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.
33 JH/01 Move the wait-for-next-tick (needed for unique messmage IDs) from
34 after reception to before a subsequent reception. This should
35 mean slightly faster delivery, and also confirmation of reception
38 JH/02 Move from using the pcre library to pcre2. The former is no longer
39 being developed or supported (by the original developer).
41 JH/03 Constification work in the filters module required a major version
42 bump for the local-scan API. Specifically, the "headers_charset"
43 global which is visible via the API is now const and may therefore
44 not be modified by local-scan code.
46 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
47 sendfile() didi not account for the way the ClamAV driver code called it.
49 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
50 time check was being done for every character; replace that with one
53 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
54 sent was prefixed with a length byte.
56 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
57 RFC 5321. Previously Dovecot (at least) would log errors during
60 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
61 in 4.61 without a reason logged. Binaries will be bigger, which might
62 matter on diskspace-constrained systems, but debug is easier.
64 JH/09 Fix macro-definition during "-be" expansion testing. The move to
65 write-protected store for macros had not accounted for these runtime
66 additions; fix by removing this protection for "-be" mode.
68 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
69 handing out large-numbered file descriptors, violating the usual Unix
70 assumption (and required by Posix) that the lowest possible number will be
71 allocated by the kernel when a new one is needed. In the daemon, and any
72 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
73 are not useable for FD_SET() [and hence select()] and overwrite the stack.
74 Assorted crashes happen.
76 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
77 main-section options or in a connect ACL, the value from the first ever
78 connection was never replaced for subsequent connections. Found by
81 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
82 though only once PCRE2 was introduced: the memory accounting used under
83 debug offset allocations by an int, giving a hard trap in early startup.
84 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
86 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
87 with underbars is given. The write-protection of configuration introduced
88 in 4.95 trapped when normalisation was applied to an option not needing
91 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
93 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
94 failing to close the certificates directory, every hour or any time it
97 JH/16 Debugging initiated by an ACL control now continues through into routing
98 and transport processes. Previously debugging stopped any time Exim
99 re-execs, or for processing a queued message.
101 JH/17 The "expand" debug selector now gives more detail, specifically on the
102 result of expansion operators and items.
104 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
105 bad comparison between the option value and the name of the file to
106 be included was done, and a mismatch was wrongly identified.
107 4.88 to 4.95 are affected.
109 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
111 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
112 supplied as the db file-pair basename. Previously, if a directory
113 path was given, for example via the autoreply "once" option, the DB
114 file.pag and file.dir files would be created in that directory's
117 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
118 "taint" log_selector. These were previously deprecated.
120 JH/22 Fix static address-list lookups to properly return the matched item.
121 Previously only the domain part was returned.
123 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
124 the call into OpenSSL to send a TLS Close was being repeated; this
125 resulted in the library waiting for the peer's Close. If that was never
126 sent we waited forever. Fix by tracking send calls.
128 JH/24 The ${run} expansion item now expands its command string elements after
129 splitting. Previously it was before; the new ordering makes handling
130 zero-length arguments simpler. The old ordering can be obtained by
131 appending a new option "preexpand", after a comma, to the "run".
133 JH/25 Taint-check exec arguments for transport-initiated external processes.
134 Previously, tainted values could be used. This affects "pipe", "lmtp" and
135 "queryprogram" transport, transport-filter, and ETRN commands.
136 The ${run} expansion is also affected: in "preexpand" mode no part of
137 the command line may be tainted, in default mode the executable name
140 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
141 the the facility was not passed across execs, and only the first message
142 passed over a connection could use BDAT; any further ones using DATA.
144 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
145 uses $sending_ip_address and an interface is specified.
146 Previously any use of the local address in the EHLO name disabled
147 PIPECONNECT, the common case being to use the rDNS of it.
149 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
150 resumption. Previously verify failed because no certificate status is
151 passed on the wire for the restarted session. Fix by using the recorded
152 ocsp status of the stored session for the new connection.
154 JH/29 TLS resumption: the key for session lookup in the client now includes
155 more info that a server could potentially use in configuring a TLS
156 session, avoiding oferring mismatching sessions to such a server.
157 Previously only the server IP was used.
159 JH/30 Fix string_copyn() for limit greater than actual string length.
160 Previously the copied amount was the limit, which could result in a
161 overlapping memcpy for newly allocated destination soon after a
162 source string shorter than the limit. Found/investigated by KM.
164 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
165 close; it may be needed for a subsequent connection. This caused a
166 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
168 JH/32 Fix CHUNKING for a second message on a connection when the first was
169 rejected. Previously we did not reset the chunking-offered state, and
170 erroneously rejected the BDAT command. Investigation help from
173 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
174 an empty address. Previously the expansion returned an error.
176 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
177 proxy. Previously these were misparsed, leading to paniclog entries.
183 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
184 related applications. Previously an "H" was used where available info
185 says that "M" should be, so change to match.
187 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
188 as arguments, so an implementation trying to copy these into a local
189 buffer was taking a taint-enforcement trap. Fix by using dynamically
190 created buffers. Similar fix for radius expansion condition.
192 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
193 reasonable, eg. to count headers. Fix by using dynamically created
194 buffers rather than a local. Do similar fixes for ACL actions "dcc",
195 "log_reject_target", "malware" and "spam"; the arguments are expanded
196 so could be handling tainted values.
198 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
199 broken the (no-op) support for this sendmail command. Restore it
200 to doing nothing, silently, and returning good status.
202 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
203 record path was given (or the default used) without a leading directory
204 path, an error occurred on trying to open it. Use the transport's working
207 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
208 transport. Previously it was the name on the DNS A-record; use instead
209 the head of the CNAME chain leading there (if there is one). This seems
210 to align better with RFC 6125.
212 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
213 smtp_accept_max_per_host allocated resources which were not released
214 when the limit was exceeded. This eventually crashed the daemon. Fix
215 by adding a release action in that path.
217 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
218 expanded; previously using tainted values was rejected. Fix by using
219 dynamically-created buffers.
221 JH/09 Relax restrictions on ACL verify condition needing access to message
222 headers. Previously they were only permitted in data and non-smtp ACLs;
223 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
224 not_blind, header_sender and header_names_ascii verification.
226 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
227 Previously a macro used one argument twice; when called with the
228 argument as an expression having side-effects, incorrect operation
229 resulted. Use an inlineable function.
231 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
232 held open for a verify callout. Previously this wan not accounted for
233 and a corrupt onward SMTP conversation resulted.
235 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
236 Previously they were embedded naively in the constructed address; when
237 needed, strip the quoting and quote the entire local_part.
238 Also make the inbound_srs expansion condition handle quoting.
240 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
241 excluded, not matching the documentation.
243 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
244 was given for the sqlite_dbfile a trap resulted.
246 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
247 "name" argument resulted in a trap. There is no reason to disallow such;
248 this was a coding error.
250 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
251 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
252 spent suspended, ignoring the POSIX definition. Previously we assumed
253 it did and a constant offset from real time could be used as a correction.
254 Change to using the same clock source for the start-of-message and the
255 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
256 exists, just to get a clock slightly more aligned to reality.
258 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
259 RFC says it is optional some validators care. The missing char was not
260 intended but triggered by a line-wrap alignment. Discovery and fix by
261 Guillaume Outters, hacked on by JH.
263 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
264 name being quoted was tainted a trap would be taken. Fix by using
265 dynamically created buffers. The routine could have been called by a
266 rewrite with the "h" flag, by using the "-F" command-line option, or
267 by using a "name=" option on a control=submission ACL modifier.
269 JH/19 SPF: change the Authentication-Results expansion component to give
270 smtp.helo when the sender domain is empty. Previously it gave
273 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
274 not in 127.0.0.0/8 to help in spotting list domains taken over by a
275 domain-parking registrar.
277 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
278 Previously when a whitespace character was specified it was not inserted
279 after removing the newline.
281 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
282 the domain part of the recipient address. This overrides any tls_sni
283 option set, which was previously used.
285 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
288 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
289 is_tainted() had an off-by-one error in the overenthusiastic direction.
290 Find and fix by Gavan. Although NetBSD is not a supported platform for
291 4.94 this bug could affect other platforms.
293 PP/01 Fix default prime selection to be consistent.
294 One path used ike23 still, instead of exim.dev.20160529.3; now both
295 execution flows will use the same DH primes (currently
296 exim.dev.20160529.3).
298 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
299 option in smtp transport, to match the documentation. Previously
300 verification was not being done in some cases where it should have been.
302 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
303 than one server was defined and depending on the platform memory layout
304 details, an internal consistency trap could be hit while walking the list
307 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
308 files. The value is used by the authresults expansion item. Previously
309 if this was used in a router or transport, a crash could result.
311 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
312 have a sufficient incidence of EINTR returns from select that an
313 interaction with other operations done by the main daemon loop exposed
314 a bug in the error-handling. This was benign apart from the log
317 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
318 for consistency with delivered "=>" and failed "**" lines. While we're
319 there, handle PRX and TFO.
321 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
322 applied. This resulted, if any header-line rewrite rules were configured,
323 in a panic-log triggerable by sending a message with a long address in
324 a header. Fix by increasing the arbitrary limit to larger than a single
325 (dewrapped) 5322 header line maximum size.
327 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
328 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
329 RFC 6648 which deprecates X- options in protocols as a general practice.
330 Changeover between the implementations is handled by the mechanisms
333 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
334 a remote which uses callout/hold. Previously the local was queued.
336 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
339 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
340 log line. The asterisk is now consistently appended to the remote IP
341 (and port, if given), and will also be provided on defer and fail log
342 lines. Previously it could be placed on the local IP if that was being
343 logged, and was only provided on delivery lines.
345 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
347 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
348 authenticator client_send option. Previously the next char, after a pair
349 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
350 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
351 documentation. There is still no way to get a leading ^ immediately
352 after a NUL (ie. for the password of a PLAIN method authenticator.
354 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
355 files. For bad sizes read, delete the record and whine to paniclog.
357 JH/38 When logging an AUTH failure, as server, do not include sensitive
358 information. Previously, the credentials would be included if given
359 as part of the AUTH command line and an ACL denied authentication.
361 JH/39 Bug 2691: fix $local_part_data. When the matching list element
362 referred to a file, bad data was returned. This likely also affected
365 JH/40 The gsasl authenticator now supports caching of the salted password
366 generated by the client-side implementation. This required the addition
367 of a new variable: $auth4.
369 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
370 left undeleted; the attempt to re-create it then failed - resulting in
371 the usual "SIGHUP tp have daemon reload configuration" to not work.
372 This affected any platform not supporting "abstract" Unix-domain
373 sockets (i.e. not Linux).
375 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
376 recipients" response to RCPT in a later response, with a 250. The
377 previous coding assumed this would not happen, and under PIPELINING
378 would result in both lost and duplicate recipients for a message.
380 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
381 Previously the weighting was incorrectly applied. Similar fix for socks
382 proxies. Found and fixed by Heiko Schlichting.
384 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
385 not handle sub-lists included using the +namedlist syntax. While
386 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
387 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
390 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
391 Previously the permanent pool was used, so the sore could not be freed.
392 This meant a connection with many messages would use continually-growing
395 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
396 per-pool so as not to waste too much space. Previously a constant size
397 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
398 DOS attacks harder. The cost is wasted memory use in the larger blocks.
400 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
401 same space repeatedly, and a smaller process footprint.
403 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
404 for GnuTLS. Previously the name started "%s" which, while not a bug,
405 looked as if if might be one.
407 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
408 "continued connection") make the $proxy_* variables available. Previously
409 the information was not passed across the exec() call for subsequent
410 transport executions. This also mean that the log lines for the
411 messages can show the proxy information.
413 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
414 receive time. With modern systems the difference is significant.
415 The historical behaviour can be restored by disabling (a new) log_selector
416 "queue_time_exclusive".
418 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
419 content) were specifically tested for. Now, also cover expansions
420 resulting in ACL names and inline ACL content.
422 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
423 making it unusable in complex expressions.
425 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
426 fail with a taint-check on the spoolfile name, and leave the message
429 HS/01 Enforce absolute PID file path name.
431 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
433 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
435 PP/02 Bug 2643: Correct TLS DH constants.
436 A missing NUL termination in our code-generation tool had led to some
437 incorrect Diffie-Hellman constants in the Exim source.
