the
4797 IP address twice in Received: lines, even though the IP addresses were
4798 identical. For example:
4800 Received: from [1.2.3.4] (helo=[1.2.3.4])
4802 However, if the real host name was known, it was omitting the HELO data
4803 if it matched the actual IP address. This has been tidied up so that it
4804 doesn't show the same IP address twice.
4806 PH/07 When both +timestamp and +memory debugging was on, the value given by
4807 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4808 called by the debug printing, thereby overwriting the timestamp buffer.
4809 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4811 PH/08 When the original message was included in an autoreply transport, it
4812 always said "this is a copy of the message, including all the headers",
4813 even if body_only or headers_only was set. It now gives an appropriate
4816 PH/09 Applied a patch from the Sieve maintainer which:
4818 o fixes some comments
4819 o adds the (disabled) notify extension core
4820 o adds some debug output for the result of if/elsif tests
4821 o points to the current vacation draft in the documentation
4822 and documents the missing references header update
4826 o fixes a bug in processing the envelope test (when testing
4827 multiple envelope elements, the last element determined the
4830 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4831 Electronic Mail") by including:
4833 Auto-submitted: auto-generated
4835 in the messages that it generates (bounce messages and others, such as
4836 warnings). In the case of bounce messages for non-SMTP messages, there was
4837 also a typo: it was using "Auto_submitted" (underscore instead of
4838 hyphen). Since every message generated by Exim is necessarily in response
4839 to another message, thes have all been changed to:
4841 Auto-Submitted: auto-replied
4843 in accordance with these statements in the RFC:
4845 The auto-replied keyword:
4847 - SHOULD be used on messages sent in direct response to another
4848 message by an automatic process,
4850 - MUST NOT be used on manually-generated messages,
4852 - MAY be used on Delivery Status Notifications (DSNs) and Message
4853 Disposition Notifications (MDNs),
4855 - MUST NOT be used on messages generated by automatic or periodic
4856 processes, except for messages which are automatic responses to
4859 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4860 to the default Received: header definition.
4862 PH/12 Added log selector acl_warn_skipped (default on).
4864 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4865 variables because (a) they are in the wrong storage pool and (b) even if
4866 they were copied, it wouldn't work properly because of the caching.
4868 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4869 checking when decoding. Apparently there are clients that generate
4870 overlong encoded strings. Why am I not surprised?
4872 PH/15 If the first argument of "${if match_address" was not empty, but did not
4873 contain an "@" character, Exim crashed. Now it writes a panic log message
4874 and treats the condition as false.
4876 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4878 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4879 extension "envelope-auth". The code is finished and in agreement with
4880 other implementations, but there is no documentation so far and in fact,
4881 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4882 not changing the active code.
4884 Print executed "if" and "elsif" statements when debugging is used. This
4885 helps a great deal to understand what a filter does.
4887 Document more things not specified clearly in RFC3028. I had all this
4888 sorted out, when out of a sudden new issues came to my mind. Oops."
4890 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4893 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4894 canonical form (as documented). However, after a host name lookup from
4895 the IP address, check_host() was doing a simple string comparison with
4896 addresses acquired from the DNS when checking that the found name did
4897 have the original IP as one of its addresses. Since any found IPv6
4898 addresses are likely to be in abbreviated form, the comparison could
4899 fail. Luckily, there already exists a function for doing the comparison
4900 by converting both addresses to binary, so now that is used instead of
4901 the text comparison.
4903 PH/20 There was another similar case to PH/19, when a complete host name was
4904 given in a host list; looking up its IP address could give an abbreviated
4905 form, whereas the current host's name might or might not be abbreviated.
4906 The same fix has been applied.
4912 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4913 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4916 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4917 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4919 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4921 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4922 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4923 Also, for "Auto-Submitted", treat anything other than "no" as
4924 non-personal, in accordance with RFC 3834. (Previously it treated
4925 anything starting "auto-" as non-personal.)
4927 TF/01 The control=submission/name=... option had a problem with syntax
4928 errors if the name included a slash character. The /name= option
4929 now slurps the rest of the string, so it can include any characters
4930 but it must come last in the list of options (after /sender_retain
4933 PH/05 Some modifications to the interface to the fake nameserver for the new
4941 TK/01 Added the "success_on_redirect" address verification option. See
4942 NewStuff for rationale and an example.
4944 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4946 PH/02 Patch to exigrep to allow it to work on syslog lines.
4948 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4949 fread() to skip over the body file's header line, because in Cygwin the
4950 header line is locked and is inaccessible.
4952 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4953 co-exist for some time) to make it clear that it is the Exim ID that is
4954 referenced, not the Message-ID: header line.
4956 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4957 string_format() function, because snprintf() does not exist on all
4960 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4961 ${stat: expansion item.
4963 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4964 protocol synchronization error", to keep the pedants happy.
4966 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4967 well as for IRIX systems, when gcc is being used. See the host.c source
4970 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4972 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4975 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4976 transport and to the smtp transport in LMTP mode.
4978 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4980 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4981 run a filter in a subprocess. This could lead to confusion in subsequent
4982 lookups in the parent process. There should also be a search_tidyup() at
4983 the end of the subprocess.
4985 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4986 only if the host matched helo_try_verify_hosts, which caused the
4987 verification to occur when the EHLO/HELO command was issued. The ACL just
4988 tested the remembered result. Now, if a previous verification attempt has
4989 not happened, "verify = helo" does it there and then.
4991 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4993 TK/03 Fix log output including CR from clamd.
4995 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4996 exist provoked a memory error which could cause a segfault.
4998 PH/15 Installed PCRE 6.2
5000 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5002 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5003 of the problem. Specifically, suggested +O2 rather than +O1 for the
5006 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5008 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5009 record, the port was not correctly logged when the outgoing_port log
5010 selector was set (it logged the transort's default port).
5012 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5013 fallback_hosts, and "hosts" in the smtp transport.
5015 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5016 host errors such as "Connection refused".
5018 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5019 authentication with radiusclient 0.4.9:
5021 - Error returned from rc_read_config was caught wrongly
5022 - Username/password not passed on to radius server due to wrong length.
5024 The presumption is that some radiusclient API changes for 4.51/PH/17
5025 were not taken care of correctly. The code is still untested by me (my
5026 Linux distribution still has 0.3.2 of radiusclient), but it was
5027 contributed by a Radius user.
5029 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5030 expanding the "port" option of the smtp transport.
5032 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5033 while reading a MIME header. Thanks to Tom Hughes for a patch.
5035 PH/24 Include config.h inside local_scan.h so that configuration settings are
5038 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5039 that in an ACL for RCPT (for example), you can examine exactly what was
5042 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5043 commands, but it was not correctly comparing the address with the actual
5044 client host address. Thus, it would show the EHLO address in Received:
5045 header lines when this was not necessary.
5047 PH/27 Added the % operator to ${eval:}.
5049 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5050 it should be ignoring failures (because with -C, for example, it has lost
5051 privilege). It wasn't ignoring creation failures other than "already
5054 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5057 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5058 because an input line was too long, either on its own, or by virtue of
5059 too many continuations, the temporary file was not being removed, and the
5060 return code was incorrect.
5062 PH/31 Missing "BOOL" in function definition in filtertest.c.
5064 PH/32 Applied Sieve patches from the maintainer.
5066 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5068 PH/33 Added "verify = not_blind".
5070 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5071 Local/Makefile (with some defaults set). These are used in built scripts
5072 such as exicyclog, but they have never been used in the exim_install
5073 script (though there are many overriding facilities there). I have
5074 arranged that the exim_install script now takes note of these two
5077 PH/35 Installed configuration files for Dragonfly.
5079 PH/36 When a locally submitted message by a trusted user did not contain a
5080 From: header, and the sender address was obtained from -f or from an SMTP
5081 MAIL command, and the trusted user did not use -F to supply a sender
5082 name, $originator_name was incorrectly used when constructing a From:
5083 header. Furthermore, $originator_name was used for submission mode
5084 messages from external hosts without From: headers in a similar way,
5085 which is clearly wrong.
5087 PH/37 Added control=suppress_local_fixups.
5089 PH/38 When log_selector = +received_sender was set, and the addition of the
5090 sender made the log line's construction buffer exactly full, or one byte
5091 less than full, an overflow happened when the terminating "\n" was
5094 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5095 when the result of a list match is failure because a DNS lookup failed.
5097 PH/40 RM_COMMAND is now used in the building process.
5099 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5100 the "build-* directories that it finds.
5102 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5103 address in a domain literal was a prefix of an interface address.
5105 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5106 when verifying a sender address, unless rewrite_headers is false.
5108 PH/44 Wrote a long comment about why errors_to addresses are verified as
5109 recipients, not senders.
5111 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5112 the ratelimit ACL was added.
5114 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5116 PH/46 Added extra information about PostgreSQL errors to the error string.
5118 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5119 avoiding the need to install special zones in a real server. This is
5120 backwards compatible; if it can't find the fake resolver, it drops back.
5121 Thus, both old and new test suites can be run.
5123 TF/02 Added util/ratelimit.pl
5125 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5126 clock is set back in time.
5128 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5129 Candler <B.Candler@pobox.com>.
5131 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5132 OK for addresses that are the result of redirections.
5134 PH/48 A number of further additions for the benefit of the new test suite,
5135 including a fake gethostbyname() that interfaces to the fake DNS resolver
5138 TF/06 The fix for widen_domains has also been applied to qualify_single and
5139 search_parents which are the other dnslookup options that can cause
5142 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5145 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5147 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5148 service" if it encountered an MX record with an empty target hostname.
5149 The message is now "an MX or SRV record indicated no SMTP service".
5151 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5152 if the DNS of the sending site is misconfigured. This is quite a
5153 common situation. This change restores the behaviour of treating a
5154 helo verification defer as a failure.
5156 PH/51 If self=fail was set on a router, the bounce message did not include the
5157 actual error message.
5163 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5165 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5166 command itself ended in error, the underlying message about the transport
5167 filter timeout was being overwritten with the pipe command error. Now the
5168 underlying error message should be appended to the second error message.
5170 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5172 PH/02 Reduce the amount of output that "make" produces by default. Full output
5173 can still be requested.
5175 PH/03 The warning log line about a condition test deferring for a "warn" verb
5176 was being output only once per connection, rather than after each
5177 occurrence (because it was using the same function as for successful
5178 "warn" verbs). This seems wrong, so I have changed it.
5180 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5181 it should not have, which might have caused a crash in the right
5182 circumstances, but probably never did.
5184 PH/04 Installed a modified version of Tony Finch's patch to make submission
5185 mode fix the return path as well as the Sender: header line, and to
5186 add a /name= option so that you can make the user's friendly name appear
5189 TF/03 Added the control = fakedefer ACL modifier.
5191 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5192 Mark Lowes for thorough testing.
5194 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5196 TK/03 Merged latest SRS patch from Miles Wilton.
5198 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5199 with the definition in sysexits.h (which is #included earlier).
5200 Fortunately, Exim does not actually use EX_OK. The code used to try to
5201 preserve the sysexits.h value, by assuming that macro definitions were
5202 scanned for macro replacements. I have been disabused of this notion,
5203 so now the code just undefines EX_OK before #including unistd.h.
5205 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5206 in the smtp transport. When a block could not be written in a single
5207 write() function, the timeout was being re-applied to each part-write.
5208 This seems wrong - if the receiver was accepting one byte at a time it
5209 would take for ever. The timeout is now adjusted when this happens. It
5210 doesn't have to be particularly precise.
5212 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5213 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5215 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5216 without a domain if the check to <postmaster@domain> fails.
5218 SC/01 Eximstats: added -xls and the ability to specify output files
5219 (patch written by Frank Heydlauf).
5221 SC/02 Eximstats: use FileHandles for outputting results.
5223 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5225 SC/04 Eximstats: fixed display of large numbers with -nvr option
5227 SC/05 Eximstats: fixed merging of reports with empty tables.
5229 SC/06 Eximstats: added the -include_original_destination flag
5231 SC/07 Eximstats: removed tabs and trailing whitespace.
5233 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5235 TK/06 MBOX spool code: Add real "From " MBOX separator line
5236 so the .eml file is really in mbox format (even though
5237 most programs do not really care). Patch from Alex Miller.
5239 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5240 The latter is generated from $received_to and is only set if the
5241 message has one envelope recipient. SA can use these headers,
5242 obviously out-of-the-box. Patch from Alex Miller.
5244 PH/08 The ${def test on a variable was returning false if the variable's
5245 value was "0", contrary to what the specification has always said!
5246 The result should be true unless the variable is empty.
5248 PH/09 The syntax error of a character other than { following "${if
5249 def:variable_name" (after optional whitespace) was not being diagnosed.
5250 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5251 accidental colon was present, for example, could give incorrect results.
5253 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5254 result is used (not including appendfile, where other changes are about
5257 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5258 This involved changing a lot of size variables from int to off_t. It
5259 should work with maildirs and everything.
5261 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5262 spamd dying while we are connected to it.
5264 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5267 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5268 function for BDB 4.3.
5270 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5272 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5273 the sources for PCRE have been re-arranged and more files are now
5276 PH/15 The code I had for printing potentially long long variables in PH/11
5277 above was not the best (it lost precision). The length of off_t variables
5278 is now inspected at build time, and an appropriate printing format (%ld
5279 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5280 to be "long long int" or "long int". This is needed for the internal
5281 formatting function string_vformat().
5283 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5284 the configuration file to be ":syslog", then the script "guesses" where
5285 the logs files are, rather than using the compiled in default. In our
5286 case the guess is not the same as the compiled default, so the script
5287 suddenly stopped working when I started to use syslog. The patch checks
5288 to see if log_file_path is "". If so, it attempts to read it from exim
5289 with no configuration file to get the compiled in version, before it
5290 falls back to the previous guessing code."
5292 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5293 implementing BATV in an Exim configuration. See NewStuff for the gory
5296 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5297 Makefile that are specific to HP-UX.
5299 PH/18 If the "use_postmaster" option was set for a recipient callout together
5300 with the "random" option, the postmaster address was used as the MAIL
5301 FROM address for the random test, but not for the subsequent recipient
5302 test. It is now used for both.
5304 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5305 patch removes a few documentation additions to RFC 3028, because the
5306 latest draft now contains them. It adds the new en;ascii-case comparator
5307 and a new error check for 8bit text in MIME parts. Comparator and
5308 require names are now matched exactly. I enabled the subaddress
5309 extension, but it is not well tested yet (read: it works for me)."
5311 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5312 rework some of the code of TK/09 above to avoid the hardwired use of
5313 "%lld" and "long long". Replaced the call to snprintf() with a call to
5316 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5317 records point to non-existent hosts", "retry timeout exceeded", and
5318 "retry time not reached for any host after a long failure period".