438 Reported by kylon94, code-gen tool fix by Simon Arlott.
440 PP/03 Impose security length checks on various command-line options.
441 Fixes CVE-2020-SPRSS reported by Qualys.
443 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
444 better. Reported by Qualys.
446 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
447 providing a particularly obnoxious sender full name.
450 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
452 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
455 PP/08 Change default for recipients_max from unlimited to 50,000.
457 PP/09 Fix security issue with too many recipients on a message (to remove a
458 known security problem if someone does set recipients_max to unlimited,
459 or if local additions add to the recipient list).
460 Fixes CVE-2020-RCPTL reported by Qualys.
462 PP/10 Fix security issue in SMTP verb option parsing
463 Fixes CVE-2020-EXOPT reported by Qualys.
465 PP/11 Fix security issue in BDAT state confusion.
466 Ensure we reset known-good where we know we need to not be reading BDAT
467 data, as a general case fix, and move the places where we switch to BDAT
468 mode until after various protocol state checks.
469 Fixes CVE-2020-BDATA reported by Qualys.
471 HS/03 Die on "/../" in msglog file names
473 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
474 the Exim runtime user are allowed to create files.
476 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
479 QS/03 When reading the output from interpreted forward files we do not
480 pass the pipe between the parent and the interpreting process to
481 executed child processes (if any).
483 QS/04 Always die if requested from internal logging, even is logging is
486 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
487 the API; compilation noo longer completes with DMARC support included.
488 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
489 been reported on other platforms.
491 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
493 JH/56 Make the majority of info read from config files readonly, for defence-in-
494 depth against exploits. Suggestion by Qualys.
495 Not supported on Solaris 10.
497 JH/57 Fix control=fakreject for a custom message containing tainted data.
498 Previously this resulted in a log complaint, due to a re-expansion present
499 since fakereject was originally introduced.
501 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
502 modifier was given, a loop resulted.
504 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
505 pipelined SMTP command followed the BDAT LAST then it would be
506 incorrectly treated as part of the message body, causing a verification
509 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
510 require looser validation rules than those for 821-level addresses,
511 which only permit IP addresses.
517 JH/01 Avoid costly startup code when not strictly needed. This reduces time
518 for some exim process initialisations. It does mean that the logging
519 of TLS configuration problems is only done for the daemon startup.
521 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
523 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
524 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
527 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
528 DKIM signing is being done. Previously a transport_filter would always
529 disable CHUNKING, falling back to traditional DATA.
531 JH/05 Regard command-line recipients as tainted.
533 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
535 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
536 PAM library frees one of the arguments given to it, despite the
537 documentation. Therefore a plain malloc must be used.
539 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
540 on-stack buffers were used, resulting in a taint trap when DSN information
541 copied from a received message was written into the buffer.
543 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
544 the ordering of its ARC headers. This caused a crash.
546 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
547 a new record was being constructed with information from the peer, a trap
550 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
551 installation would get error messages from DMARC verify, when it hit the
552 nonexistent file indicated by the default. Distros wanting DMARC enabled
553 should both provide the file and set the option.
554 Also enforce no DMARC verification for command-line sourced messages.
556 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
557 could, depending on the platform, hang at the STARTTLS response.
559 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
560 message on a connection. Previously if one message had ARC headers and
561 the following one did not, a crash could result when adding an
562 Authentication-Results: header.
564 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
565 Exim main code and Exim-related utities. The introduction of taint
566 tracking also did many adjustments to string handling. Since then, eximon
567 frequently terminated with an assert failure.
569 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
570 check for 452 responses. This slightly helps the inefficieny of doing
571 a large alias-expansion into a recipient-limited target. The max_rcpt
572 transport option still applies (and at the current default, will override
573 the new feature). The check is done for either cause of synch, and forces
574 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
575 connection. The new facility is not tunable at this time.
577 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
578 library live data was being used, so the results became garbage. Make
579 copies while it is still usable.
581 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
582 on delivery deferred (==) and failed (**) lines (if a delivery was
583 attemtped). Previously it was only on completion (=>) lines.
585 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
586 for the expansion of the server_scram_iter and server_scram_salt options.
588 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
589 are now specifically given a NO_DATA response without hitting the system
590 resolver. The library goes on to do the now-standard TXT lookup.
591 Use of dnsdb lookups is not affected.
593 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
594 only retrieve the errormessage once. Previously two calls to dlerror()
595 were used, and the second one (for mainlog/paniclog) retrieved null
598 JH/20 Taint checking: disallow use of tainted data for
599 - the appendfile transport file and directory options
600 - the pipe transport command
601 - the autoreply transport file, log and once options
602 - file names used by the redirect router (including filter files)
604 - paths used by single-key lookups
605 Previously this was permitted.
607 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
608 adjusted the size of a major service buffer; this failed because the
609 buffer was in use at the time. Change to a compile-time increase in the
610 buffer size, when this authenticator is compiled into exim.
612 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
613 previous fast-mode was untenable in the face of glibs using mmap to
614 support larger malloc requests.
616 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
617 New values supported, if defined on system where compiled:
618 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
619 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
621 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
622 running a limited number of proceses in parallel, a benefit is gained. The
623 amount varies with the platform hardware and load. The use of the option
624 queue_run_in_order means we cannot do this, as ordering becomes
627 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
628 had introduced a string-copy (for ensuring NUL-termination) which was not
629 appropriate for that case, which can include embedded NUL bytes in the
630 block of data. Investigation showed the copy to actually be needless, the
631 data being length-specified.
633 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
634 done during a receiving connection, and both used TLS, global info was
635 used rather than per-connection info for tracking the state of data
636 queued for transmission. This could result in a connection hang.
638 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
639 Previously, when delivering serveral messages down a single connection
640 only the first would provide a SIZE. This was due to the size information
641 not being properly tracked.
643 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
644 TAI (at 37 seconds currently), pretend to be in UTC for time-related
645 expansion and logging. Previously, spurious values such as a future
646 minute could be seen.
648 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
649 it could crash from a null-deref. This could also affect the
650 ${addresses: } operator and ${readsock } item.
652 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
653 message following a mime one, the variable was not reset.
655 JH/30 When an pipelined-connect fails at the first response, assume incorrect
656 cached capability (perhaps the peer reneged?) and immediately retry in
659 JH/31 Fix spurious detection of timeout while writing to transport filter.
661 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
662 an attempt to copy the string was made before checking it.
664 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
665 taint of the lookup key was maintained; we now regard the presence in the
666 filesystem as sufficient validation.
668 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
669 argument is supplied.
671 JH/35 The dsearch lookup now requires that the directory is an absolute path.
672 Previously this was not checked, and nonempty relative paths made an
673 access under Exim's current working directory.
675 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
676 Previously no event was raised.
678 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
679 parameter supplied by the sender MAIL FROM command. Previously it was
680 ignored, and only the check_spool_space option value for the required
683 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
684 the size of the signing public-key. Previously it was instead giving
685 the size of the signature hash.
687 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
688 the default. See the (new) dkim_verify_min_keysizes option.
690 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
691 and an ACL use a lookup for checking either the local_part or domain,
692 stale data could be accessed. Ensure that variable references are
693 dropped between messages.
695 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
696 by the client was not checked as pointing within response data before
697 being used. A malicious client could thus cause an out-of-bounds read and
698 possibly gain authentication. Fix by adding the check.
700 JH/42 Internationalisation: change the default for downconversion in the smtp
701 transport to be "if needed". Previously it was "as previously set" for
702 the message, which usually meant "if needed" for message-submission but
703 "no" for everything else. However, MTAs have been seen using SMTPUTF8
704 even when the envelope addresses did not need it, resulting in forwarding
705 failures to non-supporting MTAs. A downconvert in such cases will be
706 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
707 transport. The change does mean that addresses needing conversion will
708 be converted when previously a delivery failure would occur.
710 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
711 response was received it would be used unchecked in a fail-DSN, violating
712 standards on line-length limits. Truncate if needed.
714 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
715 doesn't work. (Additionally add a new main config option to configure the
722 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
723 side, to decode a TLS 1.3 packet capture.
725 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
726 Previously the default library behaviour applied, sending two, each in
729 JH/03 Debug output for ACL now gives the config file name and line number for
732 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
734 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
736 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
737 buffer overrun for (non-chunking) other transports.
739 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
740 TLS1.3, means that a server rejecting a client certificate is not visible
741 to the client until the first read of encrypted data (typically the
742 response to EHLO). Add detection for that case and treat it as a failed
743 TLS connection attempt, so that the normal retry-in-clear can work (if
744 suitably configured).
746 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
747 and/or domain. Found and fixed by Jason Betts.
749 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
750 configuration). If a CNAME target was not a wellformed name pattern, a
753 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
754 the OS reports them interleaved with other addresses.
756 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
757 used both for input and for a verify callout, both encrypted, SMTP
758 responses being sent by the server could be lost. This resulted in
759 dropped connections and sometimes bounces generated by a peer sending
762 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
763 string. Previously it was possible to cause undefined behaviour in a
764 library routine (usually a crash). Found by "zerons".
766 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
769 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
770 API was removed, so update to use the newer ones.
772 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
773 any timeout set, is taking a long time. Previously we would hang on to a
774 rotated logfile "forever" if the input was arriving with long gaps
775 (a previous attempt to fix addressed lack, for a long time, of initial
778 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
779 shared (NFS) environment. The length of the tempfile name is now
780 4 + 16 ("hdr.$message_exim_id") which might break on file
781 systems which restrict the file name length to lower values.
784 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
785 shared (NFS) environment.
787 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
788 did for all versions <4.90). Notably -M, -m, --invert, -I may be
791 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
792 on some platforms for bit 31.
794 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
795 to changes apparently associated with TLS1.3 handling some of the APIs
796 previously used were either nonfunctional or inappropriate. Strings
797 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
798 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
799 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
800 This affects log line X= elements, the $tls_{in,out}_cipher variables,
801 and the use of specific cipher names in the encrypted= ACL condition.
803 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
805 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
806 verification result was not updated unless hosts_require_ocsp applied.
808 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
809 queue_list_requires_admin set to false, non-admin users were denied the
812 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
813 directory-of-certs mode. Previously they were advertised despite the
816 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
817 A single TCP connection by a client will now hold a TLS connection open
818 for multiple message deliveries, by default. Previously the default was to
821 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
822 default. If built with the facility, DANE will be used. The facility
823 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
825 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
826 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
827 must be defined and you must still, unless you define DISABLE_TLS, manage
828 the the include-dir and library-file requirements that go with that
829 choice. Non-TLS builds are still supported.
831 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
834 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
835 default. If the platform supports and has the facility enabled, it will
836 be requested on all coneections.
838 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
839 controlled by the build-time option SUPPORT_PIPE_CONNECT.
841 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
843 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
844 success-DSN messages. Previously the From: header was always the default
845 one for these; the option was ignored.
847 JH/28 Fix the timeout on smtp response to apply to the whole response.
848 Previously it was reset for every read, so a teergrubing peer sending
849 single bytes within the time limit could extend the connection for a
850 long time. Credit to Qualsys Security Advisory Team for the discovery.
852 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
853 delivery address, which leaked information of the results of local
854 forwarding. Change to the original envelope recipient address, per
857 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
858 requested. Previously not bounce was generated and a log entry of
859 error ignored was made.
861 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
863 JH/32 Introduce a general tainting mechanism for values read from the input
864 channel, and values derived from them. Refuse to expand any tainted
865 values, to catch one form of exploit.
867 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
868 was unused and the unexpanded text used for the test. Found and
869 fixed by Ruben Jenster.
871 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
872 an attempt to use a TLS library read routine dereffed a nul pointer,
875 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
876 them in our smtp response.
878 JH/36 Have the general router option retry_use_local_part default to true when
879 any of the restrictive preconditions are set (to anything). Previously it
880 was only for check_local user. The change removes one item of manual
881 configuration which is required for proper retries when a remote router
882 handles a subset of addresses for a domain.
884 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
885 link count into consideration.
887 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
888 caused the extension of big_buffer, the following lines were ignored.
890 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
891 accordance with RFC 2308. Previously there was no expiry, so a longlived
892 receive process (eg. due to ACL delays) versus a short SOA value could
895 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
897 JH/39 Promote DMARC support to mainline.
899 JH/40 Bug 2452: Add a References: header to DSNs.
901 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
902 parameters. The relevant library call is documented as "Deprecated: This
903 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
904 3.6.0, DH parameters are negotiated following RFC7919."