5320 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5321 experimental DomainKeys support:
5323 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5324 (2) On an error such as an illegally used "control", the wrong name for
5325 the control was given.
5327 These problems did NOT occur unless DomainKeys support was compiled.
5329 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5331 PH/24 Added ${if match_ip condition.
5333 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5334 databases so that it will be absolutely obvious if a crash occurs in the
5335 DB library. This is a regular occurrence (often caused by mis-matched
5338 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5339 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5340 picked up on a user's system that detects such things. There doesn't seem
5341 to be a gcc warning option for this - only an attribute that has to be
5342 put on the function's prototype. It seems that in Fedora Core 4 they have
5343 set this on a number of new functions. No doubt there will be more in due
5346 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5347 specify a transport. However, if an address that was verified by such a
5348 router was the subject of a callout, Exim crashed because it tried to
5349 read the rcpt_include_affixes from the non-existent transport. Now it
5350 just assumes that the setting of that option is false. This bug was
5351 introduced by 4.51/PH/31.
5353 PH/28 Changed -d+all to exclude +memory, because that information is very
5354 rarely of interest, but it makes the output a lot bigger. People tend to
5355 do -d+all out of habit.
5357 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5358 code in os-type was giving problems when libc.so lives in lib64, like on
5361 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5362 aren't the modern standard, and it seems that some systems' include files
5363 don't always have them. Exim was already checking for some of the newer
5364 ones like T_AAAA, and defining it itself. I've added checks for all the
5365 record types that Exim uses.
5367 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5368 not automatically generating a new one, as it is supposed to. This
5369 prevented TLS from working. If the file did exist, but contained invalid
5370 data, a new version was generated, as expected. It was only the case of a
5371 non-existent file that was broken.
5373 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5374 with a change in libdomainkeys > 0.64.
5376 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5377 from DNS. If the selector record carries the flag, it now has
5378 precedence over the domain-wide flag.
5380 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5382 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5383 the use of a port name that isn't defined in /etc/services) occurred, the
5384 message was deferred as in a normal delivery, and thus remained on the
5385 spool, instead of being failed because of the mua_wrapper setting. This
5386 is now fixed, and I tidied up some of the mua_wrapper messages at the
5389 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5390 the messages in a hash of arrays rather than using individual hashes.
5391 This is a bit cleaner and results in dramatic memory savings, albeit
5392 at a slight CPU cost.
5394 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5395 as requested by Marc Sherman.
5397 SC/10 Eximstats: added histograms for user specified patterns as requested
5400 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5402 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5403 fopen() in the content-scanning modules that did not already have it.
5409 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5410 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5412 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5414 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5416 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5417 to test Sieve filters that use "vacation".
5419 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5420 that changes the way the GnuTLS parameters are stored in the cache file.
5421 The new format can be generated externally. For backward compatibility,
5422 if the data in the cache doesn't make sense, Exim assumes it has read an
5423 old-format file, and it generates new data and writes a new file. This
5424 means that you can't go back to an older release without removing the
5427 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5428 work if there are any delivery delays because "one_time" forces the
5429 parent to be marked "delivered", so its unseen clone is never tried
5430 again. For this reason, Exim now forbids the simultaneous setting of
5433 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5434 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5435 entry if you want to know the details. The fix, however, neglected to
5436 consider the case when local delivery batching is involved. The test for
5437 "previously delivered" was not happening when checking to see if an
5438 address could be batched with a previous (undelivered) one; under
5439 certain circumstances this could lead to multiple deliveries to the same
5442 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5443 in its include files, and this causes problems building Exim.
5445 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5446 header_syntax) but Exim was just ignoring anything given after a slash.
5447 In particular, this caused confusion with an attempt to use "verify =
5448 reverse_host_lookup/defer_ok". An error is now given when options are
5449 supplied for verify items that do not have them. (Maybe reverse_host_
5450 lookup should have a defer_ok option, but that's a different point.)
5452 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5453 defined by RFC 821) to 2048, because there were problems with some AUTH
5454 commands, and RFC 1869 says the size should be increased for extended
5455 SMTP commands that take arguments.
5457 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5460 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5461 "unknown" error; now it says that the functionality isn't in the binary.
5463 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5464 an address' error message when a string expansion fails (syntax or
5465 whatever). Otherwise the password may appear in the log. Following change
5466 PH/42 below, there is no longer a chance of it appearing in a bounce
5469 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5471 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5472 of its addresses was ever tried. (Bugzilla bug #2.)
5474 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5475 the result incorrectly in the debug output. (It correctly added a newline
5476 to what was transported.)
5478 TF/01 Added $received_time.
5480 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5481 commented out examples of how to interface to a virus scanner and to
5482 SpamAssassin. Also added commented examples of av_scanner and
5483 spamd_address settings.
5485 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5486 and controls are allowed in which ACLs. There were a couple of minor
5487 errors. Some of the entries in the conditions table (which is a table of
5488 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5489 negation of where the condition IS allowed.
5491 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5493 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5494 header file does not have a version number, so I've had to invent a new
5495 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5496 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5497 radiusclient), but it was contributed by a Radius user.
5499 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5500 files or directories, for OpenSSL.
5502 PH/19 When an Exim process that is running as root has to create an Exim log
5503 file, it does so in a subprocess that runs as exim:exim so as to get the
5504 ownership right at creation (otherwise, other Exim processes might see
5505 the file with the wrong ownership). There was no test for failure of this
5506 fork() call, which would lead to the process getting stuck as it waited
5507 for a non-existent subprocess. Forks do occasionally fail when resources
5508 run out. I reviewed all the other calls to fork(); they all seem to check
5511 PH/20 When checking for unexpected SMTP input at connect time (before writing
5512 the banner), Exim was not dealing correctly with a non-positive return
5513 from the read() function. If the client had disconnected by this time,
5514 the result was a log entry for a synchronization error with an empty
5515 string after "input=" when read() returned zero. If read() returned -1
5516 (an event I could not check), uninitialized data bytes were printed.
5517 There were reports of junk text (parts of files, etc) appearing after
5520 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5522 PH/22 Added support for macro redefinition, and (re)definition in between
5523 driver and ACL definitions.
5525 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5526 forgetting to use the resulting value; it was using the unexpanded value.
5528 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5529 hadn't been configured. The fix is from Juergen Kreileder, who
5530 understands it better than I do:
5532 "Here's what I see happening with three configured cyrus_sasl
5533 authenticators configured (plain, login, cram-md5):
5535 On startup auth_cyrus_sasl_init() gets called for each of these.
5536 This means three calls to sasl_listmech() without a specified mech_list.
5537 => SASL tests which mechs of all available mechs actually work
5538 => three warnings about OTP not working
5539 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5541 With the patch, sasl_listmech() also gets called three times. But now
5542 SASL's mech_list option is set to the server_mech specified in the the
5543 authenticator. Or in other words, the answer from sasl_listmech()
5544 gets limited to just the mech you're testing for (which is different
5546 => the return list contains just 'plain' or 'login', 'cram-md5' or
5547 nothing depending on the value of ob->server_mech.
5549 I've just tested the patch: Authentication still works fine,
5550 unavailable mechs specified in the exim configuration are still
5551 caught, and the auth.log warnings about OTP are gone."
5553 PH/25 When debugging is enabled, the contents of the command line are added
5554 to the debugging output, even when log_selector=+arguments is not
5557 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5558 answer is "GNU", and only if the return is "GNU/something" is the answer
5561 PH/27 $acl_verify_message is now set immediately after the failure of a
5562 verification in an ACL, and so is available in subsequent modifiers. In
5563 particular, the message can be preserved by coding like this:
5565 warn !verify = sender
5566 set acl_m0 = $acl_verify_message
5568 Previously, $acl_verify_message was set only while expanding "message"
5569 and "log_message" when a very denied access.
5571 PH/28 Modified OS/os.c-Linux with
5573 -#ifndef OS_LOAD_AVERAGE
5574 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5576 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5577 nomenclature these days.)
5579 PH/29 Installed patch from the Sieve maintainer that adds the options
5580 sieve_useraddress and sieve_subaddress to the redirect router.
5582 PH/30 In these circumstances:
5583 . Two addresses routed to the same list of hosts;
5584 . First host does not offer TLS;
5585 . First host accepts first address;
5586 . First host gives temporary error to second address;
5587 . Second host offers TLS and a TLS session is established;
5588 . Second host accepts second address.
5589 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5590 and peerdn, if requested) that were in fact used only for the second
5593 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5594 attention to any local part prefix or suffix that was matched by the
5595 router that accepted the address. It now behaves in the same way as it
5596 does for delivery: the affixes are removed from the local part unless
5597 rcpt_include_affixes is set on the transport.
5599 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5600 timeout during the DATA phase of an incoming message.
5602 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5603 applied a patch sanctioned by the Sieve maintainer.
5605 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5606 the uid or gid is negative. A case of a negative gid caused this to be
5607 noticed. The fix allows for either to be negative.
5609 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5610 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5613 PH/36 The change PH/12 above was broken. Fixed it.
5615 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5616 the grounds that routing the same address twice would always produce the
5617 same answer. This might have been true once, but it is certainly no
5618 longer true now. Routing a child address may depend on the previous
5619 routing that produced that child. Some complicated redirection strategies
5620 went wrong when messages had multiple recipients, and made Exim's
5621 behaviour dependent on the order in which the addresses were given.
5623 I have moved the duplicate checking until after the routing is complete.
5624 Exim scans the addresses that are assigned to local and remote
5625 transports, and removes any duplicates. This means that more work will be
5626 done, as duplicates will always all be routed, but duplicates are
5627 presumably rare, so I don't expect this is of any significance.
5629 For deliveries to pipes, files, and autoreplies, the duplicate checking
5630 still happens during the routing process, since they are not going to be
5633 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5634 It corrects a timeout issue with spamd. This is Ian's comment: "The
5635 background is that sometimes spamd either never reads data from a
5636 connection it has accepted, or it never writes response data. The exiscan
5637 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5638 blindly assumes that writes won't block so it may never time out."
5640 PH/39 Allow G after quota size as well as K and M.
5642 PH/40 The value set for $authenticated_id in an authenticator may not contain
5643 binary zeroes or newlines because the value is written to log lines and
5644 to spool files. There was no check on this. Now the value is run through
5645 the string_printing() function so that such characters are converted to
5646 printable escape sequences.
5648 PH/41 $message_linecount is a new variable that contains the total number of
5649 lines in the message. Compare $body_linecount, which is the count for the
5652 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5653 bounce and delay warning messages, except in certain special cases, which
5656 (a) An SMTP error message from a remote host;
5657 (b) A message specified in a :fail: redirection;
5658 (c) A message specified in a "fail" command in a system filter;
5659 (d) A message specified in a FAIL return from the queryprogram router;
5660 (e) A message specified by the cannot_route_message router option.
5662 In these cases only, Exim does include the error details in bounce and
5663 warning messages. There are also a few cases where bland messages such
5664 as "unrouteable address" or "local delivery error" are given.
5666 PH/43 $value is now also set for the "else" part of a ${run expansion.
5668 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5669 being worked on, but at least Exim now implements the latest version to
5672 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5673 process to complete was treated as a delivery failure, a timeout while
5674 writing the message to the pipe was logged, but erroneously treated as a
5675 successful delivery. Such timeouts include transport filter timeouts. For
5676 consistency with the overall process timeout, these timeouts are now
5677 treated as errors, giving rise to delivery failures by default. However,
5678 there is now a new Boolean option for the pipe transport called
5679 timeout_defer, which, if set TRUE, converts the failures into defers for
5680 both kinds of timeout. A transport filter timeout is now identified in
5683 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5684 systems where "make" and "gmake" are different, calling "gmake" at top
5685 level broke things. I've arranged for the value of $(MAKE) to be passed
5686 from the Makefile to this script so that it can call the same version of
5690 A note about Exim versions 4.44 and 4.50
5691 ----------------------------------------
5693 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5694 changes of various kinds. As a consequence, a big documentation update was
5695 needed. This delayed the release for rather longer than seemed good, especially
5696 in the light of a couple of (minor) security issues. Therefore, the changes
5697 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5698 release. So 4.44 and 4.50 are in effect two different branches that both start
5701 I have left the 4.50 change log unchanged; it contains all the changes since
5702 4.43. The change log for 4.44 is below; many of its items are identical to
5703 those for 4.50. This seems to be the most sensible way to preserve the
5704 historical information.
5710 1. Minor wording change to the doc/README.SIEVE file.
5712 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5713 computation of the current number of files was incorrect.
5715 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5716 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5719 4. Give more explanation in the error message when the command for a transport
5720 filter fails to execute.
5722 5. There are several places where Exim runs a non-Exim command in a
5723 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5724 was being done only for the command run by the queryprogram router. It is
5725 now done for all such subprocesses. The other cases are: ${run, transport
5726 filters, and the commands run by the lmtp and pipe transports.
5728 6. Added CONFIGURE_GROUP build-time option.
5730 7. Some older OS have a limit of 256 on the maximum number of file
5731 descriptors. Exim was using setrlimit() to set 1000 as a large value
5732 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5733 systems. I've change it so that if it can't get 1000, it tries for 256.
5735 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5736 was an oversight, and furthermore, ever since the addition of extra
5737 controls (e.g. 4.43/32), the checks on when to allow different forms of
5738 "control" were broken. There should now be diagnostics for all cases when a
5739 control that does not make sense is encountered.
5741 9. Added the /retain_sender option to "control=submission".
5743 10. $recipients is now available in the predata ACL (oversight).
5745 11. Tidy the search cache before the fork to do a delivery from a message
5746 received from the command line. Otherwise the child will trigger a lookup
5747 failure and thereby defer the delivery if it tries to use (for example) a
5748 cached ldap connection that the parent has called unbind on.
5750 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5751 of $address_data from the recipient verification was clobbered by the
5752 sender verification.
5754 13. The value of address_data from a sender verification is now available in
5755 $sender_address_data in subsequent conditions in the ACL statement.
5757 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5759 15. Added a new option "connect=<time>" to callout options, to set a different
5762 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5763 was its contents. (It was OK if the option was not defined at all.)
5765 17. A "Completed" log line is now written for messages that are removed from
5766 the spool by the -Mrm option.
5768 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5769 information about exactly what failed.
5771 19. Added -dd to debug only the daemon process.
5773 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5774 handles timeouts, both on the server side and network timeouts. Renamed the
5775 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5777 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5778 It is now set to "smtps".
5780 22. $host_address is now set to the target address during the checking of
5781 ignore_target_hosts.
5783 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5784 being passed; this would have caused $sender_host_name to have been used if
5785 matching the list had actually called for a host name (not very likely,
5786 since this list is usually IP addresses). A host name is now passed as
5789 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5790 code that specifies a non-restarting handler (typically sigaction() in
5791 modern systems) in an attempt to fix a rare and obscure crash bug.
5793 25. Narrowed the window for a race in the daemon that could cause it to ignore
5794 SIGCHLD signals. This is not a major problem, because they are used only to
5795 wake it up if nothing else does.
5797 26. A malformed maildirsize file could cause Exim to calculate negative values
5798 for the mailbox size or file count. Odd effects could occur as a result.
5799 The maildirsize information is now recalculated if the size or filecount
5802 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5803 support for a long time. Removed HAVE_SYS_VFS_H.
5805 28. Installed the latest version of exipick from John Jetmore.
5807 29. In an address list, if the pattern was not a regular expression, an empty
5808 subject address (from a bounce message) matched only if the pattern was an
5809 empty string. Non-empty patterns were not even tested. This was the wrong
5810 because it is perfectly reasonable to use an empty address as part of a
5811 database query. An empty address is now tested by patterns that are
5812 lookups. However, all the other forms of pattern expect the subject to
5813 contain a local part and a domain, and therefore, for them, an empty
5814 address still always fails if the pattern is not itself empty.
5816 30. Exim went into a mad DNS loop when attempting to do a callout where the
5817 host was specified on an smtp transport, and looking it up yielded more
5818 than one IP address.
5820 31. Re-factored the code for checking spool and log partition space into a
5821 function that finds that data and another that does the check. The former
5822 is then used to implement four new variables: $spool_space, $log_space,
5823 $spool_inodes, and $log_inodes.
5825 32. The RFC2047 encoding function was originally intended for short strings
5826 such as real names; it was not keeping to the 75-character limit for
5827 encoded words that the RFC imposes. It now respects the limit, and
5828 generates multiple encoded words if necessary. To be on the safe side, I
5829 have increased the buffer size for the ${rfc2047: expansion operator from
5832 33. It is now permitted to omit both strings after an "if" condition; if the
5833 condition is true, the result is "true". As before, when the second string
5834 is omitted, a false condition yields an empty string. This makes it less
5835 cumbersome to write custom ACL and router conditions.
5837 34. Failure to deliver a bounce message always caused it to be frozen, even if
5838 there was an errors_to setting on the router. The errors_to setting is now
5841 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5842 canonical form (fully expanded) before being placed in
5843 $sender_host_address.
5845 36. The table in the code that translates DNS record types into text (T_A to
5846 "A" for instance) was missing entries for NS and CNAME. It is just possible
5847 that this could have caused confusion if both these types were looked up
5848 for the same domain, because the text type is used as part of Exim's
5849 per-process caching. But the chance of anyone hitting this buglet seems
5852 37. The dnsdb lookup has been extended in a number of ways.
5854 (1) There is a new type, "zns", which walks up the domain tree until it
5855 finds some nameserver records. It should be used with care.
5857 (2) There is a new type, "mxh", which is like "mx" except that it returns
5858 just the host names, not the priorities.
5860 (3) It is now possible to give a list of domains (or IP addresses) to be
5861 looked up. The behaviour when one of the lookups defers can be
5862 controlled by a keyword.
5864 (4) It is now possible to specify the separator character for use when
5865 multiple records are returned.
5867 38. The dnslists ACL condition has been extended: it is now possible to supply
5868 a list of IP addresses and/or domains to be looked up in a particular DNS
5871 39. Added log_selector=+queue_time_overall.
5873 40. When running the queue in the test harness, wait just a tad after forking a
5874 delivery process, to get repeatability of debugging output.
5876 41. Include certificate and key file names in error message when GnuTLS fails
5877 to set them up, because the GnuTLS error message doesn't include the name
5878 of the failing file when there is a problem reading it.
5880 42. Allow both -bf and -bF in the same test run.
5882 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5884 44. The "Exiscan patch" is now merged into the mainline Exim source.
5886 45. Sometimes the final signoff response after QUIT could fail to get
5887 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5888 < 0 before doing a fflush(). This bug looks as though it goes back to the
5889 introduction of TLS in release 3.20, but "sometimes" must have been rare
5890 because the tests only now provoked it.
5892 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5893 (this can affect the format of dates).
5895 47. exim_tidydb, when checking for the continued existence of a message for
5896 which it has found a message-specific retry record, was not finding
5897 messages that were in split spool directories. Consequently, it was
5898 deleting retry records that should have stayed in existence.
5900 48. Steve fixed some bugs in eximstats.
5902 49. The SPA authentication driver was not abandoning authentication and moving
5903 on to the next authenticator when an expansion was forced to fail,
5904 contradicting the general specification for all authenticators. Instead it
5905 was generating a temporary error. It now behaves as specified.
5907 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5908 (the order specifies the preference for clients). The order is now AES256,
5909 AES128, 3DES, ARCFOUR128.
5911 51. Small patch to Sieve code - explicitly set From: when generating an
5914 52. Exim crashed if a remote delivery caused a very long error message to be
5915 recorded - for instance if somebody sent an entire SpamAssassin report back
5916 as a large number of 550 error lines. This bug was coincidentally fixed by
5917 increasing the size of one of Exim's internal buffers (big_buffer) that
5918 happened as part of the Exiscan merge. However, to be on the safe side, I
5919 have made the code more robust (and fixed the comments that describe what
5922 53. Now that there can be additional text after "Completed" in log lines (if
5923 the queue_time_overall log selector is set), a one-byte patch to exigrep
5924 was needed to allow it to recognize "Completed" as not the last thing in
5927 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5928 patch that reportedly fixes this has been added. I am not expert enough to
5929 create a test for it. This is what the patch creator wrote:
5931 "I found a little strange behaviour of ldap code when working with
5932 Windows 2003 AD Domain, where users was placed in more than one
5933 Organization Units. When I tried to give exim partial DN, the exit code
5934 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5935 But simultaneously result of request was absolutely normal ldap result,
5936 so I produce this patch..."
5938 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5939 REFERENCE, so I have modified the code to exclude the patch when that macro
5942 55. Some experimental protocols are using DNS PTR records for new purposes. The
5943 keys for these records are domain names, not reversed IP addresses. The
5944 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5945 leaves it alone. Component reversal etc. now happens only for IP addresses.
5948 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5950 57. Double the size of the debug message buffer (to 2048) so that more of very
5951 long debug lines gets shown.
5953 58. The exicyclog utility now does better if the number of log files to keep
5954 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5956 59. Two changes related to the smtp_active_hostname option:
5958 (1) $smtp_active_hostname is now available as a variable.
5959 (2) The default for smtp_banner uses $smtp_active_hostname instead
5960 of $primary_hostname.
5962 60. The host_aton() function is supposed to be passed a string that is known
5963 to be a valid IP address. However, in the case of IPv6 addresses, it was
5964 not checking this. This is a hostage to fortune. Exim now panics and dies
5965 if the condition is not met. A case was found where this could be provoked
5966 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5967 components; fortuitously, this particular loophole had already been fixed
5968 by change 4.50/55 above.
5970 If there are any other similar loopholes, the new check in host_aton()
5971 itself should stop them being exploited. The report I received stated that
5972 data on the command line could provoke the exploit when Exim was running as
5973 exim, but did not say which command line option was involved. All I could
5974 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5975 running as the user.
5978 61. There was a buffer overflow vulnerability in the SPA authentication code
5979 (which came originally from the Samba project). I have added a test to the
5980 spa_base64_to_bits() function which I hope fixes it.
5983 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5984 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5986 63. The daemon start-up calls getloadavg() while still root for those OS that
5987 need the first call to be done as root, but it missed one case: when
5988 deliver_queue_load_max is set with deliver_drop_privilege. This is
5989 necessary for the benefit of the queue runner, because there is no re-exec
5990 when deliver_drop_privilege is set.
5992 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5993 This has been fixed.
5995 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5996 was in use, was not putting the data itself into the right store pool;
5997 consequently, it could be overwritten for a subsequent message in the same
5998 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6001 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6003 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6004 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6005 the function were treating the return as a boolean value, which happened to
6006 work because 0=false and not-0=true, but is not correct code.
6008 68. The host_aton() function was not handling scoped IPv6 addresses (those
6009 with, for example, "%eth0" on the end) correctly.
6011 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6012 negated items (that is, ~something) in unsigned ints. Some compilers
6013 apparently mutter when there is no cast.
6015 70. If an address verification called from an ACL failed, and did not produce a
6016 user-specific message (i.e. there was only a "system" message), nothing was
6017 put in $acl_verify_message. In this situation, it now puts the system
6020 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6021 session; change 4.31/43 added the unwanted input to the log line - except
6022 that it did not do this in the start of session case. It now does.
6024 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6025 This is wrong and can cause the other end to generate a synchronization
6026 error if it is another Exim or anything else that does the synchronization
6027 check. A QUIT command is no longer sent after a timeout.
6029 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6030 during host lookups.
6032 74. The defer_ok option of callout verification was not working if it was used
6033 when verifying addresses in header lines, that is, for this case:
6035 verify = header_sender/callout=defer_ok
6037 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6038 those file descriptors could be used for SMTP connections. If anything
6039 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6040 could be sent to the SMTP client, causing chaos. The daemon now opens
6041 stdin, stdout, and stderr to /dev/null when it puts itself into the
6044 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6045 log by default. The user can override this with suitable Perl magic.
6047 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6048 the log message when discard triggers, was not working for the DATA ACL or
6049 for the non-SMTP ACL.
6051 78. Error message wording change in sieve.c.
6053 79. If smtp_accept_max_per_host was set, the number of connections could be
6054 restricted to fewer than expected, because the daemon was trying to set up
6055 a new connection before checking whether the processes handling previous
6056 connections had finished. The check for completed processes is now done
6057 earlier. On busy systems, this bug wouldn't be noticed because something
6058 else would have woken the daemon, and it would have reaped the completed
6061 80. If a message was submitted locally by a user whose login name contained one
6062 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6063 It caused a spool format error. I have fixed the spool reading code. A
6064 related problem was that the "from" clause in the Received: line became
6065 illegal because of the space(s). It is now covered by ${quote_local_part.
6067 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6070 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6072 83. Patch from Sieve maintainer to fix unterminated string problem in
6073 "vacation" handling.
6075 84. Some minor changes to the Linux configuration files to help with other
6076 OS variants using glibc.
6078 85. One more patch for Sieve to update vacation handling to latest spec.
6081 ----------------------------------------------------
6082 See the note above about the 4.44 and 4.50 releases.
6083 ----------------------------------------------------
6089 1. Change 4.43/35 introduced a bug that caused file counts to be
6090 incorrectly computed when quota_filecount was set in an appendfile
6093 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6094 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6097 3. Give more explanation in the error message when the command for a transport
6098 filter fails to execute.
6100 4. There are several places where Exim runs a non-Exim command in a
6101 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6102 was being done only for the command run by the queryprogram router. It is
6103 now done for all such subprocesses. The other cases are: ${run, transport
6104 filters, and the commands run by the lmtp and pipe transports.
6106 5. Some older OS have a limit of 256 on the maximum number of file
6107 descriptors. Exim was using setrlimit() to set 1000 as a large value
6108 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6109 systems. I've change it so that if it can't get 1000, it tries for 256.
6111 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6112 was an oversight, and furthermore, ever since the addition of extra
6113 controls (e.g. 4.43/32), the checks on when to allow different forms of
6114 "control" were broken. There should now be diagnostics for all cases when a
6115 control that does not make sense is encountered.
6117 7. $recipients is now available in the predata ACL (oversight).
6119 8. Tidy the search cache before the fork to do a delivery from a message
6120 received from the command line. Otherwise the child will trigger a lookup
6121 failure and thereby defer the delivery if it tries to use (for example) a
6122 cached ldap connection that the parent has called unbind on.
6124 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6125 of $address_data from the recipient verification was clobbered by the
6126 sender verification.
6128 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6129 was its contents. (It was OK if the option was not defined at all.)
6131 11. A "Completed" log line is now written for messages that are removed from
6132 the spool by the -Mrm option.
6134 12. $host_address is now set to the target address during the checking of
6135 ignore_target_hosts.
6137 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6138 being passed; this would have caused $sender_host_name to have been used if
6139 matching the list had actually called for a host name (not very likely,
6140 since this list is usually IP addresses). A host name is now passed as
6143 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6144 code that specifies a non-restarting handler (typically sigaction() in
6145 modern systems) in an attempt to fix a rare and obscure crash bug.
6147 15. Narrowed the window for a race in the daemon that could cause it to ignore
6148 SIGCHLD signals. This is not a major problem, because they are used only to
6149 wake it up if nothing else does.
6151 16. A malformed maildirsize file could cause Exim to calculate negative values
6152 for the mailbox size or file count. Odd effects could occur as a result.
6153 The maildirsize information is now recalculated if the size or filecount
6156 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6157 support for a long time. Removed HAVE_SYS_VFS_H.
6159 18. Updated exipick to current release from John Jetmore.
6161 19. Allow an empty sender to be matched against a lookup in an address list.
6162 Previously the only cases considered were a regular expression, or an
6165 20. Exim went into a mad DNS lookup loop when doing a callout where the
6166 host was specified on the transport, if the DNS lookup yielded more than
6169 21. The RFC2047 encoding function was originally intended for short strings
6170 such as real names; it was not keeping to the 75-character limit for
6171 encoded words that the RFC imposes. It now respects the limit, and
6172 generates multiple encoded words if necessary. To be on the safe side, I
6173 have increased the buffer size for the ${rfc2047: expansion operator from
6176 22. Failure to deliver a bounce message always caused it to be frozen, even if
6177 there was an errors_to setting on the router. The errors_to setting is now
6180 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6181 canonical form (fully expanded) before being placed in
6182 $sender_host_address.
6184 24. Updated eximstats to version 1.33
6186 25. Include certificate and key file names in error message when GnuTLS fails
6187 to set them up, because the GnuTLS error message doesn't include the name
6188 of the failing file when there is a problem reading it.
6190 26. Expand error message when OpenSSL has problems setting up cert/key files.
6193 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6194 (this can affect the format of dates).
6196 28. exim_tidydb, when checking for the continued existence of a message for
6197 which it has found a message-specific retry record, was not finding
6198 messages that were in split spool directories. Consequently, it was
6199 deleting retry records that should have stayed in existence.
6201 29. eximstats updated to version 1.35
6202 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6203 1.35 - bugfix such that pie charts by volume are generated correctly
6205 30. The SPA authentication driver was not abandoning authentication and moving
6206 on to the next authenticator when an expansion was forced to fail,
6207 contradicting the general specification for all authenticators. Instead it
6208 was generating a temporary error. It now behaves as specified.