906 HS/06 Change the default of dnssec_request_domains to "*"
908 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
909 carried on and emitted a BDAT command, even when PIPELINING was not
912 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
913 buffer was used for the filename, resulting in a trap when tainted
914 arguments (eg. $domain) were used.
916 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
917 recommended to avoid a possible server-load attack. The feature can be
918 re-enabled via the openssl_options main cofiguration option.
920 JH/45 local_scan API: documented the current smtp_printf() call. This changed
921 for version 4.90 - adding a "more data" boolean to the arguments.
922 Bumped the ABI version number also, this having been missed previously;
923 release versions 4.90 to 4.92.3 inclusive were effectively broken in
924 respect of usage of smtp_printf() by either local_scan code or libraries
925 accessed via the ${dlfunc } expansion item. Both will need coding
926 adjustment for any calls to smtp_printf() to match the new function
927 signature; a FALSE value for the new argument is always safe.
929 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
930 the file-offset (which the Linux syscall does, and exim expects); this
931 resulted in an indefinite loop.
933 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
934 to do ARC verification. The Authentication-Results: header line added
935 by the configuration then had no ARC item.
937 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
938 an incorrect mode was used for reading the result, resulting in it being
945 JH/01 Remove code calling the customisable local_scan function, unless a new
946 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
948 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
949 non-signal-safe functions being used.
951 JH/03 Bug 2269: When presented with a received message having a stupidly large
952 number of DKIM-Signature headers, disable DKIM verification to avoid
953 a resource-consumption attack. The limit is set at twenty.
955 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
956 report of oldest_pass in ${authres } in consequence, and separate out
957 some descriptions of reasons for verification fail.
959 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
960 files in the spool were present and unlocked. A queue-runner could spot
961 them, resulting in a duplicate delivery. Fix that by doing the unlock
962 after the unlink. Investigation by Tim Stewart. Take the opportunity to
963 add more error-checking on spoolfile handling while that code is being
966 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
967 No known attacks, no CVE, this is defensive hardening.
969 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
970 a queue-runner could start a delivery while other operations were ongoing.
971 Cutthrough delivery was a common victim, resulting in duplicate delivery.
972 Found and investigated by Tim Stewart. Fix by using the open message data
973 file handle rather than opening another, and not locally closing it (which
974 releases a lock) for that case, while creating the temporary .eml format
975 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
977 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
978 $sender_verify_failure/$recipient_verify_failure to "random".
980 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
983 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
984 Previously this would segfault.
986 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
989 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
990 like zero, since the resolver should be doing this for us, But we need one
991 as a CNAME but no MX presence gets the CNAME returned; we need to check
992 that doesn't point to an MX to declare it "no MX returned" rather than
993 "error, loop". A new main option is added so the older capability of
994 following some limited number of chain links is maintained.
996 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
998 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
999 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1000 now-deprecated earlier definitions used only specified the range up to TLS
1001 1.2 (in the older-version library docs).
1003 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1005 JH/15 Rework TLS client-side context management. Stop using a global, and
1006 explicitly pass a context around. This enables future use of TLS for
1007 connections to service-daemons (eg. malware scanning) while a client smtp
1008 connection is using TLS; with cutthrough connections this is quite likely.
1010 JH/16 Fix ARC verification to do AS checks in reverse order.
1012 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1014 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1015 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1016 Previously the "utf8" would be re-prepended for every additional message.
1018 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1019 Previously thery were accepted, resulting in issues when attempting to
1020 forward messages to a non-supporting MTA.
1022 PP/02 Let -n work with printing macros too, not just options.
1024 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1025 one parent address was copied, and bogus data was used at delivery-logging
1026 time. Either a crash (after delivery) or bogus log data could result.
1027 Discovery and analysis by Tim Stewart.
1029 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1030 Previously if the string ended mid-character, we did not insert the
1031 promised '?' replacement.
1033 PP/04 Documentation: current string operators work on bytes, not codepoints.
1035 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1036 should pack well giving a smaller memory footprint so better caching and
1037 therefore performance. Group the declarations where this can't be done so
1038 that the byte-sized flag variables are not interspersed among pointer
1039 variables, giving a better chance of good packing by the compiler.
1041 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1042 non-null, to avoid issues with sites running BATV. Previously reports were
1043 sent with an empty envelope sender so looked like bounces.
1045 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1046 The ignore_error flag wasn't being returned from the filter subprocess so
1047 was not set for later routers. Investigation and fix by Matthias Kurz.
1049 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1050 and a msg:complete for the whole, when a message is manually removed using
1051 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1053 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1054 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1055 hope that is portable enough.
1057 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1058 requiring a known-CA anchor certificate; make it now rely entirely on the
1059 TLSA as an anchor. Checking the name on the leaf cert against the name
1060 on the A-record for the host is still done for TA (but not for EE mode).
1062 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1063 included in delivery lines for non-proxied connections, when compiled with
1064 SUPPORT_SOCKS and running with proxy logging enabled.
1066 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1067 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1068 move the existing event to fire before the normal logging of message
1069 failure so that custom logging is bracketed by normal logging.
1071 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1072 msg:fail:internal event. Developement by Matthias Kurz.
1074 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1075 far too small for todays use of crypto signatures stored there. Go all
1076 the way to the max DNS message size of 64kB, even though this might be
1077 overmuch for IOT constrained device use.
1079 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1080 copy a string over itself. The library routine is documented as not
1081 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1083 JH/32 For main options check_spool_space and check_inode_space, where the
1084 platform supports 64b integers, support more than the previous 2^31 kB
1085 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1086 the previous G, M, k.
1088 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1089 $authenticated_fail_id variable on authentication failure. Previously
1092 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1093 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1094 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1095 GNUTLS_SEC_PARAM_MEDIUM.
1097 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1098 side. Previously we would continue as if no SNI had been received.
1100 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1101 "<" character, which should be a list-separator specification, we walked
1102 off past the nul-terimation.
1104 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1105 causes) even when the retry time is not yet met. Previously they were
1106 not, meaning that when (say) an account was over-quota and temp-rejecting,
1107 and multiple senders' messages were queued, only one sender would get
1108 notified on each configured delay_warning cycle.
1110 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1112 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1113 cipher-suites, an error can be left on the stack even for a succeeding
1114 accept; this results in impossible error messages when a later operation
1117 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1118 return error codes indicating retry. Under TLS1.3 this becomes required.
1120 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1121 it only wrote the new authenticators, resulting in a lack of tracking of
1122 peer changes of ESMTP extensions until the next cache flush.
1124 JH/41 Fix the loop reading a message header line to check for integer overflow,
1125 and more-often against header_maxsize. Previously a crafted message could
1126 induce a crash of the recive process; now the message is cleanly rejected.
1128 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1129 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1135 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1136 When redis_servers is set to a list of > 1 element, and the Redis servers
1137 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1138 case of MOVED into a DEFER case instead, thus moving the query onto the
1139 next server in the list. For a cluster of N elements, all N servers must
1140 be defined in redis_servers.
1142 GF/02 Catch and remove uninitialized value warning in exiqsumm
1143 Check for existence of @ARGV before looking at $ARGV[0]
1145 JH/01 Replace the store_release() internal interface with store_newblock(),
1146 which internalises the check required to safely use the old one, plus
1147 the allocate and data copy operations duplicated in both (!) of the
1148 extant use locations.
1150 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1151 modifier. This matches the restriction on the commandline.
1153 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1154 Previously only the last row was returned.
1156 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1157 we assumed that tags in the header were well-formed, and parsed the
1158 element content after inspecting only the first char of the tag.
1159 Assumptions at that stage could crash the receive process on malformed
1162 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1163 While running the DKIM ACL we operate on the Permanent memory pool so that
1164 variables created with "set" persist to the DATA ACL. Also (at any time)
1165 DNS lookups that fail create cache records using the Permanent pool. But
1166 expansions release any allocations made on the current pool - so a dnsdb
1167 lookup expansion done in the DKIM ACL releases the memory used for the
1168 DNS negative-cache, and bad things result. Solution is to switch to the
1169 Main pool for expansions.
1170 While we're in that code, add checks on the DNS cache during store_reset,
1171 active in the testsuite.
1172 Problem spotted, and debugging aided, by Wolfgang Breyha.
1174 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1175 When none of the hosts presented to a transport match an already-open
1176 connection, close it and proceed with the list. Previously we would
1177 queue the message. Spotted by Lena with Yahoo, probably involving
1180 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1181 Previously a spurious "250 OK id=" response was appended to the proper
1184 JH/08 The "support for" informational output now, which built with Content
1185 Scanning support, has a line for the malware scanner interfaces compiled
1186 in. Interface can be individually included or not at build time.
1188 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1189 by the template makefile "src/EDITME". The "STREAM" support for an older
1190 ClamAV interface method is removed.
1192 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1193 rows affected is given instead).
1195 JH/11 The runtime Berkeley DB library version is now additionally output by
1196 "exim -d -bV". Previously only the compile-time version was shown.
1198 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1199 SMTP connection. Previously, when one had more recipients than the
1200 first, an abortive onward connection was made. Move to full support for
1201 multiple onward connections in sequence, handling cutthrough connection
1202 for all multi-message initiating connections.
1204 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1205 routers. Previously, a multi-recipient message would fail to match the
1206 onward-connection opened for the first recipient, and cause its closure.
1208 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1209 a timeout on read on a GnuTLS initiating connection, resulting in the
1210 initiating connection being dropped. This mattered most when the callout
1211 was marked defer_ok. Fix to keep the two timeout-detection methods
1214 JH/15 Relax results from ACL control request to enable cutthrough, in
1215 unsupported situations, from error to silently (except under debug)
1216 ignoring. This covers use with PRDR, frozen messages, queue-only and
1219 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1221 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1222 metadata, resulting in a crash in free().
1224 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1225 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1226 Broken also in d185889f4, with init system revamp.
1228 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1229 Previously we abruptly closed the connection after reading a malware-
1230 found indication; now we go on to read the "scan ok" response line,
1233 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1234 ACL. Previously, a crash would result.
1236 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1237 macro text after a replacement (previously it was only once per line) and
1238 by skipping builtin macros when searching for an uppercase lead character.
1240 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1241 for the build is renamed.
1243 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1244 was allocated for every new TLS startup, meaning one per message. Fix
1245 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1247 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1248 reported the original. Fix to report (as far as possible) the ACL
1249 result replacing the original.
1251 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1252 OpenSSL. Certificate information is loaded for every new TLS startup,
1253 and the resources needed to be freed.
1255 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1257 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1260 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1261 DATA response info to the (existing) per-recipient response info for
1262 the "C=" log element. It can have useful tracking info from the
1263 destination system. Patch from Simon Arlott.
1265 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1266 length value. Previously this would segfault.
1268 HS/02 Support Avast multiline protoocol, this allows passing flags to
1269 newer versions of the scanner.
1271 JH/28 Ensure that variables possibly set during message acceptance are marked
1272 dead before release of memory in the daemon loop. This stops complaints
1273 about them when the debug_store option is enabled. Discovered specifically
1274 for sender_rate_period, but applies to a whole set of variables.
1275 Do the same for the queue-runner and queue-list loops, for variables set
1276 from spool message files. Do the same for the SMTP per-message loop, for
1277 certain variables indirectly set in ACL operations.
1279 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1280 as a multi-recipient message from a mailinglist manager). The coding had
1281 an arbitrary cutoff number of characters while checking for more input;
1282 enforced by writing a NUL into the buffer. This corrupted long / fast
1283 input. The problem was exposed more widely when more pipelineing of SMTP
1284 responses was introduced, and one Exim system was feeding another.
1285 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1286 receiving system, and refused recipients seen by the sending system
1287 (propating to people being dropped from mailing lists).
1288 Discovered and pinpointed by David Carter.
1290 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1291 replaced by the ${authresults } expansion.
1293 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1295 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1296 allows proper process termination in container environments.
1298 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1299 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1301 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1302 and "err_perm", deprecated since 4.83 when the RFC-defined words
1303 "temperror" and "permerror" were introduced.
1305 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1306 transport-filters or DKIM-signing. The restriction was lost in the
1307 consolidation of verify-callout and delivery SMTP handling.
1308 Extend the restriction to also cover ARC-signing.
1310 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1311 in defer=pass mode supply a 450 to the initiator. Previously the message
1314 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1315 tls_require_ciphers is used as before.