6210 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6211 (the order specifies the preference for clients). The order is now AES256,
6212 AES128, 3DES, ARCFOUR128.
6214 31. Small patch to Sieve code - explicitly set From: when generating an
6217 32. Exim crashed if a remote delivery caused a very long error message to be
6218 recorded - for instance if somebody sent an entire SpamAssassin report back
6219 as a large number of 550 error lines. This bug was coincidentally fixed by
6220 increasing the size of one of Exim's internal buffers (big_buffer) that
6221 happened as part of the Exiscan merge. However, to be on the safe side, I
6222 have made the code more robust (and fixed the comments that describe what
6225 33. Some experimental protocols are using DNS PTR records for new purposes. The
6226 keys for these records are domain names, not reversed IP addresses. The
6227 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6228 leaves it alone. Component reversal etc. now happens only for IP addresses.
6231 34. The host_aton() function is supposed to be passed a string that is known
6232 to be a valid IP address. However, in the case of IPv6 addresses, it was
6233 not checking this. This is a hostage to fortune. Exim now panics and dies
6234 if the condition is not met. A case was found where this could be provoked
6235 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6236 components; fortuitously, this particular loophole had already been fixed
6237 by change 4.50/55 or 4.44/33 above.
6239 If there are any other similar loopholes, the new check in host_aton()
6240 itself should stop them being exploited. The report I received stated that
6241 data on the command line could provoke the exploit when Exim was running as
6242 exim, but did not say which command line option was involved. All I could
6243 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6244 running as the user.
6247 35. There was a buffer overflow vulnerability in the SPA authentication code
6248 (which came originally from the Samba project). I have added a test to the
6249 spa_base64_to_bits() function which I hope fixes it.
6252 36. The daemon start-up calls getloadavg() while still root for those OS that
6253 need the first call to be done as root, but it missed one case: when
6254 deliver_queue_load_max is set with deliver_drop_privilege. This is
6255 necessary for the benefit of the queue runner, because there is no re-exec
6256 when deliver_drop_privilege is set.
6258 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6259 was in use, was not putting the data itself into the right store pool;
6260 consequently, it could be overwritten for a subsequent message in the same
6261 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6264 38. Sometimes the final signoff response after QUIT could fail to get
6265 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6266 < 0 before doing a fflush(). This bug looks as though it goes back to the
6267 introduction of TLS in release 3.20, but "sometimes" must have been rare
6268 because the tests only now provoked it.
6274 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6275 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6276 Now it returns an integer. A number of calls were still expecting a T/F
6277 return. Fortuitously, in all cases, the tests worked in OK situations,
6278 which is the norm. However, things would have gone wrong on any write
6279 failures on the smtp file descriptor. This function is used when sending
6280 messages over SMTP and also when doing verify callouts.
6282 2. When Exim is called to do synchronous delivery of a locally submitted
6283 message (the -odf or -odi options), it no longer closes stderr before doing
6286 3. Implemented the mua_wrapper option.
6288 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6290 5. Implemented the functions header_remove(), header_testname(),
6291 header_add_at_position(), and receive_remove_recipient(), and exported them
6294 6. If an ACL "warn" statement specified the addition of headers, Exim already
6295 inserted X-ACL-Warn: at the start if there was no header name. However, it
6296 was not making this test for the second and subsequent header lines if
6297 there were newlines in the string. This meant that an invalid header could
6298 be inserted if Exim was badly configured.
6300 7. Allow an ACL "warn" statement to add header lines at the start or after all
6301 the Received: headers, as well as at the end.
6303 8. Added the rcpt_4xx retry error code.
6305 9. Added postmaster_mailfrom=xxx to callout verification option.
6307 10. Added mailfrom=xxxx to the callout verification option, for verify=
6310 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6311 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6313 12. Inserted some casts to stop certain compilers warning when using pointer
6314 differences as field lengths or precisions in printf-type calls (mostly
6315 affecting debugging statements).
6317 13. Added optional readline() support for -be (dynamically loaded).
6319 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6320 same clock tick as a message's arrival, so that its received time was the
6321 same as the "first fail" time on the retry record, and that message
6322 remained on the queue past the ultimate address timeout, every queue runner
6323 would try a delivery (because it was past the ultimate address timeout) but
6324 after another failure, the ultimate address timeout, which should have then
6325 bounced the address, did not kick in. This was a "< instead of <=" error;
6326 in most cases the first failure would have been in the next clock tick
6327 after the received time, and all would be well.
6329 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6330 being recognized when the domain list was tested by the match_domain
6331 condition in an expansion string.
6333 16. Added the ${str2b64: operator.
6335 17. Exim was always calling setrlimit() to set a large limit for the number of
6336 processes, without checking whether the existing limit was already
6337 adequate. (It did check for the limit on file descriptors.) Furthermore,
6338 errors from getrlimit() and setrlimit() were being ignored. Now they are
6339 logged to the main and panic logs, but Exim does carry on, to try to do its
6340 job under whatever limits there are.
6342 18. Imported PCRE 5.0.
6344 19. Trivial typo in log message " temporarily refused connection" (the leading
6347 20. If the log selector return_path_on_delivery was set and an address was
6348 redirected to /dev/null, the delivery process crashed because it assumed
6349 that a return path would always be set for a "successful" delivery. In this
6350 case, the whole delivery is bypassed as an optimization, and therefore no
6353 21. Internal re-arrangement: the function for sending a challenge and reading
6354 a response while authentication was assuming a zero-terminated challenge
6355 string. It's now changed to take a pointer and a length, to allow for
6356 binary data in such strings.
6358 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6360 23. Exim was not respecting finduser_retries when seeking the login of the
6361 uid under which it was called; it was always trying 10 times. (The default
6362 setting of finduser_retries is zero.) Also, it was sleeping after the final
6363 failure, which is pointless.
6365 24. Implemented tls_on_connect_ports.
6367 25. Implemented acl_smtp_predata.
6369 26. If the domain in control=submission is set empty, Exim assumes that the
6370 authenticated id is a complete email address when it generates From: or
6371 Sender: header lines.
6373 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6374 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6375 chown and chgrp in /bin and hostname in /usr/bin.
6377 28. Exim was keeping the "process log" file open after each use, just as it
6378 does for the main log. This opens the possibility of it remaining open for
6379 long periods when the USR1 signal hits a daemon. Occasional processlog
6380 errors were reported, that could have been caused by this. Anyway, it seems
6381 much more sensible not to leave this file open at all, so that is what now
6384 29. The long-running daemon process does not normally write to the log once it
6385 has entered its main loop, and it closes the log before doing so. This is
6386 so that log files can straightforwardly be renamed and moved. However,
6387 there are a couple of unusual error situations where the daemon does write
6388 log entries, and I had neglected to close the log afterwards.
6390 30. The text of an SMTP error response that was received during a remote
6391 delivery was being truncated at 512 bytes. This is too short for some of
6392 the long messages that one sometimes sees. I've increased the limit to
6395 31. It is now possible to make retry rules that apply only when a message has a
6396 specific sender, in particular, an empty sender.
6398 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6399 it possible to be selective about when SMTP synchronization is enforced.
6401 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6403 32. Implemented hosts_connection_nolog.
6405 33. Added an ACL for QUIT.
6407 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6410 35. Added mailbox_size and mailbox_filecount to appendfile.
6412 36. Added control = no_multiline_responses to ACLs.
6414 37. There was a bug in the logic of the code that waits for the clock to tick
6415 in the case where the clock went backwards by a substantial amount such
6416 that the microsecond fraction of "now" was more than the microsecond
6417 fraction of "then" (but the whole seconds number was less).
6419 38. Added support for the libradius Radius client library this is found on
6420 FreeBSD (previously only the radiusclient library was supported).
6426 1. When certain lookups returned multiple values in the form name=value, the
6427 quoting of the values was not always being done properly. Specifically:
6428 (a) If the value started with a double quote, but contained no whitespace,
6430 (b) If the value contained whitespace other than a space character (i.e.
6431 tabs or newlines or carriage returns) it was not quoted.
6432 This fix has been applied to the mysql and pgsql lookups by writing a
6433 separate quoting function and calling it from the lookup code. The fix
6434 should probably also be applied to nisplus, ibase and oracle lookups, but
6435 since I cannot test any of those, I have not disturbed their existing code.
6437 2. A hit in the callout cache for a specific address caused a log line with no
6438 reason for rejecting RCPT. Now it says "Previous (cached) callout
6439 verification failure".
6441 3. There was an off-by-one bug in the queryprogram router. An over-long
6442 return line was truncated at 256 instead of 255 characters, thereby
6443 overflowing its buffer with the terminating zero. As well as fixing this, I
6444 have increased the buffer size to 1024 (and made a note to document this).
6446 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6447 when Exim is waiting for an SMTP response from a remote server, Exim
6448 restarts its select() call on the socket, thereby resetting its timeout.
6449 This is not a problem when such interrupts are rare. Somebody set up a cron
6450 job to run exiwhat every 2 minutes, which is less than the normal select()
6451 timeout (5 or 10 minutes). This meant that the select() timeout never
6452 kicked in because it was always reset. I have fixed this by comparing the
6453 time when an interrupt arrives with the time at the start of the first call
6454 to select(). If more time than the timeout has elapsed, the interrupt is
6455 treated as a timeout.
6457 5. Some internal re-factoring in preparation for the addition of Sieve
6458 extensions (by MH). In particular, the "personal" test is moved to a
6459 separate function, and given an option for scanning Cc: and Bcc: (which is
6460 not set for Exim filters).
6462 6. When Exim created an email address using the login of the caller as the
6463 local part (e.g. when creating a From: or Sender: header line), it was not
6464 quoting the local part when it contained special characters such as @.
6466 7. Installed new OpenBSD configuration files.
6468 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6469 try to make them clearer.
6471 9. Callout options, other than the timeout value, were being ignored when
6472 verifying sender addresses in header lines. For example, when using
6474 verify = header_sender/callout=no_cache
6476 the cache was (incorrectly) being used.
6478 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6479 only the Cygwin environment.
6481 11. When return_path_on_delivery was set as a log selector, if different remote
6482 addresses in the same message used different return paths and parallel
6483 remote delivery occurred, the wrong values would sometimes be logged.
6484 (Whenever a remote delivery process finished, the return path value from
6485 the most recently started remote delivery process was logged.)
6487 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6488 lines when AUTH and/or TLS are in use. This is the "received protocol"
6489 field. Exim used to use "asmtp" for authenticated SMTP, without any
6490 indication (in the protocol name) for TLS use. Now it follows the RFC and
6491 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6492 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6493 names appear in log lines as well as in Received: header lines.
6495 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6496 extensions, and comparison tests, and to fix some bugs.
6498 14. Changes to the "personal" filter test:
6500 (1) The test was buggy in that it was just doing the equivalent of
6501 "contains" tests on header lines. For example, if a user's address was
6502 anne@some.where, the "personal" test would incorrectly be true for
6504 To: susanne@some.where
6506 This test is now done by extracting each address from the header in turn,
6507 and checking the entire address. Other tests that are part of "personal"
6508 are now done using regular expressions (for example, to check local parts
6509 of addresses in From: header lines).
6511 (2) The list of non-personal local parts in From: addresses has been
6512 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6513 taken from the Sieve specification recommendations.
6515 (3) If the message contains any header line starting with "List-" it is
6516 treated as non-personal.
6518 (4) The test for "circular" in the Subject: header line has been removed
6519 because it now seems ill-conceived.
6521 15. Minor typos in src/EDITME comments corrected.
6523 16. Installed latest exipick from John Jetmore.
6525 17. If headers_add on a router specified a text string that was too long for
6526 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6527 of string_sprintf() is now avoided.
6529 18. $message_body_size was not set (it was always zero) when running the DATA
6530 ACL and the local_scan() function.
6532 19. For the "mail" command in an Exim filter, no default was being set for
6533 the once_repeat time, causing a random time value to be used if "once" was
6534 specified. (If the value happened to be <= 0, no repeat happened.) The
6535 default is now 0s, meaning "never repeat". The "vacation" command was OK
6536 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6537 (I found it when inspecting the code).
6539 20. There is now an overall timeout for performing a callout verification. It
6540 defaults to 4 times the callout timeout, which applies to individual SMTP
6541 commands during the callout. The overall timeout applies when there is more
6542 than one host that can be tried. The timeout is checked before trying the
6543 next host. This prevents very long delays if there are a large number of
6544 hosts and all are timing out (e.g. when the network connections are timing
6545 out). The value of the overall timeout can be changed by specifying an
6546 additional sub-option for "callout", called "maxwait". For example:
6548 verify = sender/callout=5s,maxwait=20s
6550 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6551 to the end before writing, but this should make it even safer).
6553 22. Exim was forgetting that it had advertised PIPELINING for the second and
6554 subsequent messages on an SMTP connection. It was also not resetting its
6555 memory on STARTTLS and an internal HELO.
6557 23. When Exim logs an SMTP synchronization error within a session, it now
6558 records whether PIPELINING has been advertised or not.
6560 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6561 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6562 rather than long int.
6564 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6566 26. Added the never_mail option to autoreply.
6572 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6573 crash if the getsockname() call failed; this can happen if a connection is
6574 closed very soon after it is established. The problem was simply in the
6575 order in which certain operations were done, causing Exim to try to write
6576 to the SMTP stream before it had set up the file descriptor. The bug has
6577 been fixed by making things happen in the correct order.
6583 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6584 before the connection was closed, thus losing the rejection response.
6586 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6587 some early Solaris releases, but causes trouble in current releases where
6588 socklen_t is defined.
6590 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6593 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6596 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6597 that releasing the top of it at the end releases what was used for sub-
6598 expansions (unless the block got too big). However, discard this block if
6599 the first thing is a variable or header, so that we can use its block when
6600 it is dynamic (useful for very large $message_headers, for example).
6602 6. Lookups now cache *every* query, not just the most recent. A new, separate
6603 store pool is used for this. It can be recovered when all lookup caches are
6604 flushed. Lookups now release memory at the end of their result strings.
6605 This has involved some general refactoring of the lookup sources.
6607 7. Some code has been added to the store_xxx() functions to reduce the amount
6608 of flapping under certain conditions.
6610 8. log_incoming_interface used to affect only the <= reception log lines. Now
6611 it causes the local interface and port to be added to several more SMTP log
6612 lines, for example "SMTP connection from", and rejection lines.
6614 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6616 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6618 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6619 could be overwritten at the end of the current message (or the start of a
6620 new message if it was set in a HELO ACL). The value is now preserved for
6621 the duration of the SMTP connection.