1317 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1318 "pass_unscanned". Only tmpfails from the scanner are written to
1319 the paniclog, as they may require admin intervention (permission
1320 denied, license issues). Other scanner errors (like decompression
1321 bombs) do not cause a paniclog entry.
1323 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1324 Previously it was possible to log spurious information in receive log
1327 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1328 triggered odd behaviour from Outlook Express clients.
1330 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1333 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1334 since the IETF WG has not yet settled on that versus the original
1335 "bare" representation.
1337 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1338 Previously the millisecond value corrupted the output.
1339 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1340 corrupted the output.
1346 JH/01 Rework error string handling in TLS interface so that the caller in
1347 more cases is responsible for logging. This permits library-sourced
1348 string to be attached to addresses during delivery, and collapses
1349 pairs of long lines into single ones.
1351 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1352 during configuration. Wildcards are allowed and expanded.
1354 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1355 This permits better logging.
1357 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1358 do not close it down and have the child transport start it up again on
1359 the passed-on TCP connection. Instead, proxy the child (and any
1360 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1361 affected: the continued delivery log lines do not have any DNSSEC, TLS
1362 Certificate or OCSP information. TLS cipher information is still logged.
1364 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1365 identical IP addresses on different listening ports. Will also affect
1368 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1369 add noisy ifdef guards to special-case this sillyness.
1370 Patch from Bernd Kuhls.
1372 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1373 than 255 are no longer allowed.
1375 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1376 Disable the session-cache too, which might reduce our load. Since we
1377 currrectly use a new context for every connection, both as server and
1378 client, there is no benefit for these.
1379 GnuTLS appears to not support tickets server-side by default (we don't
1380 call gnutls_session_ticket_enable_server()) but client side is enabled
1381 by default on recent versions (3.1.3 +) unless the PFS priority string
1384 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1385 <https://reproducible-builds.org/specs/source-date-epoch/>.
1387 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1388 the check for any unsuccessful recipients did not notice the limit, and
1389 erroneously found still-pending ones.
1391 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1392 MSG_MORE. Only in-clear (not on TLS connections).
1394 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1395 transport-filter is involved we can buffer the headers in memory for
1396 creating the signature, and read the spool data file once for the
1397 signature and again for transmission.
1399 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1400 4.77 as the kernel support then wasn't solid, having issues in 64bit
1401 mode. Now, it's been long enough. Add support for FreeBSD also.
1403 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1404 case where the routing stage had gathered several addresses to send to
1405 a host before calling the transport for the first, we previously failed
1406 to close down TLS in the old transport process before passing the TCP
1407 connection to the new process. The new one sent a STARTTLS command
1408 which naturally failed, giving a failed delivery and bloating the retry
1409 database. Investigation and fix prototype from Wolfgang Breyha.
1411 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1412 false-negatives in the check that the sender had not preempted a response
1413 or prompt from Exim (running as a server), due to that code's lack of
1414 awareness of the SMTP input buffering.
1416 PP/04 Add commandline_checks_require_admin option.
1417 Exim drops privileges sanely, various checks such as -be aren't a
1418 security problem, as long as you trust local users with access to their
1419 own account. When invoked by services which pass untrusted data to
1420 Exim, this might be an issue. Set this option in main configuration
1421 AND make fixes to the calling application, such as using `--` to stop
1424 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1425 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1426 responses to those, into a single TLS record each way (this usually means
1427 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1430 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1431 affects you only if you're dancing at the edge of the param size limits.
1432 If you are, and this message makes sense to you, then: raise the
1433 configured limit or use OpenSSL 1.1. Nothing we can do for older
1436 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1437 cmdline element to get the documented default one. Previously it was
1440 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1441 are unsuable for non-message-specific reasons - eg. connection timeout,
1442 banner-time rejection.
1444 JH/16 Fix logging of delivery remote port, when specified by router, under
1447 PP/06 Repair manualroute's ability to take options in any order, even if one
1448 is the name of a transport.
1451 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1453 JH/17 Change the list-building routines interface to use the expanding-string
1454 triplet model, for better allocation and copying behaviour.
1456 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1457 Previously it was constructed the first time a possibly-matching string
1458 was met in the configuration file input during startup; now it is done
1461 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1462 compatible one, to avoid the (poorly documented) possibility of a config
1463 file in the working directory redirecting the DB files, possibly correpting
1464 some existing file. CVE-2017-10140 assigned for BDB.
1466 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1467 cache-hot. Previously, although the result was properly cached, the
1468 initial verify call returned a defer.
1470 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1471 the main verify for receipient in uncached-mode.
1473 JH/22 Retire historical build files to an "unsupported" subdir. These are
1474 defined as "ones for which we have no current evidence of testing".
1476 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1477 if present. Previously it was ignored.
1479 JH/24 Start using specified-initialisers in C structure init coding. This is
1480 a C99 feature (it's 2017, so now considered safe).
1482 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1483 if was a fixed-sized field and bitmask ops via macros; it is now more
1486 PP/07 GitHub PR 56: Apply MariaDB build fix.
1487 Patch provided by Jaroslav Škarvada.
1489 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1490 during Coverity cleanups [4.87 JH/47]
1491 Diagnosis and fix provided by Michael Fischer v. Mollard.
1493 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1494 the right size to place the terminating semicolon on its own folded
1495 line, the header hash was calculated to an incorrect value thanks to
1496 the (relaxed) space the fold became.
1498 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1499 and confused the parent.
1501 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1502 which could crash as a result. This could lead to undeliverable messages.
1504 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1507 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1508 matters on fast-turnover and PID-randomising systems, which were getting
1509 out-of-order delivery.
1511 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1512 a possibly-overlapping copy. The symptom was that "Remote host closed
1513 connection in response to HELO" was logged instead of the actual 4xx
1516 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1517 Previously only that bufferd was discarded, resulting in SYMTP command
1520 JH/32 DKIM: when a message has multiple signatures matching an identity given
1521 in dkim_verify_signers, run the dkim acl once for each. Previously only
1522 one run was done. Bug 2189.
1524 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1525 "panic the current process" to "deliberately defer". The panic log is
1526 still written with the problem list name; the mail and reject logs now
1527 get a temp-reject line for the message that was being handled, saying
1528 something like "domains check lookup or other defer". The SMTP 451
1529 message is still "Temporary local problem".
1531 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1532 A crafted sequence of BDAT commands could result in in-use memory beeing
1533 freed. CVE-2017-16943.
1535 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1536 from SMTP input. Previously it was always done; now only done for DATA
1537 and not BDAT commands. CVE-2017-16944.
1539 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1540 to the message (such as an overlong header line). Previously this was
1541 not done and we did not exit BDAT mode. Followon from the previous item
1542 though a different problem.
1548 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1549 than -2003 did; needs libidn2 in addition to libidn.
1551 JH/02 The path option on a pipe transport is now expanded before use.
1553 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1554 Patch provided by "Björn", documentation fix added too.
1556 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1557 missing a wire-to-host endian conversion.
1559 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1560 close after a BDAT command line could be taken as a following command,
1561 giving a synch failure. Fix by only checking for synch immediately
1562 before acknowledging the chunk.
1564 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1565 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1566 macro. Patches provided by Josh Soref.
1568 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1569 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1570 by IANA as a service extension. However, John Klensin suggests that we
1573 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1574 the dkim code may be unix-mode line endings rather than smtp wire-format
1575 CRLF, so prepend a CR to any bare LF.
1577 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1578 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1580 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1581 the first were themselves being wrongly included in the feed into dkim
1582 processing; with most chunk sizes in use this resulted in an incorrect
1583 body hash calculated value.
1585 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1586 DKIM signature block, for verification. Although advised against by
1587 standards it is specifically not ruled illegal.
1589 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1591 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1592 missing a body hash (the bh= tag).
1594 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1595 It seems that HAProxy sends the Proxy Protocol information in clear and
1596 only then does a TLS startup, so do the same.
1598 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1599 TCP connections (such as for Spamd) unless the daemon successfully set
1600 Fast Open mode on its listening sockets. This fixes breakage seen on
1601 too-old kernels or those not configured for Fast Open, at the cost of
1602 requiring both directions being enabled for TFO, and TFO never being used
1603 by non-daemon-related Exim processes.
1605 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1606 endings, at least on the first header line. Try to canonify any that get
1607 past that check, despite the cost.
1609 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1610 now limited to an arbitrary five deep, while parsing addresses with the
1611 strip_excess_angle_brackets option enabled.
1613 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1614 instead leave the unprompted TLS handshake in socket buffer for the
1615 TLS library to consume.
1617 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1619 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1621 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1622 region in one operation, for speed, and this leaves assigned pointers
1623 dangling. Add checks run only under the testsuite which checks all
1624 variables at a store-reset and panics on a dangling pointer; add code
1625 explicitly nulling out all the variables discovered. Fixes one known
1626 bug: a transport crash, where a dangling pointer for $sending_ip_address
1627 originally assigned in a verify callout, is re-used.
1629 PP/06 Drop '.' from @INC in various Perl scripts.
1631 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1633 PP/08 Reduce a number of compilation warnings under clang; building with
1634 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1635 should be warning-free.
1637 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1639 HS/01 Fix portability problems introduced by PP/08 for platforms where
1640 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1642 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1643 chunk. This allows us to accept broken chunked messages. We need a more
1644 general solution here.
1646 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1647 already-broken messages in the queue.
1649 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1651 JH/19 Fix reference counting bug in routing-generated-address tracking.
1657 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1658 supports it and a size is available (ie. the sending peer gave us one).
1660 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1661 years of being deprecated). The replacements are the ACLs
1662 acl_smtp_mime and acl_not_smtp_mime.
1664 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1665 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1666 or even an in-clear connection were permitted. Now, if the host lookup
1667 was dnssec and dane was requested then the host is only used if the
1668 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1669 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1670 if one fails this test.
1671 This means that a poorly-configured remote DNS will make it incommunicado;
1672 but it protects against a DNS-interception attack on it.
1674 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1675 non-noisy when a race steals the message being considered.
1677 JH/05 If main configuration option tls_certificate is unset, generate a
1678 self-signed certificate for inbound TLS connections.
1680 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1681 in rewrites and routers.
1683 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1684 and logged a warning sing 4.83; now they are a configuration file error.
1686 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1687 (lacking @domain). Apply the same qualification processing as RCPT.
1689 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1691 JH/10 Support ${sha256:} applied to a string (as well as the previous
1694 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1695 a cutthrough deliver is pending, as we always want to make a connection.
1696 This also avoids re-routing the message when later placing the cutthrough
1697 connection after a verify cache hit.
1698 Do not update it with the verify result either.
1700 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1701 when routing results in more than one destination address.
1703 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1704 signing (which inhibits the cutthrough capability). Previously only
1705 the presence of an option was tested; now an expansion evaluating as
1706 empty is permissible (obviously it should depend only on data available
1707 when the cutthrough connection is made).
1709 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1710 the relevant preceding SMTP command did not note the pipelining mode.
1712 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1713 Previously they were not counted.
1715 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1716 as one having no matching records. Previously we deferred the message
1717 that needed the lookup.
1719 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1720 distinguished as "(=".
1722 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1723 for missing MX records. Previously it only worked for missing A records.
1725 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1727 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1728 after the data-go-ahead and data-ack. Patch from Jason Betts.
1730 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1731 even for a "none" policy. Patch from Tony Meyer.
1733 JH/22 Fix continued use of a connection for further deliveries. If a port was
1734 specified by a router, it must also match for the delivery to be
1737 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1738 When one of the recipients of a message was unsuitable for the connection
1739 (has no matching addresses), we lost track of needing to mark it
1740 deferred. As a result mail would be lost.
1742 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1744 JH/25 Decoding ACL controls is now done using a binary search; the source code
1745 takes up less space and should be simpler to maintain. Merge the ACL
1746 condition decode tables also, with similar effect.
1748 JH/26 Fix problem with one_time used on a redirect router which returned the
1749 parent address unchanged. A retry would see the parent address marked as
1750 delivered, so not attempt the (identical) child. As a result mail would
1753 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1754 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1755 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1758 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1759 Main config options check_{log,spool}_{inodes,space} are now
1760 100 inodes, 10MB unless set otherwise in the configuration.
1762 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1763 Previously it only applied to the main-section connection policy
1766 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1768 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1769 by me. Added RFC7919 DH primes as an alternative.
1771 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1772 are not in the system include path.
1774 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1775 GnuTLS, when a session startup failed (eg because the client disconnected)
1776 Exim did stdio operations after fclose. This was exposed by a recent
1777 change which nulled out the file handle after the fclose.
1779 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1780 signed directly by the cert-signing cert, rather than an intermediate
1781 OCSP-signing cert. This is the model used by LetsEncrypt.
1783 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1785 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1786 an incoming connection.
1788 HS/02 Bug 1802: Do not half-close the connection after sending a request
1791 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1792 fallback to "prime256v1".
1794 JH/34 SECURITY: Use proper copy of DATA command in error message.
1795 Could leak key material. Remotely exploitable. CVE-2016-9963.
1801 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1802 and 3.4.4 - once the server is enabled to respond to an OCSP request
1803 it does even when not requested, resulting in a stapling non-aware
1804 client dropping the TLS connection.
1806 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1807 support variable-length bit vectors. No functional change.
1809 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1810 The I= interface field on outgoing lines is now after the H= remote
1811 host field, same as incoming lines. There is a separate
1812 outgoing_interface log selector which allows you to disable the
1815 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1816 If not running log_selector +smtp_connection the mainlog would be held
1817 open indefinitely after a "too many connections" event, including to a
1818 deleted file after a log rotate. Leave the per net connection logging
1819 leaving it open for efficiency as that will be quickly detected by the
1820 check on the next write.
1822 HS/01 Bug 1671: Fix post transport crash.
1823 Processing the wait-<transport> messages could crash the delivery
1824 process if the message IDs didn't exist for some reason. When
1825 using 'split_spool_directory=yes' the construction of the spool
1826 file name failed already, exposing the same netto behaviour.
1828 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1829 mime_regex ACL conditions.
1831 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1832 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1833 to HELO, local diagnostic string.
1835 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1836 log line to debug. Even when configured with a tls authenticator many
1837 client connections are expected to not authenticate in this way, so
1838 an authenticate fail is not an error.
1840 HS/02 Add the Exim version string to the process info. This way exiwhat
1841 gives some more detail about the running daemon.
1843 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1844 matter for fast-change records such as DNSBLs.
1846 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1847 retry record, even if constant. There may be multiple transports with
1848 different interface settings and the retry behaviour needs to be kept
1851 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1853 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1855 JH/10 Bug 840: fix log_defer_output option of pipe transport
1857 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1858 hosts_randomize. This matters a lot when combined with mua_wrapper.
1860 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1861 ${quote_pgsql:<string>} operator.
1863 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1865 JH/14 Change header file name for freeradius-client. Relevant if compiling
1866 with Radius support; from the Gentoo tree and checked under Fedora.
1868 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1870 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1871 option transport_filter, meaning no filtering.
1873 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1875 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1876 defaults to "*" (all hosts). The variable is now available when not built
1877 with TLS, default unset, mainly to enable keeping the testsuite sane.
1878 If a server certificate is not supplied (via tls_certificate) an error is
1879 logged, and clients will find TLS connections fail on startup. Presumably
1880 they will retry in-clear.
1881 Packagers of Exim are strongly encouraged to create a server certificate
1882 at installation time.
1884 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1885 with the $config_file variable.
1887 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1888 in transport context, after the attempt, and per-recipient. The latter type
1889 is per host attempted. The event data is the error message, and the errno
1890 information encodes the lookup type (A vs. MX) used for the (first) host,
1891 and the trailing two digits of the smtp 4xx response.
1893 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1894 to write to mainlog (or rejectlog, paniclog) in the window between file
1895 creation and permissions/ownership being changed. Particularly affects
1896 installations where exicyclog is run as root, rather than exim user;
1897 result is that the running daemon panics and dies.
1899 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1901 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1902 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1903 "pri" and "weight". Note that the previous implicit priority given by the
1904 list order is no longer honoured.
1906 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1907 for DKIM processing.
1909 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1910 by defining SUPPORT_SOCKS.
1912 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1913 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1914 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1915 variables are renamed to proxy_{local,external}_{address,port}.
1917 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1918 for a build by defining SUPPORT_I18N
1920 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1921 of the query string, and make ${quote_redis:} do that quoting.
1923 JH/29 Move Events support from Experimental to mainline, enabled by default
1924 and removable for a build by defining DISABLE_EVENT.
1926 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1928 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1929 cached by the daemon.
1931 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1932 by defining LOOKUP_REDIS. The libhiredis library is required.
1934 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1935 keys are given for lookup.
1937 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1938 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1939 only supported when built with TLS support. The PolarSSL SHA routines
1940 are still used when the TLS library is too old for convenient support.
1942 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1943 openssl_options), for security. OpenSSL forces this from version 1.1.0
1944 server-side so match that on older versions.
1946 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1947 allocation for $value could be released as the expansion processing
1948 concluded, but leaving the global pointer active for it.
1950 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1951 and to use the domains and local_parts ACL conditions.
1953 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1954 incorrectly not doubled on cutthrough transmission, hence seen as a
1955 body-termination at the receiving system - resulting in truncated mails.
1956 Commonly the sender saw a TCP-level error, and retransmitted the message
1957 via the normal store-and-forward channel. This could result in duplicates
1958 received - but deduplicating mailstores were liable to retain only the
1959 initial truncated version.
1961 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1963 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1965 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1966 we're in there, support oversigning also; bug 1309.
1968 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1970 HS/04 Add support for keep_environment and add_environment options.
1972 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1973 either intentional arithmetic overflow during PRNG, or testing config-
1976 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1977 delivery resulted in actual delivery. Cancel cutthrough before DATA
1980 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1981 rejects a recipient: pass the reject to the originator.
1983 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1984 Many were false-positives and ignorable, but it's worth fixing the
1987 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1988 for the new environment-manipulation done at startup. Move the routines
1989 from being local to tls.c to being global via the os.c file.
1991 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1992 an extract embedded as result-arg for a map, the first arg for extract
1993 is unavailable so we cannot tell if this is a numbered or keyed
1994 extraction. Accept either.
2000 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2003 JH/02 The smtp transport option "multi_domain" is now expanded.
2005 JH/03 The smtp transport now requests PRDR by default, if the server offers
2008 JH/04 Certificate name checking on server certificates, when exim is a client,
2009 is now done by default. The transport option tls_verify_cert_hostnames
2010 can be used to disable this per-host. The build option
2011 EXPERIMENTAL_CERTNAMES is withdrawn.
2013 JH/05 The value of the tls_verify_certificates smtp transport and main options
2014 default to the word "system" to access the system default CA bundle.
2015 For GnuTLS, only version 3.0.20 or later.
2017 JH/06 Verification of the server certificate for a TLS connection is now tried
2018 (but not required) by default. The verification status is now logged by
2019 default, for both outbound TLS and client-certificate supplying inbound
2022 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2025 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2026 Status Notification (bounce) messages are now MIME format per RFC 3464.
2027 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2028 under the control of the dsn_advertise_hosts option, and routers may
2029 have a dsn_lasthop option.
2031 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2032 default, modifiable by a malware= option. The list separator for
2033 the options can now be changed in the usual way. Bug 68.
2035 JH/10 The smtp_receive_timeout main option is now expanded before use.
2037 JH/11 The incoming_interface log option now also enables logging of the
2038 local interface on delivery outgoing connections.
2040 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2041 if the interface and destination host and port all match.
2043 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2046 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2047 Patch from Andrew Lewis.
2049 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2050 now supports optional time-restrictions, weighting, and priority
2051 modifiers per server. Patch originally by <rommer@active.by>.
2053 JH/16 The spamd_address main option now supports a mixed list of local
2054 and remote servers. Remote servers can be IPv6 addresses, and
2055 specify a port-range.
2057 JH/17 Bug 68: The spamd_address main option now supports an optional
2058 timeout value per server.
2060 JH/18 Bug 1581: Router and transport options headers_add/remove can
2061 now have the list separator specified.
2063 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2066 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2069 JH/21 Support for the A6 type of dns record is withdrawn.
2071 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2072 rather than the verbs used.
2074 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2075 from 255 to 1024 chars.
2077 JH/24 Verification callouts now attempt to use TLS by default.
2079 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2080 are generic router options now. The defaults didn't change.
2082 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2083 Original patch from Alexander Shikoff, worked over by JH.
2085 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2086 files if ZCAT_COMMAND is not executable.
2088 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2090 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2092 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2093 Normally benign, it bites when the pair was led to by a CNAME;
2094 modern usage is to not canonicalize the domain to a CNAME target
2095 (and we were inconsistent anyway for A-only vs AAAA+A).
2097 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2099 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2100 when evaluating $sender_host_dnssec.
2102 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2103 $sender_helo_dnssec variable.
2105 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2107 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2109 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2111 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2112 documented as working, but never had. Support all but $spam_report.
2114 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2115 added for tls authenticator.
2117 HS/03 Add perl_taintmode main config option
2123 TL/01 When running the test suite, the README says that variables such as
2124 no_msglog_check are global and can be placed anywhere in a specific
2125 test's script, however it was observed that placement needed to be near
2126 the beginning for it to behave that way. Changed the runtest perl
2127 script to read through the entire script once to detect and set these
2128 variables, reset to the beginning of the script, and then run through
2129 the script parsing/test process like normal.
2131 TL/02 The BSD's have an arc4random API. One of the functions to induce
2132 adding randomness was arc4random_stir(), but it has been removed in
2133 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2134 function when detected.
2136 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2137 cause callback expansion.
2139 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2140 syntax errors in an expansion can be treated as a string instead of
2141 logging or causing an error, due to the internal use of bool_lax
2142 instead of bool when processing it.
2144 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2145 server certificates when making smtp deliveries.
2147 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2149 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2151 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2153 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2154 Merged patch from Sebastian Wiedenroth.
2156 JH/05 Fix results-pipe from transport process. Several recipients, combined
2157 with certificate use, exposed issues where response data items split
2158 over buffer boundaries were not parsed properly. This eventually
2159 resulted in duplicates being sent. This issue only became common enough
2160 to notice due to the introduction of connection certificate information,
2161 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2163 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2164 size buffer was used, resulting in syntax errors when an expansion
2167 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2168 version 3.3.6 or later.
2170 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2171 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2172 both become "event_action", the variables become $event_name, $event_data
2173 and $event_defer_errno. There is a new variable $verify_mode, usable in
2174 routers, transports and related events. The tls:cert event is now also
2175 raised for inbound connections, if the main configuration event_action
2178 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2179 early OCSP support, but no stapling (appears to be less than 1.0.0).
2181 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2182 server certificate names available under the smtp transport option
2183 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2186 JH/10 Time-related extraction expansions from certificates now use the main
2187 option "timezone" setting for output formatting, and are consistent
2188 between OpenSSL and GnuTLS compilations. Bug 1541.
2190 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2191 encoded parameter in the incoming message. Bug 1558.
2193 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2194 include certificate info, eximon was claiming there were spoolfile
2197 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2199 JH/14 Log delivery-related information more consistently, using the sequence
2200 "H=<name> [<ip>]" wherever possible.
2202 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2203 are problematic for Debian distribution, omit them from the release
2206 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2208 JH/16 Fix string representation of time values on 64bit time_t architectures.
2211 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2212 output list separator was used.
2217 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2218 checkers that were complaining about end of non-void function with no
2221 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2222 This was a regression introduced in 4.83 by another bugfix.
2224 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2226 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2227 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2233 TF/01 Correctly close the server side of TLS when forking for delivery.
2235 When a message was received over SMTP with TLS, Exim failed to clear up
2236 the incoming connection properly after forking off the child process to
2237 deliver the message. In some situations the subsequent outgoing
2238 delivery connection happened to have the same fd number as the incoming
2239 connection previously had. Exim would try to use TLS and fail, logging
2240 a "Bad file descriptor" error.
2242 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2243 utilities have not been installed.
2245 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2246 temporary space as the ACL may create new global variables.
2248 TL/01 LDAP support uses per connection or global context settings, depending
2249 upon the detected version of the libraries at build time.
2251 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2252 to extract and use the src ip:port in logging and expansions as if it
2253 were a direct connection from the outside internet. PPv2 support was
2254 updated based on HAProxy spec change in May 2014.
2256 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2258 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2259 Properly escape header and check for NULL return.
2261 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2264 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2266 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2267 characters in header names, implemented as a verify condition.