6623 12. If a transport had a headers_rewrite setting, and a matching header line
6624 contained an unqualified address, that address was qualified, even if it
6625 did not match any rewriting rules. The underlying bug was that the values
6626 of the flags that permit the existence of unqualified sender and recipient
6627 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6628 non-local messages, and by -bnq for local messages) were not being
6629 preserved with the message after it was received.
6631 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6632 "next input=" as part of the text comprising the host identity instead of
6633 the correct text. The code was using the same buffer for two different
6634 strings. However, depending on which order the printing function evaluated
6635 its arguments, the bug did not always show up. Under Linux, for example, my
6636 test suite worked just fine.
6638 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6639 doesn't work with some older versions of Perl. It has been changed to "my",
6640 which in any case is probably the better facility to use.
6642 15. A really picky compiler found some instances of statements for creating
6643 error messages that either had too many or two few arguments for the format
6646 16. The size of the buffer for calls to the DNS resolver has been increased
6647 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6648 for addresses that have a lot of PTR records. This alleviates a problem; it
6649 does not fully solve it.
6651 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6652 buffer now truncates the list and logs the incident, which is the same
6653 action as happens when Exim is looking up a host name and its aliases.
6654 Previously in this situation something unpredictable would happen;
6655 sometimes it was "internal error: store_reset failed".
6657 18. If a server dropped the connection unexpectedly when an Exim client was
6658 using GnuTLS and trying to read a response, the client delivery process
6659 crashed while trying to generate an error log message.
6661 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6662 string, for example:
6664 warn message = H1: something\nH2: something
6666 the text was added as a single header line from Exim's point of view
6667 though it ended up OK in the delivered message. However, searching for the
6668 second and subsequent header lines using $h_h2: did not work. This has been
6669 fixed. Similarly, if a system filter added multiple headers in this way,
6670 the routers could not see them.
6672 20. Expanded the error message when iplsearch is called with an invalid key to
6673 suggest using net-iplsearch in a host list.
6675 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6676 ACLs are no longer actually imposed (and a message to that effect is
6679 22. If a "gecos" field in a passwd entry contained escaped characters, in
6680 particular, if it contained a \" sequence, Exim got it wrong when building
6681 a From: or a Sender: header from that name. A second bug also caused
6682 incorrect handling when an unquoted " was present following a character
6683 that needed quoting.
6685 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6686 was not being matched caselessly.
6688 24. Arranged for all hyphens in the exim.8 source to be escaped with
6691 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6692 properly. Recipient callout cache records were still being keyed to include
6693 the sender, even when use_sender was set false. This led to far more
6694 callouts that were necessary. The sender is no longer included in the key
6695 when use_sender is false.
6697 26. Added "control = submission" modifier to ACLs.
6699 27. Added the ${base62d: operator to decode base 62 numbers.
6701 28. dnsdb lookups can now access SRV records.
6703 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6704 the configuration file.
6706 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6707 (-v) mode. This makes the output for a verbose queue run more intelligible.
6709 31. Added a use_postmaster feature to recipient callouts.
6711 32. Added the $body_zerocount variable, containing the number of binary zero
6712 bytes in the message body.
6714 33. The time of last modification of the "new" subdirectory is now used as the
6715 "mailbox time last read" when there is a quota error for a maildir
6718 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6720 35. Added +ignore_unknown as a special item in host lists.
6722 36. Code for decoding IPv6 addresses in host lists is now included, even if
6723 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6724 address was recognized as an IP address, but was then not correctly decoded
6725 into binary, causing unexpected and incorrect effects when compared with
6732 1. Very minor rewording of debugging text in manualroute to say "list of
6733 hosts" instead of "hostlist".
6735 2. If verify=header_syntax was set, and a header line with an unqualified
6736 address (no domain) and a large number of spaces between the end of the
6737 name and the colon was received, the reception process suffered a buffer
6738 overflow, and (when I tested it) crashed. This was caused by some obsolete
6739 code that should have been removed. The fix is to remove it!
6741 3. When running in the test harness, delay a bit after writing a bounce
6742 message to get a bit more predictability in the log output.
6744 4. Added a call to search_tidyup() just before forking a reception process. In
6745 theory, someone could use a lookup in the expansion of smtp_accept_max_
6746 per_host which, without the tidyup, could leave open a database connection.
6748 5. Added the variables $recipient_data and $sender_data which get set from a
6749 lookup success in an ACL "recipients" or "senders" condition, or a router
6750 "senders" option, similar to $domain_data and $local_part_data.
6752 6. Moved the writing of debug_print from before to after the "senders" test
6755 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6756 problems for message scanning, either using a data ACL, or using
6757 local_scan() because the Received: header was not generated till after they
6758 were called (in order to set the time as the time of reception completion).
6759 I have revised the way this works. The header is now generated after the
6760 body is received, but before the ACL or local_scan() are called. After they
6761 are run, the timestamp in the header is updated.
6767 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6768 before starting a queue runner without re-exec. This happened only when
6769 deliver_drop_privilege was set or when the Exim user was set to root. The
6770 effect of the bug was that timeouts during subsequent deliveries caused
6771 crashes instead of being properly handled. The handler is now left at its
6772 default (and expected) setting.
6774 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6775 message, again when deliver_drop_privilege is set or Exim is run as root.
6776 The bug described in (1) was not present in this case, but the tidying up
6777 of the other signals was missing. I have made the two cases consistent.
6779 3. The ignore_target_hosts setting on a manualroute router was being ignored
6780 for hosts that were looked up using the /MX notation.
6782 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6785 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6786 operated on the sender address. After changing the $sender_address to <>
6787 for the sender address verify, Exim was re-instated it as the original
6788 (before rewriting) address, but remembering that it had rewritten it, so it
6789 wasn't rewriting it again. This bug also had the effect of breaking the
6790 sender address verification caching when the sender address was rewritten.
6792 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6793 This has been changed so that if the ip literal address matches
6794 ignore_target_hosts, the router declines.
6796 7. Added expansion conditions match_domain, match_address, and match_local_
6797 part (NOT match_host).
6799 8. The placeholder for the Received: header didn't have a length field set.
6801 9. Added code to Exim itself and to exim_lock to test for a specific race
6802 condition that could lead to file corruption when using MBX delivery. The
6803 issue is with the lockfile that is created in /tmp. If this file is removed
6804 after a process has opened it but before that process has acquired a lock,
6805 there is the potential for a second process to recreate the file and also
6806 acquire a lock. This could lead to two Exim processes writing to the file
6807 at the same time. The added code performs the same test as UW imapd; it
6808 checks after acquiring the lock that its file descriptor still refers to
6809 the same named file.
6811 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6812 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6815 11. Added the smtp_active_hostname option. If used, this will typically be made
6816 to depend on the incoming interface address. Because $interface_address is
6817 not set up until the daemon has forked a reception process, error responses
6818 that can happen earlier (such as "too many connections") no longer contain
6821 12. If an expansion in a condition on a "warn" statement fails because a lookup
6822 defers, the "warn" statement is abandoned, and the next ACL statement is
6823 processed. Previously this caused the whole ACL to be aborted.
6825 13. Added the iplsearch lookup type.
6827 14. Added ident_timeout as a log selector.
6829 15. Added tls_certificate_verified as a log selector.
6831 16. Added a global option tls_require_ciphers (compare the smtp transport
6832 option of the same name). This controls incoming TLS connections.
6834 17. I finally figured out how to make tls_require_ciphers do a similar thing
6835 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6836 before starting the TLS session.
6838 18. Tabs are now shown as \t in -bP output.
6840 19. If the log selector return_path_on_delivery was set, Exim crashed when
6841 bouncing a message because it had too many Received: header lines.
6843 20. If two routers both had headers_remove settings, and the first one included
6844 a superfluous trailing colon, the final name in the first list and the
6845 first name in the second list were incorrectly joined into one item (with a
6846 colon in the middle).
6852 1. Added -C and -D options to the exinext utility, mainly to make it easier
6853 to include in the automated testing, but these could be helpful when
6854 multiple configurations are in use.
6856 2. The exinext utility was not formatting the output nicely when there was
6857 an alternate port involved in the retry record key, nor when there was a
6858 message id as well (for retries that were specific to a specific message
6859 and a specific host). It was also confused by IPv6 addresses, because of
6860 the additional colons they contain. I have fixed the IPv4 problem, and
6861 patched it up to do a reasonable job for IPv6.
6863 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6864 delivery, the log line now contains "pipelined" if PIPELINING was used.
6866 4. An SMTP transport process used to panic and die if the bind() call to set
6867 an explicit outgoing interface failed. This has been changed; it is now
6868 treated in the same way as a connect() failure.
6870 5. A reference to $sender_host_name in the part of a conditional expansion
6871 that was being skipped was still causing a DNS lookup. This no longer
6874 6. The def: expansion condition was not recognizing references to header lines
6875 that used bh_ and bheader_.
6877 7. Added the _cache feature to named lists.
6879 8. The code for checking quota_filecount in the appendfile transport was
6880 allowing one more file than it should have been.
6882 9. For compatibility with Sendmail, the command line option
6890 and sets the incoming protocol and host name (for trusted callers). The
6891 host name and its colon can be omitted when only the protocol is to be set.
6892 Note the Exim already has two private options, -pd and -ps, that refer to
6893 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6894 "s", but I don't think that's a major issue.
6896 10. A number of refactoring changes to the code, none of which should affect
6899 (a) The number of logging options was getting close to filling up the
6900 32-bit word that was used as a bit map. I have split them into two classes:
6901 those that are passed in the argument to log_write(), and those that are
6902 only ever tested independently outside of that function. These are now in
6903 separate 32-bit words, so there is plenty of room for expansion again.
6904 There is no change in the user interface or the logging behaviour.
6906 (b) When building, for example, log lines, the code previously used a
6907 macro that called string_cat() twice, in order to add two strings. This is
6908 not really sufficiently general. Furthermore, there was one instance where
6909 it was actually wrong because one of the argument was used twice, and in
6910 one call a function was used. (As it happened, calling the function twice
6911 did not affect the overall behaviour.) The macro has been replaced by a
6912 function that can join an arbitrary number of extra strings onto a growing
6915 (c) The code for expansion conditions now uses a table and a binary chop
6916 instead of a serial search (which was left over from when there were very
6917 few conditions). Also, it now recognizes conditions like "pam" even when
6918 the relevant support is not compiled in: a suitably worded error message is
6919 given if an attempt is made to use such a condition.
6921 11. Added ${time_interval:xxxxx}.
6923 12. A bug was causing one of the ddress fields not to be passed back correctly
6924 from remote delivery subprocesses. The field in question was not being
6925 subsequently used, so this caused to problems in practice.
6927 13. Added new log selectors queue_time and deliver_time.
6929 14. Might have fixed a bug in maildirsizefile handling that threw up
6930 "unexpected character" debug warnings, and recalculated the data
6931 unnecessarily. In any case, I expanded the warning message to give more
6934 15. Added the message "Restricted characters in address" to the statements in
6935 the default ACL that block characters like @ and % in local parts.
6937 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6938 Three changes have been made:
6940 (a) There was a serious bug; a negative response to MAIL caused the whole
6941 recipient domain to be cached as invalid, thereby blocking all messages
6942 to all local parts at the same domain, from all senders. This bug has
6943 been fixed. The domain is no longer cached after a negative response to
6944 MAIL if the sender used is not empty.
6946 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6949 (c) A new callout option, "use_sender" has been added for people who want
6950 the modified behaviour.
6956 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6959 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6960 indeed breaks things for older releases.
6962 3. Added additional logging to the case where there is a problem reading data
6963 from a filter that is running in a subprocess using a pipe, in order to
6964 try to track down a specific problem.
6966 4. Testing facility fudge: when running in the test harness and attempting
6967 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6968 getting "No route to host". Convert this to a timeout.
6970 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6973 6. Some OS don't have socklen_t but use size_t instead. This affects the
6974 fifth argument of getsockopt() amongst other things. This is now
6975 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6976 can be set for individual OS. I have set it for SunOS5, OSF1, and
6977 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6978 some earlier ones do not.
6980 7. Change 4.30/15 was not doing the test caselessly.
6982 8. The standard form for an IPv6 address literal was being rejected by address
6983 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6984 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6985 this, as well as the form without the "IPv6" on the front (but only when
6986 address literals are enabled, of course).
6988 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6990 10. Exim crashed if a message with an empty sender address specified by -f
6991 encountered a router with an errors_to setting. This could be provoked only
6992 by a command such as
6996 where an empty string was supplied; "<>" did not hit this bug.
6998 11. Installed PCRE release 4.5.
7000 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7001 remained set. It is now erased.
7003 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7004 times from message ids (which are base 36 rather than the normal 62).
7006 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7007 were being counted as actual protocol errors, and logged if the log
7008 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7009 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7010 and DATA following a set of rejected RCPTs do not count as protocol errors.
7011 In other words, Exim assumes they were pipelined, though this may not
7012 actually be the case. Of course, in all cases the client gets an
7013 appropriate error code.
7015 15. If a lookup fails in an ACL condition, a message about the failure may
7016 be available; it is used if testing the ACL cannot continue, because most
7017 such messages specify what the cause of the deferral is. However, some
7018 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7019 that caused an old message to be retained and used if a later statement
7020 caused a defer, replacing the real cause of the deferral.
7022 16. If an IP address had so many PTR records that the DNS lookup buffer
7023 was not large enough to hold them, Exim could crash while trying to process
7024 the truncated data. It now detects and logs this case.
7026 17. Further to 4.21/58, another change has been made: if (and only if) the
7027 first line of a message (the first header line) ends with CRLF, a bare LF
7028 in a subsequent header line has a space inserted after it, so as not to
7029 terminate the header.
7031 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7032 unnecessarily, used atoi() instead of strtol(), and didn't check the
7033 termination when getting file sizes from file names by regex.
7035 19. Completely re-implemented the support for maildirsize files, in the light
7036 of a number of problems with the previous contributed implementation
7037 (4.30/29). In particular:
7039 . If the quota is zero, the maildirsize file is maintained, but no quota is
7042 . If the maildir directory does not exist, it is created before any attempt
7043 to write a maildirsize file.
7045 . The quota value in the file is just a cache; if the quota is changed in
7046 the transport, the new value overrides.
7048 . A regular expression is available for excluding directories from the
7051 20. The autoreply transport checks the characters in options that define the
7052 message's headers; it allows continued headers, but it was checking with
7053 isspace() after an embedded newline instead of explicitly looking for a
7056 21. If all the "regular" hosts to which an address was routed had passed their
7057 expiry times, and had not reached their retry times, the address was
7058 bounced, even if fallback hosts were defined. Now Exim should go on to try
7061 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7062 equivalent code in the SMTP transport. Some hosts send humungous responses
7063 to HELO/EHLO, more than 1024 it seems.