2268 Contributed by Michael Fischer v. Mollard.
2270 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2271 results permerror and temperror. Previous values are deprecated but
2272 still accepted. In a future release, err_perm and err_temp will be
2273 completely removed, which will be a backward incompatibility if the
2274 ACL tests for either of these two old results. Patch contributed by
2275 user bes-internal on the mailing list.
2277 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2279 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2280 selectors, in both main and reject logs.
2282 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2285 JH/07 Add malware type "sock" for talking to simple daemon.
2287 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2289 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2290 routers/transports under cutthrough routing.
2292 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2293 numbers. Touch up "bool" conditional to keep the same definition.
2295 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2297 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2299 JH/12 Expand items in router/transport headers_add or headers_remove lists
2300 individually rather than the list as a whole. Bug 1452.
2302 Required for reasonable handling of multiple headers_ options when
2303 they may be empty; requires that headers_remove items with embedded
2304 colons must have them doubled (or the list-separator changed).
2306 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2307 view the policy declared in the DMARC record. Currently, $dmarc_status
2308 is a combined value of both the record presence and the result of the
2311 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2313 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2314 dnslookup router and the smtp transport (applying to the forward
2317 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2318 of ldap servers used for a specific lookup. Patch provided by Heiko
2321 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2322 New variable $lookup_dnssec_authenticated for observability.
2324 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2325 Patch submitted by Lars Timman.
2327 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2329 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2330 Requires trusted mode and valid format message id, aborts otherwise.
2331 Patch contributed by Heiko Schlichting.
2333 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2334 certextract with support for various fields. Bug 1358.
2336 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2337 is requested by default, modifiable by smtp transport option
2340 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2341 operate on certificate variables to give certificate fingerprints
2342 Also new ${sha256:cert_variable}.
2344 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2346 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2347 Christian Aistleitner.
2349 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2351 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2352 file. Patch from Wolfgang Breyha.
2354 JH/25 Expand the coverage of the delivery $host and $host_address to
2355 client authenticators run in verify callout. Bug 1476.
2357 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2358 align with daemon_smtp_ports. Bug 72.
2360 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2361 support and error reporting did not work properly.
2363 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2364 and is readable. Patch from Andrew Colin Kissa.
2366 TL/14 Enhance documentation of ${run expansion and how it parses the
2367 commandline after expansion, particularly in the case when an
2368 unquoted variable expansion results in an empty value.
2370 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2372 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2373 by renaming away from it. Addresses GH issue 15, reported by
2376 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2378 TL/15 SECURITY: prevent double expansion in math comparison functions
2379 (can expand unsanitized data). Not remotely exploitable.
2386 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2388 PP/02 Make -n do something, by making it not do something.
2389 When combined with -bP, the name of an option is not output.
2391 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2394 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2395 $sender_host_name and config options to manage this, and basic check
2398 PP/05 DSCP support for outbound connections and control modifier for inbound.
2400 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2401 (Only plugin which currently uses this is kerberos4, which nobody should
2402 be using, but we should make it available and other future plugins might
2403 conceivably use it, even though it would break NAT; stuff *should* be
2404 using channel bindings instead).
2406 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2407 name; added for Sendmail compatibility; requires admin caller.
2408 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2409 just ignore it); requires trusted caller.
2410 Also parse but ignore: -Ac -Am -X<logfile>
2413 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2415 TL/02 Add +smtp_confirmation as a default logging option.
2417 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2418 Patch by Magnus Holmgren from 2007-02-20.
2420 TL/04 Bugzilla 1281 - Spec typo.
2421 Bugzilla 1283 - Spec typo.
2422 Bugzilla 1290 - Spec grammar fixes.
2424 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2426 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2428 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2429 mailing list by by Dmitry Isaikin.
2431 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2433 JH/02 Support "G" suffix to numbers in ${if comparisons.
2435 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2437 NM/01 Bugzilla 1197 - Spec typo
2438 Bugzilla 1196 - Spec examples corrections
2440 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2442 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2443 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2446 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2447 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2449 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2450 "acl {{name}{arg}...}", and optional args on acl condition
2453 JH/05 Permit multiple router/transport headers_add/remove lines.
2455 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2457 JH/07 Avoid using a waiting database for a single-message-only transport.
2458 Performance patch from Paul Fisher. Bugzilla 1262.
2460 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2463 JH/09 Add $headers_added variable, with content from use of ACL modifier
2464 add_header (but not yet added to the message). Bugzilla 199.
2466 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2467 Pulled from Bugzilla 817 by Wolfgang Breyha.
2469 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2471 (nb: this is the same fix as in Exim 4.80.1)
2473 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2476 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2479 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2481 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2482 advertise SMTP AUTH mechanism to us, instead of a generic
2483 protocol violation error. Also, make Exim more robust to bad
2484 data from the Dovecot auth socket.
2486 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2488 When a queue runner is handling a message, Exim first routes the
2489 recipient addresses, during which it prunes them based on the retry
2490 hints database. After that it attempts to deliver the message to
2491 any remaining recipients. It then updates the hints database using
2494 So if a recipient address works intermittently, it can get repeatedly
2495 deferred at routing time. The retry hints record remains fresh so the
2496 address never reaches the final cutoff time.
2498 This is a fairly common occurrence when a user is bumping up against
2499 their storage quota. Exim had some logic in its local delivery code
2500 to deal with this. However it did not apply to per-recipient defers
2501 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2503 This change adds a proper retry rule check during routing so that the
2504 final cutoff time is checked against the message's age. We only do
2505 this check if there is an address retry record and there is not a
2506 domain retry record; this implies that previous attempts to handle
2507 the address had the retry_use_local_parts option turned on. We use
2508 this as an approximation for the destination being like a local
2509 delivery, as in LMTP.
2511 I suspect this new check makes the old local delivery cutoff check
2512 redundant, but I have not verified this so I left the code in place.
2514 TF/02 Correct gecos expansion when From: is a prefix of the username.
2516 Test 0254 submits a message to Exim with the header
2520 When I ran the test suite under the user fanf2, Exim expanded
2521 the header to contain my full name, whereas it should have added
2522 a Resent-Sender: header. It erroneously treats any prefix of the
2523 username as equal to the username.
2525 This change corrects that bug.
2527 GF/01 DCC debug and logging tidyup
2528 Error conditions log to paniclog rather than rejectlog.
2529 Debug lines prefixed by "DCC: " to remove any ambiguity.
2531 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2533 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2534 Bug spotted by Jeremy Harris; was flawed since initial commit.
2535 Would have resulted in OCSP responses post-SNI triggering an Exim
2536 NULL dereference and crash.
2538 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2540 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2541 Bug detection, analysis and fix by Samuel Thibault.
2542 Bugzilla 1331, Debian bug #698092.
2544 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2546 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2547 Server implementation by Todd Lyons, client by JH.
2548 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2549 config variable "prdr_enable" controls whether the server
2550 advertises the facility. If the client requests PRDR a new
2551 acl_data_smtp_prdr ACL is called once for each recipient, after
2552 the body content is received and before the acl_smtp_data ACL.
2553 The client is controlled by both of: a hosts_try_prdr option
2554 on the smtp transport, and the server advertisement.
2555 Default client logging of deliveries and rejections involving
2556 PRDR are flagged with the string "PRDR".
2558 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2559 fclose(). Diagnosis by Todd Lyons.
2561 PP/17 Update configure.default to handle IPv6 localhost better.
2562 Patch by Alain Williams (plus minor tweaks).
2565 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2566 This is now consistent with GnuTLS, and is now documented: the
2567 previous undocumented portable approach to treating the option as
2568 unset was to force an expansion failure. That still works, and
2569 an empty string is now equivalent.
2571 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2572 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2573 not performing validation itself.
2575 PP/20 Added force_command boolean option to pipe transport.
2576 Patch from Nick Koston, of cPanel Inc.
2578 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2581 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2583 PP/21 Fix eximon continuous updating with timestamped log-files.
2584 Broken in a format-string cleanup in 4.80, missed when I repaired the
2585 other false fix of the same issue.
2586 Report and fix from Heiko Schlichting.
2589 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2590 Report from Prashanth Katuri.
2592 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2593 It's SecureTransport, so affects any MacOS clients which use the
2594 system-integrated TLS libraries, including email clients.
2596 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2597 using a MIME ACL for non-SMTP local injection.
2598 Report and assistance in diagnosis by Warren Baker.
2600 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2602 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2604 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2605 last id that failed so it may be referenced in subsequent ACL's.
2607 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2610 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2611 ldap library initialization, allowing self-signed CA's to be
2612 used. Also properly sets require_cert option later in code by
2613 using NULL (global ldap config) instead of ldap handle (per
2614 session). Bug diagnosis and testing by alxgomz.
2616 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2617 the src/util/ subdirectory.
2619 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2620 renamed to Transport Post Delivery Action by Jeremy Harris, as
2623 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2624 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2625 redis_servers = needs to be configured which will be used by the redis
2626 lookup. Patch from Warren Baker, of The Packet Hub.
2628 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2630 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2631 hostname or reverse DNS when processing a host list. Used suggestions
2632 from multiple comments on this bug.
2634 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2636 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2637 Missed a few lines, added it to make the runtest require no keyboard
2640 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2641 contains upper case chars. Make router use caseful_local_part.
2643 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2644 support when GnuTLS has been built with p11-kit.
2650 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2652 This, or similar/improved, will also be change PP/11 of 4.82.
2658 PP/01 Handle short writes when writing local log-files.
2659 In practice, only affects FreeBSD (8 onwards).
2660 Bugzilla 1053, with thanks to Dmitry Isaikin.
2662 NM/01 Bugzilla 949 - Documentation tweak
2664 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2667 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2669 PP/02 Implemented gsasl authenticator.
2671 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2673 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2674 `pkg-config foo` for cflags/libs.
2676 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2677 with rest of GSASL and with heimdal_gssapi.
2679 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2680 `pkg-config foo` for cflags/libs for the TLS implementation.
2682 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2683 properties get this fed in as external SSF. A number of robustness
2684 and debugging improvements to the cyrus_sasl authenticator.
2686 PP/08 cyrus_sasl server now expands the server_realm option.
2688 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2689 Patch by Jeremy Harris.
2691 PP/10 Added dbmjz lookup type.
2693 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2695 PP/12 MAIL args handles TAB as well as SP, for better interop with
2696 non-compliant senders.
2697 Analysis and variant patch by Todd Lyons.
2699 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2700 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2701 Patch from Dirk Mueller <dmueller@suse.com>
2703 PP/13 tls_peerdn now print-escaped for spool files.
2704 Observed some $tls_peerdn in wild which contained \n, which resulted
2705 in spool file corruption.
2707 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2708 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2709 or write after TLS renegotiation, which otherwise led to messages
2712 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2713 as a tracking header (ie: a signed header comes before the signature).
2714 Patch from Wolfgang Breyha.
2716 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2717 comma-sep list; embedded commas doubled.
2719 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2721 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2723 Report and patch from Dmitry Banschikov.
2725 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2726 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2727 failures. We appear to now support TLS1.1+ with Exim.
2729 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2730 lets Exim select keys and certificates based upon TLS SNI from client.
2731 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2732 before an outbound SMTP session. New log_selector, +tls_sni.
2734 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2735 NULL dereference. Report and patch from Alun Jones.
2737 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2738 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2739 Not seeing resolver debug output on NetBSD, but suspect this is a
2740 resolver implementation change.
2742 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2743 Left warnings. Added "eximon gdb" invocation mode.
2745 PP/21 Defaulting "accept_8bitmime" to true, not false.
2747 PP/22 Added -bw for inetd wait mode support.
2749 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2750 locate the relevant includes and libraries. Made this the default.
2752 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2753 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2755 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2756 This may cause build issues on older platforms.
2758 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2759 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2760 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2761 Added SNI support via GnuTLS too.
2762 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2764 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2766 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2767 Applied second patch from Janne, implementing suggestion to default
2768 multiple-strings-in-record handling to match SPF spec.
2770 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2772 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2773 read-only, out of scope).
2774 Patch from Wolfgang Breyha, report from Stuart Northfield.
2776 PP/29 Fix three issues highlighted by clang analyser static analysis.
2777 Only crash-plausible issue would require the Cambridge-specific
2778 iplookup router and a misconfiguration.
2779 Report from Marcin Mirosław.
2781 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2783 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2784 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2785 As part of this, removing so much warning spew let me fix some minor
2786 real issues in debug logging.
2788 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2789 assignment on my part. Fixed.