7065 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7066 gives compiler warnings in some environments. I've now done it "properly",
7069 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7070 (because of problems with the built-in one) was declared to return uschar *
7071 instead of char *, causing compiler failure.
7073 25. Fixed a file descriptor leak when processing alias/forward files.
7075 26. Fixed a minor format string issue in dbfn.c.
7077 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7079 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7080 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7083 29. When a sender address is verified, it is cached, to save repeating the test
7084 when there is more than one recipient in a message. However, when the
7085 verification involves a callout, it is possible for different callout
7086 options to be set for different recipients. It is too complicated to keep
7087 track of this in the cache, so now Exim always runs a verification when a
7088 callout is required, relying on the callout cache for the optimization.
7089 The overhead is duplication of the address routing, but this should not be
7092 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7093 to be verified with callout=postmaster, and the main callout worked but the
7094 postmaster check failed, the verification correctly failed. However, if a
7095 subsequent RCPT command asked for sender verification *without* the
7096 postmaster check, incorrect caching caused this verification also to fail,
7099 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7100 it was not caching the DNS options (qualify_single, search_parents) that
7101 were used when the lookup failed. A subsequent lookup with different
7102 options therefore always gave the same answer, though there were cases
7103 where it should not have. (Example: a "domains = !$mx_any" option on a
7104 dnslookup router: the "domains" option is always processed without any
7105 widening, but the router might have qualify_single set.) Now Exim uses the
7106 cached value only when the same options are set.
7108 32. Added John Jetmore's "exipick" utility to the distribution.
7110 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7111 than a timeout (e.g. a certificate is required, and is not provided), an
7112 Exim server now closes the connection immediately. Previously it waited for
7113 the client to close - but if the client is SSL, it seems that they each
7114 wait for each other, leading to a delay before one of them times out.
7116 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7117 maintained 0.8.x compatibility because I don't think many are using it, and
7118 it is clearly obsolete.
7120 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7123 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7124 client certificate was expired. A simple patch fixes this, though I don't
7125 understand the full logic of why the verify callback is called multiple
7128 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7129 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7130 which causes problems with some clients (such as the Certicom SSL Plus
7131 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7132 disables the coutermeasure allowing Eudora to connect."
7134 38. Exim was not checking that a write() to a log file succeeded. This could
7135 lead to Bad Things if a log got too big, in particular if it hit a file
7136 size limit. Exim now panics and dies if it cannot write to a log file, just
7137 as it does if it cannot open a log file.
7139 39. Modified OS/Makefile-Linux so that it now contains
7141 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7143 The two -D definitions ensure that Exim is compiled with large file
7144 support, which makes it possible to handle log files that are bigger than
7147 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7148 instance) a domain was checked against a named list that involved a lookup,
7149 causing $domain_data to be set, then another domain was checked against the
7150 same list, then the first domain was re-checked, the value of $domain_data
7151 after the final check could be wrong. In particular, if the second check
7152 failed, it could be set empty. This bug probably also applied to
7155 41. The strip_trailing_dot option was not being applied to the address given
7156 with the -f command-line option.
7158 42. The code for reading a message's header from the spool was incrementing
7159 $received_count, but never initializing it. This meant that the value was
7160 incorrect (doubled) while delivering a message in the same process in which
7161 it was received. In the most common configuration of Exim, this never
7162 happens - a fresh exec is done - but it can happen when
7163 deliver_drop_privilege is set.
7165 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7166 it now includes up to 150 characters of the unexpected data in the log
7169 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7170 and building data strings. The size of both of these buffers was 10 000
7171 bytes - far larger than anybody would *ever* want, thought I. Needless to
7172 say, somebody hit the limit. I have increased the maximum line length to
7173 20 000 and the maximum data length of concatenated lines to 100 000. I have
7174 also fixed two bugs, because there was no checking on these buffers. Tsk,
7175 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7176 buffer is too small.
7178 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7179 lsearch lookups. Now it does.
7181 46. When parsing a route_list item in a manualroute router, a fixed-length
7182 buffer was used for the list of hosts. I made this 1024 bytes long,
7183 thinking that nobody would ever have a list of hosts that long. Wrong.
7184 Somebody had a whole pile of complicated expansion conditions, and the
7185 string was silently truncated, leading to an expansion error. It turns out
7186 that it is easier to change to an unlimited length (owing to other changes
7187 that have happened since this code was originally written) than to build
7188 structure for giving a limitation error. The length of the item that
7189 expands into the list of hosts is now unlimited.
7191 47. The lsearch lookup could not handle data where the length of text line was
7192 more than 4095 characters. Such lines were truncated, leading to shortened
7193 data being returned. It should now handle lines of any length.
7195 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7196 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7199 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7200 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7201 no longer happens in comment lines. A list of replacements is now placed
7202 at the head of all of the source files, except those whose only change is
7203 to replace PERL_COMMAND in the very first #! line.
7205 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7206 messages on the queue, with a bottom-up merge sort, using code contributed
7207 by Michael Haardt. This should make operations like -bp somewhat faster on
7208 large queues. It won't affect queue runners, except when queue_run_in_order
7211 51. Installed eximstats 1.31 in the distribution.
7213 52. Added support for SRV lookups to the dnslookup router.
7215 53. If an ACL referred to $message_body or $message_body_end, the value was not
7216 reset for any messages that followed in the same SMTP session.
7218 54. The store-handling optimization for building very long strings was not
7219 differentiating between the different store pools. I don't think this
7220 actually made any difference in practice, but I've tidied it.
7222 55. While running the routers to verify a sender address, $sender_address
7223 was still set to the sender address. This is wrong, because when routing to
7224 send a bounce to the sender, it would be empty. Therefore, I have changed
7225 it so that, while verifying a sender address, $sender_address is set to <>.
7226 (There is no change to what happens when verifying a recipient address.)
7228 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7229 target A or AAAA records (if not already returned) without resetting the
7230 qualify_single or search_parents options of the DNS resolver. These are
7231 inappropriate in this case because the targets of MX and SRV records must
7232 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7233 target that, when qualified, matched something in the local domain. These
7234 two options are now turned off when doing these lookups.
7236 57. It seems that at least some releases of Reiserfs (which does not have the
7237 concept of a fixed number of inodes) returns zero and not -1 for the
7238 number of available inodes. This interacted badly with check_spool_inodes,
7239 which assumed that -1 was the "no such thing" setting. What I have done is
7240 to check that the total number of inodes is greater than zero before doing
7241 the test of how many are available.
7243 58. When a "warn" ACL statement has a log_message modifier, the message is
7244 remembered, and not repeated. This is to avoid a lot of repetition when a
7245 message has many recipients that cause the same warning to be written.
7246 However, Exim was preserving the list of already written lines for an
7247 entire SMTP session, which doesn't seem right. The memory is now reset if a
7248 new message is started.
7250 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7251 debugging output unless log_rewrite was also set.
7253 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7254 of (int)(handle) when we know that handle contains (void *)(-1).
7256 61. The Exim daemon panic-logs an error return when it closes the incoming
7257 connection. However "connection reset by peer" seems to be common, and
7258 isn't really an error worthy of noting specially, so that particular error
7261 62. When Exim is trying to find all the local interfaces, it used to panic and
7262 die if the ioctl to get the interface flags failed. However, it seems that
7263 on at least one OS (Solaris 9) it is possible to have an interface that is
7264 included in the list of interfaces, but for which you get a failure error
7265 for this call. This happens when the interface is not "plumbed" into a
7266 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7267 failure of the "get flags" call assumes that the interface is down.
7269 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7270 makes life easier for people who are doing arithmetic on fields extracted
7271 from dates, where you often get leading zeros that should not be
7272 interpreted as octal.
7274 64. Added qualify_domain to the redirect router, to override the global
7277 65. If a pathologically long header line contained very many addresses (the
7278 report of this problem mentioned 10 000) and each of them was rewritten,
7279 Exim could use up a very large amount of memory. (It kept on making new
7280 copies of the header line as it rewrote, and never released the old ones.)
7281 At the expense of a bit more processing, the header rewriting function has
7282 been changed so that it no longer eats memory in this way.
7284 66. The generation of the Received: header has been moved from the time that a
7285 message starts to be received, to the time that it finishes. The timestamp
7286 in the Received: header should now be very close to that of the <= log
7287 line. There are two side-effects of this change:
7289 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7290 logged header lines no longer include the local Received: line, because
7291 it has not yet been created. The same applies to a copy of the message
7292 that is returned to a non-SMTP sender when a message is rejected.
7294 (b) When a filter file is tested using -bf, no additional Received: header
7295 is added to the test message. After some thought, I decided that this
7298 This change does not affect the value of $received_for. It is still set
7299 after address rewriting, but before local_scan() is called.
7301 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7303 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7304 gave an unhelpful panic error message, and a defer error. I have managed to
7305 change this behaviour so that it now rejects any supplied certificate,
7306 which seems right, as the list of acceptable certificates is empty.
7308 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7309 gave an unhelpful defer error. I have not managed to make this reject any
7310 supplied certificates, but the error message it gives is "no certificate
7311 supplied", which is not helpful.
7313 70. exigrep's output now also includes lines that are not associated with any
7314 message, but which match the given pattern. Implemented by a patch from
7315 Martin Sluka, which also tidied up the Perl a bit.
7317 71. Recipient callout verification, like sender verification, was using <> in
7318 the MAIL FROM command. This isn't really the right thing, since the actual
7319 sender may affect whether the remote host accepts the recipient or not. I
7320 have changed it to use the actual sender in the callout; this means that
7321 the cache record is now keyed on a recipient/sender pair, not just the
7322 recipient address. There doesn't seem to be a real danger of callout loops,
7323 since a callout by the remote host to check the sender would use <>.
7324 [SEE ABOVE: changed after hitting problems.]
7326 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7327 temporary errors. However, in the case of such a code being given after
7328 the end of a data transmission (i.e. after ".") Exim was failing to write
7329 a retry record for the message. (Yes, there was some broken host that was
7330 actually sending 8xx at this point.)
7332 73. An unknown lookup type in a host list could cause Exim to panic-die when
7333 the list was checked. (An example that provoked this was putting <; in the
7334 middle of a list instead of at the start.) If this happened during a DATA
7335 ACL check, a -D file could be left lying around. This kind of configuration
7336 error no longer causes Exim to die; instead it causes a defer error. The
7337 incident is still logged to the main and panic logs.
7339 74. Buglet left over from Exim 3 conversion. The message "too many messages
7340 in one connection" was written to the rejectlog but not the mainlog, except
7341 when address rewriting (yes!) was being logged.
7343 75. Added write_rejectlog option.
7345 76. When a system filter was run not as root (that is, when system_filter_user
7346 was set), the values of the $n variables were not being returned to the
7347 main process; thus, they were not subsequently available in the $sn
7350 77. Added +return_path_on_delivery log selector.
7352 78. A connection timeout was being treated differently from recipients deferred
7353 when testing hosts_max_try with a message that was older than the host's
7354 retry timeout. (The host should not be counted, thus allowing all hosts to
7355 be tried at least once before bouncing.) This may have been the cause of an
7356 occasionally reported bug whereby a message would remain on the queue
7357 longer than the retry timeout, but would be bounced if a delivery was
7358 forced. I say "may" because I never totally pinned down the problem;
7359 setting up timeout/retry tests is difficult. See also the next item.
7361 79. The ultimate address timeout was not being applied to errors that involved
7362 a combination of host plus message (for example, a timeout on a MAIL
7363 command). When an address resolved to a number of possible hosts, and they
7364 were not all tried for each delivery (e.g. because of hosts_max_try), a
7365 message could remain on the queue longer than the retry timeout.
7367 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7370 81. Fixed an obscure SMTP outgoing bug which required at least the following
7371 conditions: (a) there was another message waiting for the same server;
7372 (b) the server returned 5xx to all RCPT commands in the first message so
7373 that the message was not completed; (c) the server dropped the connection
7374 or gave a negative response to the RSET that Exim sends to abort the
7375 transaction. The observed case was a dropped connection after DATA that had
7376 been sent in pipelining mode. That is, the server had advertised PIPELINING
7377 but was not implementing it correctly. The effect of the bug was incorrect
7378 behaviour, such as trying another host, and this could lead to a crash.
7384 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7385 and daemon.c were passed as pointers to ints; they should have been
7386 pointers to socklen_t variables (which are typically unsigned ints).
7388 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7391 3. Fixed a really odd bug that affected only the testing scheme; patching a
7392 certain fixed string in the binary changed the value of another string that
7393 happened to be identical to the end of the original first string.
7395 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7396 name", it returns that address as the IP address. On some operating
7397 systems (e.g. Solaris), it also passes back the IP address string as the
7398 "host name". However, on others (e.g. Linux), it passes back an empty
7399 string. Exim wasn't checking for this, and was changing the host name to an
7400 empty string, assuming it had been canonicalized.
7402 5. Although rare, it is permitted to have more than one PTR record for a given
7403 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7404 all the names associated with an address, because they do in Solaris.
7405 However, it seems that they do not in Linux for data that comes from the
7406 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7407 I found this out when I moved to a new Linux workstation and tried to run
7408 the Exim test suite.
7410 To get round this problem I have changed the code so that it now does its
7411 own call to the DNS to look up PTR records when searching for a host name.
7412 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7413 addresses that are only in /etc/hosts are still found.
7415 This behaviour is, however, controlled by an option called host_lookup_
7416 order, which defaults to "bydns:byaddr". If people want to use the other
7417 order, or indeed, just use one or the other means of lookup, they can
7418 specify it in this variable.
7420 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7421 some operating systems, this comes back from gethostbyaddr() as an empty
7422 string, and this is what Exim used to test for. However, it seems that in
7423 other systems, "." is yielded. Exim now tests for this case too.
7425 7. The values of check_spool_space and check_log_space are now held internally
7426 as a number of kilobytes instead of an absolute number of bytes. If a
7427 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7428 kilobyte. This means that much larger values can be stored.
7430 8. Exim monitor: an attempt to get the action menu when not actually pointing
7431 at a message produces an empty menu entitled "No message selected". This
7432 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7433 no entries in it ("Shell widget menu has zero width and/or height"). So I
7434 have added a single, blank menu entry in this case.
7436 9. Added ${quote_local_part.
7438 10. MIME decoding is now applied to the contents of Subject: header lines when
7441 11. Now that a reference to $sender_host_address automatically causes a reverse
7442 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7443 host lookup before query-style lookups in lists that might use this
7444 variable. This has therefore been abolished, and the "net-" prefix is no
7445 longer necessary for query-style lookups.
7447 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7448 have been a typo for LFLAGS, so it has been changed.