2791 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2792 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2793 Janne Snabb (who went above and beyond: thank you).
2795 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2796 string otherwise requires a connection and a bunch more work and it's
2797 relatively easy to get wrong. Should also expose TLS library linkage
2800 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2801 64-bit ${eval} (JH/03).
2803 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2804 GNU libc to support some of the 64-bit stuff, should not lead to
2805 conflicts. Defined before os.h is pulled in, so if a given platform
2806 needs to override this, it can.
2808 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2809 protection layer was required, which is not implemented.
2810 Bugzilla 1254, patch from Wolfgang Breyha.
2812 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2813 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2814 tls_dhparam take prime identifiers. Also unbreak combination of
2815 OpenSSL+DH_params+TLSSNI.
2817 PP/39 Disable SSLv2 by default in OpenSSL support.
2823 PP/01 Solaris build fix for Oracle's LDAP libraries.
2824 Bugzilla 1109, patch from Stephen Usher.
2826 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2828 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2831 TF/02 Fix a couple more cases where we did not log the error message
2832 when unlink() failed. See also change 4.74-TF/03.
2834 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2835 lock up or crash if it happened to be inside a call to libc when it
2836 got a SIGUSR1 from exiwhat.
2838 The SIGUSR1 handler appends the current process status to the process
2839 log which is later printed by exiwhat. It used to use the general
2840 purpose logging code to do this, but several functions it calls are
2841 not safe for signals.
2843 The new output code in the SIGUSR1 handler is specific to the process
2844 log, and simple enough that it's easy to inspect for signal safety.
2845 Removing some special cases also simplifies the general logging code.
2846 Removing the spurious timestamps from the process log simplifies
2849 TF/04 Improved ratelimit ACL condition.
2851 The /noupdate option has been deprecated in favour of /readonly which
2852 has clearer semantics. The /leaky, /strict, and /readonly update modes
2853 are mutually exclusive. The update mode is no longer included in the
2854 database key; it just determines when the database is updated. (This
2855 means that when you upgrade Exim will forget old rate measurements.)
2857 Exim now checks that the per_* options are used with an update mode that
2858 makes sense for the current ACL. For example, when Exim is processing a
2859 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2860 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2861 must specify per_mail/readonly. If you omit the update mode it defaults to
2862 /leaky where that makes sense (as before) or /readonly where required.
2864 The /noupdate option is now undocumented but still supported for
2865 backwards compatibility. It is equivalent to /readonly except that in
2866 ACLs where /readonly is required you may specify /leaky/noupdate or
2867 /strict/noupdate which are treated the same as /readonly.
2869 A useful new feature is the /count= option. This is a generalization
2870 of the per_byte option, so that you can measure the throughput of other
2871 aggregate values. For example, the per_byte option is now equivalent
2872 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2874 The per_rcpt option has been generalized using the /count= mechanism
2875 (though it's more complicated than the per_byte equivalence). When it is
2876 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2877 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2878 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2879 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2880 Note that using per_rcpt with a non-readonly update mode in more than
2881 one ACL will cause the recipients to be double-counted. (The per_mail
2882 and per_byte options don't have this problem.)
2884 The handling of very low rates has changed slightly. If the computed rate
2885 is less than the event's count (usually one) then this event is the first
2886 after a long gap. In this case the rate is set to the same as this event's
2887 count, so that the first message of a spam run is counted properly.
2889 The major new feature is a mechanism for counting the rate of unique
2890 events. The new per_addr option counts the number of different
2891 recipients that someone has sent messages to in the last time period. It
2892 behaves like per_rcpt if all the recipient addresses are different, but
2893 duplicate recipient addresses do not increase the measured rate. Like
2894 the /count= option this is a general mechanism, so the per_addr option
2895 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2896 example, measure the rate that a client uses different sender addresses
2897 with the options per_mail/unique=$sender_address. There are further
2898 details in the main documentation.
2900 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2902 TF/06 Removed a few PCRE remnants.
2904 TF/07 Automatically extract Exim's version number from tags in the git
2905 repository when doing development or release builds.
2907 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2908 Bugzilla 879. Patch from Paul Fisher.
2910 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2911 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2914 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2916 PP/05 Variable $av_failed, true if the AV scanner deferred.
2917 Bugzilla 1078. Patch from John Horne.
2919 PP/06 Stop make process more reliably on build failure.
2920 Bugzilla 1087. Patch from Heiko Schlittermann.
2922 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2923 Bugzilla 1089. Patch from Heiko Schlittermann.
2925 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2926 Bugzilla 1131. Patch from Holger Weiß.
2928 PP/09 Handle IPv6 addresses with SPF.
2929 Bugzilla 860. Patch from Wolfgang Breyha.
2931 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2933 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2936 PP/11 match_* no longer expand right-hand-side by default.
2937 New compile-time build option, EXPAND_LISTMATCH_RHS.
2938 New expansion conditions, "inlist", "inlisti".
2940 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2942 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2944 PP/14 fix log_write() format string regression from TF/03.
2945 Bugzilla 1152. Patch from Dmitry Isaikin.
2951 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2953 PP/02 Harmonised TLS library version reporting; only show if debugging.
2954 Layout now matches that introduced for other libraries in 4.74 PP/03.
2956 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2958 PP/04 New "dns_use_edns0" global option.
2960 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2963 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2964 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2966 TK/01 Updated PolarSSL code to 0.14.2.
2967 Bugzilla 1097. Patch from Andreas Metzler.
2969 PP/07 Catch divide-by-zero in ${eval:...}.
2970 Fixes bugzilla 1102.
2972 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2975 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2976 format-string attack -- SECURITY: remote arbitrary code execution.
2978 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2979 time unintentionally subject to list matching rules, letting the header
2980 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2981 arbitrary string expansion). This allowed for information disclosure.
2983 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2984 INT_MIN/-1 -- value coerced to INT_MAX.
2990 NM/01 Workaround for PCRE version dependency in version reporting
2993 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2994 This fixes portability to compilers other than gcc, notably
2995 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2997 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2998 makefiles for portability to HP-UX and POSIX correctness.
3000 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3001 Also via indented variable definition in the Makefile.
3002 (Debugging by Oliver Heesakkers).
3004 PP/02 Restore caching of spamd results with expanded spamd_address.
3005 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3007 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3008 Improves build reliability. Fix from: Frank Elsner
3010 NM/02 Fix wide character breakage in the rfc2047 coding
3011 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3013 NM/03 Allow underscore in dnslist lookups
3014 Fixes bug 1026. Patch from Graeme Fowler
3016 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3017 Code patches from Adam Ciarcinski of NetBSD.
3019 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3022 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3023 is logged, to avoid truncation. Patch from John Horne.
3025 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3026 Patch from Jakob Hirsch.
3028 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3029 SQL string expansion failure details.
3030 Patch from Andrey Oktyabrski.
3032 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3033 Patch from Simon Arlott.
3035 PP/09 New lookups functionality failed to compile on old gcc which rejects
3036 extern declarations in function scope.
3037 Patch from Oliver Fleischmann
3039 PP/10 Use sig_atomic_t for flags set from signal handlers.
3040 Check getgroups() return and improve debugging.
3041 Fixed developed for diagnosis in bug 927 (which turned out to be
3044 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3045 Patch from Mark Zealey.
3047 PP/12 Bugzilla 1056: Improved spamd server selection.
3048 Patch from Mark Zealey.
3050 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3051 Based on patch from Heiko Schlittermann.
3053 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3054 Patch from Uwe Doering, sign-off by Michael Haardt.
3056 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3057 variable declaration deep within a block. Bug and patch from
3060 PP/15 lookups-Makefile IRIX compatibility coercion.
3062 PP/16 Make DISABLE_DKIM build knob functional.
3064 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3065 Patch by Simon Arlott
3067 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3068 variable argument macros. Our copy now differs from upstream.
3074 TF/01 Failure to get a lock on a hints database can have serious
3075 consequences so log it to the panic log.
3077 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3078 controlled using the smtp_confirmation log selector.
3080 TF/03 Include the error message when we fail to unlink a spool file.
3082 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3083 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3084 for maintaining out-of-tree patches for some time.
3086 PP/01 Bugzilla 139: Documentation and portability issues.
3087 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3088 Handle per-OS dynamic-module compilation flags.
3090 PP/02 Let /dev/null have normal permissions.
3091 The 4.73 fixes were a little too stringent and complained about the
3092 permissions on /dev/null. Exempt it from some checks.
3093 Reported by Andreas M. Kirchwitz.
3095 PP/03 Report version information for many libraries, including
3096 Exim version information for dynamically loaded libraries. Created
3097 version.h, now support a version extension string for distributors
3098 who patch heavily. Dynamic module ABI change.
3100 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3101 privilege escalation vulnerability whereby the Exim run-time user
3102 can cause root to append content of the attacker's choosing to
3105 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3108 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3109 If dropping privileges for untrusted macros, we disabled normal logging
3110 on the basis that it would fail; for the Exim run-time user, this is not
3111 the case, and it resulted in successful deliveries going unlogged.
3112 Fixed. Reported by Andreas Metzler.
3118 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3119 only prepend for the Resent-* case. Fixes regression introduced in
3120 Exim 4.70 by NM/22 for Bugzilla 607.
3122 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3123 increasing numbers of administrators be bitten by this.
3125 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3127 PP/03 Bugzilla 994: added openssl_options main configuration option.
3129 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3131 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3133 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3135 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3136 without a peer certificate, leading to a segfault because of an
3137 assumption that peers always have certificates. Be a little more
3138 paranoid. Problem reported by Martin Tscholak.
3140 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3141 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3142 NB: ClamAV planning to remove STREAM in "middle of 2010".
3143 CL also introduces -bmalware, various -d+acl logging additions and
3144 more caution in buffer sizes.
3146 PP/09 Implemented reverse_ip expansion operator.
3148 PP/10 Bugzilla 937: provide a "debug" ACL control.
3150 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3152 PP/12 Bugzilla 973: Implement --version.
3154 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3156 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3158 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3160 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3161 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3162 ignore trailing whitespace.
3164 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3166 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3169 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3170 Notification from Dr Andrew Aitchison.
3172 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3173 ExtendedDetectionInfo response format.
3174 Notification from John Horne.
3176 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3179 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3180 XSL and documented dependency on system catalogs, with examples of how
3183 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3186 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3187 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3188 configuration file which is writeable by the Exim user or group.
3190 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3191 of configuration files to cover files specified with the -C option if
3192 they are going to be used with root privileges, not just the default
3195 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3196 option (effectively making it always true).
3198 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3199 files to be used while preserving root privileges.
3201 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3202 that rogue child processes cannot use them.
3204 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3205 run-time user, instead of root.
3207 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3208 Exim run-time user without dropping privileges.
3210 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3211 result string, instead of calling string_vformat() twice with the same
3214 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3215 for other users. Others should always drop root privileges if they use
3216 -C on the command line, even for a whitelisted configure file.
3218 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3220 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3226 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3227 $data_path, and $header_path variables; fixed documentation bugs and
3230 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3231 exipick to access non-standard spools, including the "frozen" queue
3234 NM/01 Bugzilla 965: Support mysql stored procedures.
3235 Patch from Alain Williams
3237 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3239 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3240 Patch from Andreas Metzler
3242 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3243 Patch from Kirill Miazine
3245 NM/05 Bugzilla 671: Added umask to procmail example.
3247 JJ/03 installed exipick 20100323.0, fixing doc bug
3249 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3250 directory. Notification and patch from Dan Rosenberg.
3252 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3254 TK/02 Improve log output when DKIM signing operation fails.
3256 MH/01 Treat the transport option dkim_domain as a colon separated
3257 list, not as a single string, and sign the message with each element,
3258 omitting multiple occurences of the same signer.
3260 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3261 Bugzilla 985, 986. Patch by Simon Arlott
3263 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3264 Patch by Simon Arlott
3266 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3267 MBX locking. Notification from Dan Rosenberg.
3273 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3275 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3277 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3279 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3281 NM/04 Bugzilla 915: Flag broken perl installation during build.
3287 TK/01 Added patch by Johannes Berg that expands the main option
3288 "spamd_address" if it starts with a dollar sign.
3290 TK/02 Write list of recipients to X-Envelope-Sender header when building
3291 the mbox-format spool file for content scanning (suggested by Jakob
3294 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3295 (http://www.dcc-servers.net/) support via dccifd. Activated by
3296 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3298 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3299 by Mark Daniel Reidel <mr@df.eu>.