7450 13. The install script calls Exim with "-C /dev/null" in order to find the
7451 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7452 to be output. However, since Exim outputs its version number before the
7453 error, it didn't break the script. It just looked ugly. I fixed this by
7454 always allowing "-C /dev/null" if the caller is root.
7456 14. Ignore overlarge ACL variable number when reading spool file - insurance
7457 against a later release with more variables having written the file.
7459 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7460 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7461 this, as well as the form without the "IPv6" on the front.
7463 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7464 OS/Makefile-Darwin file.
7466 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7467 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7469 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7470 variable. However, some OS use ints for inodes. I've added cast to long int
7471 to get rid of the compiler warning.
7473 19. I had forgotten to lock out "/../" in configuration file names when
7474 ALT_CONFIG_PREFIX was set.
7476 20. Routers used for verification do not need to specify transports. However,
7477 if such a router generated a host list, and callout was configured, Exim
7478 crashed, because it could not find a port number from the (non-existent)
7479 transport. It now assumes port 25 in this circumstance.
7481 21. Added the -t option to exigrep.
7483 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7484 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7485 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7486 them is set, LOOKUP_LSEARCH is forced.
7488 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7490 24. Added sender and host information to the "rejected by local_scan()" log
7491 line; previously there was no indication of these.
7493 25. Added .include_if_exists.
7495 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7496 Linux. It turns out that not all file systems support this. Apparently some
7497 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7498 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7499 not supported on the file descriptor, is now ignored when Exim is trying to
7500 sync a directory. This applies only to Linux.
7502 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7504 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7505 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7508 29. Added support for maildirsize files from supplied patch (modified a bit).
7510 30. The use of :fail: followed by an empty string could lead Exim to respond to
7511 sender verification failures with (e.g.):
7513 550 Verification failed for <xxx>
7514 550 Sender verify failed
7516 where the first response line was missing the '-' that indicates it is not
7517 the final line of the response.
7519 31. The loop for finding the name of the user that called Exim had a hardwired
7520 limit of 10; it now uses the value of finduser_retries, which is used for
7521 all other user lookups.
7523 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7526 33. Exim was neglecting to zero errno before one call of strtol() when
7527 expanding a string and expecting an integer value. On some systems this
7528 resulted in spurious "integer overflow" errors. Also, it was casting the
7529 result into an int without checking.
7531 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7532 did not work. The code looks as if it has *never* worked, though it appears
7533 to have been documented since at least release 1.62. I have made it work.
7535 35. The "timeout_DNS" error in retry rules, also documented since at least
7536 1.62, also never worked. As it isn't clear exactly what this means, and
7537 clearly it isn't a major issue, I have abolished the feature by treating it
7538 as "timeout", and writing a warning to the main and panic logs.
7540 36. The display of retry rules for -brt wasn't always showing the error code
7543 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7544 timeout_connect_A, timeout_connect_MX.
7546 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7547 to the empty sender.
7549 39. The daemon was not analysing the content of -oX till after it had closed
7550 stderr and disconnected from the controlling terminal. This meant that any
7551 syntax errors were only noted on the panic log, and the return code from
7552 the command was 0. By re-arranging the code a little, I've made the
7553 decoding happen first, so such errors now appear on stderr, and the return
7554 code is 1. However, the actual setting up of the sockets still happens in
7555 the disconnected process, so errors there are still only recorded on the
7558 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7559 connections (as happens on some IP stacks) was logged at start up time as
7560 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7561 it from "IPv6 and IPv4", which means that two separate sockets are being
7564 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7565 says whether AF_INET or AF_INET6 was passed as an argument.
7567 42. Exiwhat output was messed up when time zones were included in log
7570 43. Exiwhat now gives more information about the daemon's listening ports,
7571 and whether -tls-on-connect was used.
7573 44. The "port" option of the smtp transport is now expanded.
7575 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7576 silently ignored. Now an error message is written to the main and panic
7579 46. There's a new ACL modifier called "logwrite" which writes to a log file
7580 as soon as it is encountered.
7582 47. Added $local_user_uid and $local_user_gid at routing time.
7584 48. Exim crashed when trying to verify a sender address that was being
7587 49. Exim was recognizing only a space character after ".include". It now also
7588 recognizes a tab character.
7590 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7591 extracting the relevant information from the specification. The man page no
7592 longer contains scrambled data for the -d option, and I've added a section
7593 at the front about calling Exim under different names.
7595 51. Added "extra_headers" argument to the "mail" command in filter files.
7597 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7600 53. Installed eximstats 1.29.
7602 54. Added transport_filter_timeout as a generic transport option.
7604 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7605 Cc: header lines. This was required by RFC 822, but it not required by RFC
7608 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7609 incoming messages that did not have them. Now it does so only if the
7610 message originates locally, that is, if there is no associated remote host
7611 address. When Resent- header lines are present, this applies to the Resent-
7612 lines rather than the non-Resent- lines.
7614 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7615 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7617 58. Messages for configuration errors now include the name of the main
7618 configuration file - useful now that there may be more than one file in a
7619 list (.included file names were always shown).
7621 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7622 for those rare installations that do not start the daemon as root or run it
7623 setuid root. I've cut out the call to initgroups() if the daemon is not
7626 60. The Exim user and group can now be bound into the binary as text strings
7627 that are looked up at the start of Exim's processing.
7629 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7631 62. Added $mailstore_basename variable.
7633 63. Installed patch to sieve.c from Michael Haardt.
7635 64. When Exim failed to open the panic log after failing to open the main log,
7636 the original message it was trying to log was written to stderr and debug
7637 output, but if they were not available (the usual case in production), it
7638 was lost. Now it is written to syslog before the two lines that record the
7639 failures to open the logs.
7641 65. Users' Exim filters run in subprocesses under the user's uid. It is
7642 possible for a "deliver" command or an alias in a "personal" command to
7643 provoke an address rewrite. If logging of address rewriting is configured,
7644 this fails because the process is not running as root or exim. There may be
7645 a better way of dealing with this, but for the moment (because 4.30 needs
7646 to be released), I have disabled address rewrite logging when running a
7647 filter in a non-root, non-exim process.
7653 1. The buildconfig auxiliary program wasn't quoting the value set for
7654 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7655 not defined. This bug was masked in 4.22 by the effect that was fixed in
7658 2. Some messages that were rejected after a message id was allocated were
7659 shown as "incomplete" by exigrep. It no longer does this for messages that
7660 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7662 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7663 have allow_domain_literals set, the ID did not get logged in the <= line.
7664 Domain literals are now always recognized in Message-ID: header lines.
7666 4. The first argument for a ${extract expansion item is the key name or field
7667 number. Leading and trailing spaces in this item were not being ignored,
7668 causing some misleading effects.
7670 5. When deliver_drop_privilege was set, single queue runner processes started
7671 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7672 same command in the process it spins off) were not dropping privilege.
7674 6. When the daemon running as "exim" started a queue runner, it always
7675 re-executed Exim in the spun-off process. This is a waste of effort when
7676 deliver_drop_privilege is set. The new process now just calls the
7677 queue-runner function directly.
7683 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7686 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7687 ignored. Though the use of -oP was forcing the writing of a pid file, it
7688 was always written to the default place.
7690 3. If the message "no IP address found for host xxxx" is generated during
7691 incoming verification, it is now followed by identification of the incoming
7692 connection (so you can more easily find what provoked it).
7694 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7696 5. Added some features to "harden" Exim a bit more against certain attacks:
7698 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7699 be put in Local/Makefile. This is like the never_users runtime option,
7700 but it cannot be overridden. The default setting is "root".
7702 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7703 prefix string with which any file named in a -C command line option
7706 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7707 is retained for -C and -D only if the caller of Exim is root. Without
7708 it, the exim user may also use -C and -D and retain privilege.
7710 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7711 command line option is disabled.
7713 6. Macro names set by the -D option must start with an upper case letter, just
7714 like macro names defined in the configuration file.
7716 7. Added "dereference=" facility to LDAP.
7718 8. Two instances of the typo "uknown" in the source files are fixed.
7720 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7721 the Configure-Makefile script screwed up while processing it.
7723 10. Incorporated PCRE 4.4.
7725 11. The SMTP synchronization check was not operating right at the start of an
7726 SMTP session. For example, it could not catch a HELO sent before the client
7727 waited for the greeting. There is now a check for outstanding input at the
7728 point when the greeting is written. Because of the duplex, asynchronous
7729 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7730 way, but not yet received, when the check is performed.
7732 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7733 on TCP/IP sockets, because this apparently causes some broken clients to
7736 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7737 unchanged) from the Cygwin maintainer.
7739 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7740 was supported instead of testing for mbx. Effectively a typo.
7742 15. The spa authenticator server code was not checking that the input it
7743 received was valid base64.
7745 16. The debug output line for the "set" modifier in ACLs was not showing the
7746 name of the variable that was being set.
7748 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7750 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7751 lookup on its own. Something else was needed to trigger the lookup. For
7752 example, a match in host_lookup or the need for a host name in a host list.
7753 Now, if $sender_host_name is referenced and the host name has not yet been
7754 looked up, a lookup is performed. If the lookup fails, the variable remains
7755 empty, and $host_lookup_failed is set to "1".
7757 19. Added "eqi" as a case-independent comparison operator.
7759 20. The saslauthd authentication condition could segfault if neither service
7760 nor realm was specified.
7762 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7763 error message that was logged was misleading, and incoming SMTP
7764 connections were dropped. The message is now more accurate, and temporary
7765 errors are given to SMTP connections.
7767 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7768 (except RSET) with a 421 error, until QUIT is received. However, it was
7769 failing to send a response to QUIT.
7771 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7772 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7773 ACL is now run after the helo_try_verify_hosts code.
7775 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7776 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7777 case-independent, so other case variants are also recognized). Apparently
7778 some systems use these upper case variants.
7780 25. If more than two messages were waiting for the same host, and a transport
7781 filter was specified for the transport, Exim sent two messages over the
7782 same TCP/IP connection, and then failed with "socket operation on non-
7783 socket" when it tried to send the third.
7785 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7787 27. The extern definition of crypt16() in expand.c was not being excluded when
7788 the OS had its own crypt16() function.
7790 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7791 as a preferred synonym for return_size_limit, both as an option and as an
7794 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7796 30. Changed the default configuration ACL to relax the local part checking rule
7797 for addresses that are not in any local domains. For these addresses,
7798 slashes and pipe symbols are allowed within local parts, but the sequence
7799 /../ is explicitly forbidden.
7801 31. SPA server authentication was not clearing the challenge buffer before
7804 32. log_message in a "warn" ACL statement was writing to the reject log as
7805 well as to the main log, which contradicts the documentation and doesn't
7806 seem right (because no rejection is happening). So I have stopped it.
7808 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7809 However, I am unable to do any testing of this.
7811 34. Fixed an infelicity in the appendfile transport. When checking directories
7812 for a mailbox, to see if any needed to be created, it was accidentally
7813 using path names with one or more superfluous leading slashes; tracing
7814 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7816 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7817 contents are added to the log line that is written for every discarded
7818 recipient. (Previously a log_message setting was ignored.)
7820 36. The ${quote: operator now quotes the string if it is empty.
7822 37. The install script runs exim in order to find its version number. If for
7823 some reason other than non-existence or emptiness, which it checks, it
7824 could not run './exim', it was installing it with an empty version number,
7825 i.e. as "exim-". This error state is now caught, and the installation is
7828 38. An argument was missing from the function that creates an error message
7829 when Exim fails to connect to the socket for saslauthd authentication.
7830 This could cause Exim to crash, or give a corrupted message.
7832 39. Added isip, isip4, and isip6 to ${if conditions.
7834 40. The ACL variables $acl_xx are now saved with the message, and can be
7835 accessed later in routers, transports, and filters.
7837 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7838 strings in the file are not string-expanded.
7840 42. If a MAIL command specified a SIZE value that was too large to fit into an
7841 int variable, the check against message_size_limit failed. Such values are
7842 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7843 day this will have to be increased, but I don't think I want to be around
7844 when emails are that large.
7851 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7852 iconv() is not standard in FreeBSD.
7854 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7855 IPv6 enabled. The observed symptom was a segmentation fault on return from
7856 the function os_common_find_running_interfaces() in src/os.c.
7858 3. In the check_special_case() function in daemon.c I had used "errno" as an
7859 argument name, which causes warnings on some systems. This was basically a
7860 typo, since it was named "eno" in the comments!
7862 4. The code that waits for the clock to tick (at a resolution of some fraction
7863 of a second) so as to ensure message-id uniqueness was always waiting for
7864 at least one whole tick, when it could have waited for less. [This is
7865 almost certainly not relevant at current processor speeds, where it is
7866 unlikely to ever wait at all. But we try to future-proof.]
7868 5. The function that sleeps for a time interval that includes fractions of a
7869 second contained a race. It did not block SIGALRM between setting the
7870 timer, and suspending (a couple of lines later). If the interval was short
7871 and the sigsuspend() was delayed until after it had expired, the suspension
7872 never ended. On busy systems this could lead to processes getting stuck for
7875 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7876 process, before it forks any delivery processes. The open lookup caching
7877 mechanism meant that the open file or database connection was passed into
7878 the delivery process. The problem was that delivery processes always tidy
7879 up cached lookup data. This could cause a problem for the next delivery
7880 process started by the queue runner, because the external queue runner
7881 process does not know about the closure. So the next delivery process
7882 still has data in the lookup cache. In the case of a file lookup, there was
7883 no problem because closing a file descriptor in a subprocess doesn't affect
7884 the parent. However, if the lookup was caching a connection to a database,
7885 the connection was closed, and the second delivery process was likely to
7886 see errors such as "PGSQL: query failed: server closed the connection
7887 unexpectedly". The problem has been fixed by closing all cached lookups
7888 in a queue runner before running a delivery process.
7890 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7891 seem to have the "const" qualifier which it has on other OS. I've
7894 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7895 *of the same type* (client or server) with the same public name that an
7896 error should be diagnosed.
7898 9. When Exim looked up a host name for an IP address, but failed to find the
7899 original IP address when looking up the host name (a safety check), it
7900 output the message "<ip address> does not match any IP for NULL", which was
7901 confusing, to say the least. The bug was that the host name should have
7902 appeared instead of "NULL".
7904 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7905 user that is built into the binary, and the -C or -D options is used, root
7906 privilege is dropped before the configuration file is read. In addition,
7907 logging is switched to stderr instead of the normal log files. If the
7908 configuration then re-defines the Exim user, the unprivileged environment
7909 is probably not what is expected, so Exim logs a panic warning message (but
7912 However, if deliver_drop_privilege is set, the unprivileged state may well
7913 be exactly what is intended, so the warning has been cut out in that case,
7914 and Exim is allowed to try to write to its normal log files.
7920 1. smtp_return_error_details was not giving details for temporary sender
7921 or receiver verification errors.
7923 2. Diagnose a configuration error if two authenticators have the same public
7926 3. Exim used not to create the message log file for a message until the first
7927 delivery attempt. This could be confusing when incoming messages were held
7928 for policy or load reasons. The message log file is now created at the time
7929 the message is received, and an initial "Received" line is written to it.
7931 4. The automatically generated man page for command line options had a minor
7932 bug that caused no ill effects; however, a more serious problem was that
7933 the procedure for building the man page automatically didn't always
7934 operate. Consequently, release 4.20 contains an out-of-date version. This
7935 shouldn't happen again.
7937 5. When building Exim with embedded Perl support, the script that builds the
7938 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7939 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7941 6. The freeze_tell option was not being used for messages that were frozen on
7942 arrival, either by an ACL or by local_scan().
7944 7. Added the smtp_incomplete_transaction log selector.
7946 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7947 was accepting AUTH without a new EHLO.
7949 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7950 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7953 10. Logging of TCP/IP connections (when configured) now happens in the main
7954 daemon process instead of the child process, so that the TCP/IP connection
7955 count is more accurate (but it can never be perfect).
7957 11. The use of "drop" in a nested ACL was not being handled correctly in the
7958 outer ACL. Now, if condition failure induced by the nested "drop" causes
7959 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7960 or "require"), the connection is dropped.
7962 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7963 that yield "discard" can now be used with an "accept" or a "discard" verb,
7964 but an error is generated for any others (because I can't see a useful way
7965 to define what should happen).
7967 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7968 are now processed in the same way as lines in the Exim configuration file.
7969 In particular, continuation lines are supported.
7971 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7973 15. Added -ti meaning -t -i.
7975 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7976 domains, and warn by logging if others are found.
7978 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7979 returned from GIFCONF when Exim is trying to find the list of interfaces on
7980 a host. The code in os.c has been modified to copy each ifreq to an aligned
7981 structure in all cases.
7983 Also, in some cases, the returned ifreq's were being copied to a 'struct
7984 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7985 means the last couple of bytes of an IPv6 address could be chopped if the
7986 ifreq contained only a normal sockaddr (14 bytes storage).
7988 18. Named domain lists were not supported in the hosts_treat_as_local option.
7989 An entry such as +xxxx was not recognized, and was treated as a literal
7992 19. Ensure that header lines added by a DATA ACL are included in the reject log
7993 if the ACL subsequently rejects the message.
7995 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7996 MD5 (which is deprecated).
7998 21. When testing a filter file using -bf, Exim was writing a message when it
7999 took the sender from a "From " line in the message, but it was not doing so
8000 when it took $return_path from a Return-Path: header line. It now does.
8002 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8003 with a valid header line field name (a series of printing characters
8004 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8006 23. Changed "disc" in the source to "disk" to conform to the documentation and
8007 the book and for uniformity.
8009 24. Ignore Sendmail's -Ooption=value command line item.
8011 25. When execve() failed while trying to run a command in a pipe transport,
8012 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8013 could be confused with a return value of 69 from the command itself. This
8014 has been changed to 127, the value the shell returns if it is asked to run
8015 a non-existent command. The wording for the related log line suggests a
8016 non-existent command as the problem.
8018 26. If received_header_text expands to an empty string, do not add a Received:
8019 header line to the message. (Well, it adds a token one on the spool, but
8020 marks it "old" so that it doesn't get used or transmitted.)
8022 27. Installed eximstats 1.28 (addition of -nt option).
8024 28. There was no check for failure on the call to getsockname() in the daemon
8025 code. This can fail if there is a shortage of resources on the system, with
8026 ENOMEM, for example. A temporary error is now given on failure.
8028 29. Contrary to the C standard, it seems that in some environments, the
8029 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8030 program. Exim now does this explicitly; it affects the formatting of
8031 timestamps using strftime().
8033 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8034 complaints. I've now initialized all the variables, to avoid this.
8036 32. Header lines added by a system filter were not being "seen" during
8037 transport-time rewrites.
8039 33. The info_callback() function passed to OpenSSL is set up with type void
8040 (*)(SSL *, int, int), as described somewhere. However, when calling the
8041 function (actually a macro) that sets it up, the type void(*)() is
8042 expected. I've put in a cast to prevent warnings from picky compilers.
8044 34. If a DNS black list lookup found a CNAME record, but there were no A
8045 records associated with the domain it pointed at, Exim crashed.
8047 35. If a DNS black list lookup returned more than one A record, Exim ignored
8048 all but the first. It now scans all returned addresses if a particular IP
8049 value is being sought. In this situation, the contents of the
8050 $dnslist_value variable are a list of all the addresses, separated by a
8053 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8054 to accept a host name and all its aliases if the forward lookup for any of
8055 them yielded the IP address of the incoming connection. Now it accepts only
8056 those names whose forward lookup yields the correct IP address. Any other
8057 names are discarded. This closes a loophole whereby a rogue DNS
8058 administrator could create reverse DNS records to break through a
8059 wildcarded host restriction in an ACL.
8061 37. If a user filter or a system filter that ran in a subprocess used any of
8062 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8063 the wrong values were passed to the pipe command ($thisaddress had the
8064 value of $0, $0 had the value of $1, etc). This bug was introduced by
8065 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8067 38. Improved the line breaking for long SMTP error messages from ACLs.
8068 Previously, if there was no break point between 40 and 75 characters, Exim
8069 left the rest of the message alone. Two changes have been made: (a) I've
8070 reduced the minimum length to 35 characters; (b) if it can't find a break
8071 point between 35 and 75 characters, it looks ahead and uses the first one
8072 that it finds. This may give the occasional overlong line, but at least the
8073 remaining text gets split now.
8075 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8076 file descriptors might be low, and that setting 1000 would always raise it.
8077 It turns out that in some environments, the limit is already over 1000 and
8078 that lowering it causes trouble. So now Exim takes care not to decrease it.
8080 40. When delivering a message, the value of $return_path is set to $sender_
8081 address at the start of routing (routers may change the value). By an
8082 oversight, this default was not being set up when an address was tested by
8083 -bt or -bv, which affected the outcome if any router or filter referred to
8086 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8087 log only when "message" or "log_message" are set. However, if one of the
8088 conditions was an address verification, or a call to a nested ACL, the
8089 messages generated by the underlying test were being passed through. This
8090 no longer happens. The underlying message is available in $acl_verify_
8091 message for both "message" and "log_message" expansions, so it can be
8092 passed through if needed.
8094 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8095 new expansion $bh_ to give the encoded byte string without charset
8096 translation. Translation happens only if iconv() is available; HAVE_ICONV
8097 indicates this at build time. HEADERS_CHARSET gives the charset to
8098 translate to; headers_charset can change it in the configuration, and
8099 "headers charset" can change it in an individual filter file.
8101 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8102 that creates RFC 2047 encoded "words" labels them as that charset instead
8103 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8104 expansion operator; (ii) when Exim creates a From: line for a local
8105 message; (iii) when a header line is rewritten to include a "phrase" part.
8107 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8108 buggy, causing it to skip the first lines of messages whose message ID
8109 ended in 'D'. This would not have bitten before Exim release 4.14, because
8110 message IDs were unlikely to end in 'D' before then. The effect was to have
8111 incorrect size information for certain domains.
8113 45. #include "config.h" was missing at the start of the crypt16.c module. This
8114 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8117 46. If there was a timeout during a "random" callout check, Exim treated it as
8118 a failure of the random address, and carried on sending RSET and the real
8119 address. If the delay was just some slowness somewhere, the response to the
8120 original RCPT would be taken as a response to RSET and so on, causing
8121 mayhem of various kinds.
8123 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8124 when I implemented it. It didn't allow for the fact that some option values
8125 may legitimately be negative (e.g. size_addition), and it didn't even do
8126 the right test for positive values.
8128 48. Domain names in DNS records are case-independent. Exim always looks them up
8129 in lower case. Some resolvers return domain names in exactly the case they
8130 appear in the zone file, that is, they may contain uppercase letters. Not
8131 all resolvers do this - some return always lower case. Exim was treating a
8132 change of case by a resolver as a change of domain, similar to a widening
8133 of a domain abbreviation. This triggered its re-routing code and so it was
8134 trying to route what was effectively the same domain again. This normally
8135 caused routing to fail (because the router wouldn't handle the domain
8136 twice). Now Exim checks for this case specially, and just changes the
8137 casing of the domain that it ultimately uses when it transmits the message
8140 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8143 50. If a filter generated a file delivery with a non-absolute name (possible if
8144 no home directory exists for the router), the forbid_file option was not
8147 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8148 the existing equality matching.
8150 52. Exim was using ints instead of ino_t variables in some places where it was
8151 dealing with inode numbers.
8153 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8154 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8155 called TMPDIR, and if it finds it is different, it changes its value.
8157 54. The smtp_printf() function is now made available to local_scan() so
8158 additional output lines can be written before returning. There is also an
8159 smtp_fflush() function to enable the detection of a dropped connection.
8160 The variables smtp_input and smtp_batched_input are exported to
8163 55. Changed the default runtime configuration: the message "Unknown user"
8164 has been removed from the ACL, and instead placed on the localuser router,
8165 using the cannot_route_message feature. This means that any verification
8166 failures that generate their own messages won't get overridden. Similarly,
8167 the "Unrouteable address" message that was in the ACL for unverifiable
8168 relay addresses has also been removed.
8170 56. Added hosts_avoid_esmtp to the smtp transport.
8172 57. The exicyclog script was not checking for the esoteric option
8173 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8174 will work only if exicyclog is run under the appropriate euid.
8176 58. Following a discussion on the list, the rules by which Exim recognises line
8177 endings on incoming messages have been changed. The -dropcr and drop_cr
8178 options are now no-ops, retained only for backwards compatibility. The
8179 following line terminators are recognized: LF CRLF CR. However, special
8180 processing applies to CR:
8182 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8183 nor a local message in the state where . is a terminator.
8185 (ii) If a bare CR is encountered in a header line, an extra space is added
8186 after the line terminator so as not to end the header. The reasoning
8187 behind this is that bare CRs in header lines are most likely either
8188 to be mistakes, or people trying to play silly games.
8190 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8191 was being incorrectly given as 18 bytes larger than it should have been.
8192 This is a VOB (very old bug).
8194 60. This may never have affected anything current, but just in case it has:
8195 When the local host is found other than at the start of a list of hosts,
8196 the local host, those with the same MX, and any that follow, are discarded.
8197 When the list in question was part of a longer list of hosts, the following
8198 hosts (not currently being processed) were also being discarded. This no
8199 longer happens. I'm not sure if this situation could ever has previously
8202 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8205 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8206 header. This is something that is recommended in a new Internet Draft, and
8207 is something that is documented as being done by Sendmail. There are two
8208 possible values. For messages generated by the autoreply transport, Exim
8211 Auto-Submitted: auto-replied
8213 whereas for all other generated messages (e.g. bounces) it adds
8215 Auto-Submitted: auto-generated
8217 63. The "personal" condition in filters now includes a test for the
8218 Auto-Submitted: header. If it contains the string "auto-" the message it
8219 not considered personal.
8221 64. Added rcpt_include_affixes as a generic transport option.
8223 65. Added queue_only_override (default true).
8225 66. Added the syslog_duplication option.
8227 67. If what should have been the first header line of a message consisted of
8228 a space followed by a colon, Exim was mis-interpreting it as a header line.
8229 It isn't of course - it is syntactically invalid and should therefore be
8230 treated as the start of the message body. The misbehaviour could have
8231 caused a number of strange effects, including loss of data in subsequent
8232 header lines, and spool format errors.
8234 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8235 client host had authenticated. This control can now be exercised by an ACL
8236 for more flexibility.
8238 69. By default, callouts do not happen when testing with -bh. There is now a
8239 variant, -bhc, which does actually run the callout code, including
8240 consulting and updating the callout cache.
8242 70. Added support for saslauthd authentication, courtesy of Alexander
8245 71. If statvfs() failed on the spool or log directories while checking their
8246 size for availability, Exim confusingly gave the error "space shortage".
8247 Furthermore, in debugging mode it crashed with a floating point exception.
8248 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8249 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8250 serious problem, Exim now writes to the main and panic logs when this
8251 happens, with details of the failure. It then refuses to accept the
8252 incoming message, giving the message "spool directory problem" or "log
8253 directory problem" with a 421 code for SMTP messages.
8255 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8256 0, 1, and 2 exist, because some OS complain for execs without them (see
8257 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8258 descriptors. However, the code omitted to check that the open succeeded,
8259 causing mysterious errors if for some reason the permissions on /dev/null
8260 got screwed. Now Exim writes a message to the main and panic logs, and
8261 bombs out if it can't open /dev/null.
8263 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8264 interact so that it is all more flexible. It is supposed to remain
8265 backwards compatible. Also added extra_local_interfaces.
8267 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8268 to bomb out with an assertion failure - to the client this appears as a
8269 connection drop. This problem occurs in the part of the code that was taken
8270 from the Samba project. Fortunately, the assertion is in a very simple
8271 function, so I have fixed this by reproducing the function inline in the
8272 one place where it is called, and arranging for authentication to fail
8273 instead of killing the process with assert().
8275 75. The SPA client code was not working when the server requested OEM rather
8276 than Unicode encoding.
8278 76. Added code to make require_files with a specific uid setting more usable in
8279 the case where statting the file as root fails - usually a non-root-mounted
8280 NFS file system. When this happens and the failure is EACCES, Exim now
8281 forks a subprocess and does the per-uid checking as the relevant uid.
8283 77. Added process_log_path.
8285 78. If log_file_path was not explicitly set, a setting of check_log_space or
8286 check_log_inodes was ignored.
8288 79. If a space check for the spool or log partitions fails, the incident is now
8289 logged. Of course, in the latter case the data may get lost...
8291 80. Added the %p formatting code to string_format() so that it can be used to
8292 print addresses in debug_print(). Adjusted all the address printing in the
8293 debugging in store.c to use %p rather than %d.
8295 81. There was a concern that a line of code in smtp_in.c could overflow a
8296 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8297 initially expressed, the concern was not well-founded, because trailing
8298 spaces are removed early. However, if the trailing spaces were followed by
8299 a NULL, they did not get removed, so the overflow was possible. Two fixes
8302 (a) I re-wrote the offending code in a cleaner fashion.
8303 (b) If an incoming SMTP command contains a NULL character, it is rejected
8306 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8307 runs initgroups(), so that if the Exim user is in any additional groups,
8308 they will be used during message reception.
8314 The change log for 4.20 and earlier releases has been archived.