3301 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3302 When building exim an external PCRE library is now needed -
3303 PCRE is a system library on the majority of modern systems.
3304 See entry on PCRE_LIBS in EDITME file.
3306 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3307 conversation. Added nologin parameter to request.
3308 Patch contributed by Kirill Miazine.
3310 TF/01 Do not log submission mode rewrites if they do not change the address.
3312 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3314 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3315 log files in place. Contributed by Roberto Lima.
3317 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3319 TF/03 Bugzilla 615: When checking the local_parts router precondition
3320 after a local_part_suffix or local_part_prefix option, Exim now
3321 does not use the address's named list lookup cache, since this
3322 contains cached lookups for the whole local part.
3324 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3325 Robert Millan. Documentation is in experimental-spec.txt.
3327 TF/04 Bugzilla 668: Fix parallel build (make -j).
3329 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3331 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3332 Patch provided by Jan Srzednicki.
3334 TF/05 Leading white space used to be stripped from $spam_report which
3335 wrecked the formatting. Now it is preserved.
3337 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3338 that they are available at delivery time.
3340 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3342 TF/08 TLS error reporting now respects the incoming_interface and
3343 incoming_port log selectors.
3345 TF/09 Produce a more useful error message if an SMTP transport's hosts
3346 setting expands to an empty string.
3348 NM/06 Bugzilla 744: EXPN did not work under TLS.
3349 Patch provided by Phil Pennock.
3351 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3352 Patch provided by Richard Godbee.
3354 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3355 acl_smtp_notquit, added index entry.
3357 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3358 Patch provided by Eugene Bujak.
3360 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3361 accept(). Patch provided by Maxim Dounin.
3363 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3364 Patch provided by Phil Pennock.
3366 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3368 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3369 Patch provided by Brad "anomie" Jorsch.
3371 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3372 Patch provided by Dean Brooks.
3374 TK/05 Add native DKIM support (does not depend on external libraries).
3376 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3377 Patch provided by Graeme Fowler.
3379 NM/16 Bugzilla 851: Documentation example syntax fix.
3381 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3383 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3386 NM/19 Bugzilla 745: TLS version reporting.
3387 Patch provided by Phil Pennock.
3389 NM/20 Bugzilla 167: bool: condition support.
3390 Patch provided by Phil Pennock.
3392 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3393 clients. Patch provided by Phil Pennock.
3395 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3396 Patch provided by Brad "anomie" Jorsch.
3398 NM/23 Bugzilla 687: Fix misparses in eximstats.
3399 Patch provided by Heiko Schlittermann.
3401 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3402 Patch provided by Heiko Schlittermann.
3404 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3405 plus update to original patch.
3407 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3409 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3410 Patch provided by David Brownlee.
3412 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3414 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3416 NM/30 Bugzilla 888: TLS documentation bugfixes.
3418 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3420 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3421 Unlike the original bugzilla I have changed all shell scripts in src tree.
3423 NM/33 Bugzilla 898: Transport filter timeout fix.
3424 Patch by Todd Rinaldo.
3426 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3427 Patch by Serge Demonchaux.
3429 NM/35 Bugzilla 39: Base64 decode bug fixes.
3430 Patch by Jakob Hirsch.
3432 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3434 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3436 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3438 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3444 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3445 ALT-N's libdkim that I have put here:
3446 http://duncanthrax.net/exim-experimental/
3448 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3449 were called 'true' and it seems that C99 defines that as a reserved
3450 keyword to be used with 'bool' variable types. That means you could
3451 not include C99-style headers which use bools without triggering
3452 build errors in sieve.c.
3454 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3455 as mailq or other aliases. Changed the --help handling significantly
3456 to do whats expected. exim_usage() emits usage/help information.
3458 SC/01 Added the -bylocaldomain option to eximstats.
3460 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3462 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3464 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3470 PH/01 Another patch from the Sieve maintainer.
3472 PH/02 When an IPv6 address is converted to a string for single-key lookup
3473 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3474 dots are used instead of colons so that keys in lsearch files need not
3475 contain colons. This was done some time before quoting was made available
3476 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3477 (notated using the quote facility) so as to distinguish them from IPv4
3478 keys. This meant that lookups for IP addresses in host lists did not work
3479 for iplsearch lookups.
3481 This has been fixed by arranging for IPv6 addresses to be expressed with
3482 colons if the lookup type is iplsearch. This is not incompatible, because
3483 previously such lookups could never work.
3485 The situation is now rather anomalous, since one *can* have colons in
3486 ordinary lsearch keys. However, making the change in all cases is
3487 incompatible and would probably break a number of configurations.
3489 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3492 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3493 conversion specification without a maximum field width, thereby enabling
3494 a rogue spamd server to cause a buffer overflow. While nobody in their
3495 right mind would setup Exim to query an untrusted spamd server, an
3496 attacker that gains access to a server running spamd could potentially
3497 exploit this vulnerability to run arbitrary code as the Exim user.
3499 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3500 $primary_hostname instead of what libspf2 thinks the hosts name is.
3502 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3503 a directory entry by the name of the lookup key. Previously, if a
3504 symlink pointed to a non-existing file or a file in a directory that
3505 Exim lacked permissions to read, a lookup for a key matching that
3506 symlink would fail. Now it is enough that a matching directory entry
3507 exists, symlink or not. (Bugzilla 503.)
3509 PH/03 The body_linecount and body_zerocount variables are now exported in the
3512 PH/04 Added the $dnslist_matched variable.
3514 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3515 This means they are set thereafter only if the connection becomes
3518 PH/06 Added the client_condition to authenticators so that some can be skipped
3519 by clients under certain conditions.
3521 PH/07 The error message for a badly-placed control=no_multiline_responses left
3522 "_responses" off the end of the name.
3524 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3526 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3527 (without spaces) instead of just copying the configuration text.
3529 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3531 PH/11 Added $max_received_linelength.
3533 PH/12 Added +ignore_defer and +include_defer to host lists.
3535 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3536 way in which PCRE > 7.0 is built.
3538 PH/14 Implemented queue_only_load_latch.
3540 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3541 MAIL command. The effect was to mangle the value on 64-bit systems.
3543 PH/16 Another patch from the Sieve maintainer.
3545 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3547 PH/18 If a system quota error occurred while trying to create the file for
3548 a maildir delivery, the message "Mailbox is full" was not appended to the
3549 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3550 only to a quota excession during the actual writing of the file.
3552 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3553 characters?) which causes problems in log lines. The DN values are now
3554 passed through string_printing() before being added to log lines.
3556 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3557 and InterBase are left for another time.)
3559 PH/21 Added message_body_newlines option.
3561 PH/22 Guard against possible overflow in moan_check_errorcopy().
3563 PH/23 POSIX allows open() to be a macro; guard against that.
3565 PH/24 If the recipient of an error message contained an @ in the local part
3566 (suitably quoted, of course), incorrect values were put in $domain and
3567 $local_part during the evaluation of errors_copy.
3573 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3574 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3577 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3578 issue a MAIL command.
3580 PH/02 In an ACL statement such as
3582 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3584 if a client was not listed at all, or was listed with a value other than
3585 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3586 the condition was not true (as it should be), so access was not denied.
3587 The bug was that the ! inversion was incorrectly passed on to the second
3588 item. This has been fixed.
3590 PH/03 Added additional dnslists conditions == and =& which are different from
3591 = and & when the dns lookup returns more than one IP address.
3593 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3594 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3596 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3597 FSYNC, which compiles an option called disable_fsync that allows for
3598 bypassing fsync(). The documentation is heavily laced with warnings.
3600 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3602 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3603 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3604 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3605 including adding "make clean"; (3) Added -fPIC when compiling the test
3606 dynamically loaded module, to get rid of a warning.
3608 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3609 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3610 = 0s. The bug is otherwise harmless.
3612 PH/07 There was a bug in the dovecot authenticator such that the value of
3613 $auth1 could be overwritten, and so not correctly preserved, after a
3614 successful authentication. This usually meant that the value preserved by
3615 the server_setid option was incorrect.
3617 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3619 PH/09 Installed PCRE release 7.0.
3621 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3622 run for batched SMTP input. It is now run at the start of every message
3623 in the batch. While fixing this I discovered that the process information
3624 (output by running exiwhat) was not always getting set for -bs and -bS
3625 input. This is fixed, and it now also says "batched" for BSMTP.
3627 PH/11 Added control=no_pipelining.
3629 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3630 patch, slightly modified), and move the expansion of helo_data till after
3631 the connection is made in the smtp transport (so it can use these
3634 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3636 PH/14 Added log_selector = +pid.
3638 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3640 PH/16 Add ${if forany and ${if forall.
3642 PH/17 Added dsn_from option to vary the From: line in DSNs.
3644 PH/18 Flush SMTP output before performing a callout, unless control =
3645 no_callout_flush is set.
3647 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3648 was true (the default) a successful delivery failed to delete the retry
3649 item, thus causing premature timeout of the address. The bug is now
3652 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3654 PH/21 Long custom messages for fakedefer and fakereject are now split up
3655 into multiline responses in the same way that messages for "deny" and
3656 other ACL rejections are.
3658 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3659 with slight modification.
3661 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3662 draft, changing the syntax and factoring some duplicate code".
3664 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3665 for deliveries of the second and subsequent messages over the same SMTP
3668 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3669 ${reduce, with only minor "tidies".
3671 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3673 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3674 expansion side effects.
3676 PH/27 When a message times out after an over-quota error from an Exim-imposed
3677 quota, the bounce message says "mailbox is full". This message was not
3678 being given when it was a system quota that was exceeded. It now should
3681 MH/03 Made $recipients available in local_scan(). local_scan() already has
3682 better access to the recipient list through recipients_list[], but
3683 $recipients can be useful in postmaster-provided expansion strings.
3685 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3686 in the case of a MAIL command with additional options following the
3687 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3688 were accidentally chopped off.
3690 PH/29 SMTP synchronization checks are implemented when a command is read -
3691 there is a check that no more input is waiting when there shouldn't be
3692 any. However, for some commands, a delay in an ACL can mean that it is
3693 some time before the response is written. In this time, more input might
3694 arrive, invalidly. So now there are extra checks after an ACL has run for
3695 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3696 pipelining has not been advertised.
3698 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3700 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3701 returned in the SMTP response when smtp_return_error_details was false.
3702 This has been fixed.
3704 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3705 instead of the C I/O that was originally supplied, because problems were
3706 reported on Solaris.
3708 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3709 Exim which did not show up earlier: it was assuming that a call to
3710 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3711 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3712 was a macro that became an assignment, so it seemed to work. This has
3713 changed to a proper function call with a void return, hence the compile
3714 error. Exim's code has been fixed.
3716 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3719 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3721 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3723 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3724 because it turns out to be too big or there is a timeout) there is no
3725 "Completed" line in the log. When some messages of this type were
3726 selected by exigrep, they were listed as "not completed". Others were
3727 picked up by some special patterns. I have improved the selection
3728 criteria to be more general.
3730 PH/38 The host_find_failed option in the manualroute router can now be set
3731 to "ignore", to completely ignore a host whose IP address cannot be
3732 found. If all hosts are ignored, the behaviour is controlled by the new
3733 host_all_ignored option.
3735 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3736 homing or because of multiple MX records with /mx) generated more than
3737 one IP address, and the following item turned out to be the local host,
3738 all the secondary addresses of the first item were incorrectly removed
3739 from the list, along with the local host and any following hosts (which
3740 is what is supposed to happen).
3742 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3743 whoever called Exim into the -H file. In the case of the daemon it was
3744 behaving confusingly. When first started, it used values for whoever
3745 started the daemon, but after a SIGHUP it used the Exim user (because it
3746 calls itself on a restart). I have changed the code so that it now always
3749 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3750 message are rejected with the same error (e.g. no authentication or bad
3751 sender address), and a DATA command is nevertheless sent (as can happen
3752 with PIPELINING or a stupid MUA), the error message that was given to the
3753 RCPT commands is included in the rejection of the DATA command. This is
3754 intended to be helpful for MUAs that show only the final error to their
3757 PH/42 Another patch from the Sieve maintainer.
3759 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3760 Eximstats - Fixed some broken HTML links and added missing column headers
3762 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3763 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3765 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3767 PH/43 Yet another patch from the Sieve maintainer.
3769 PH/44 I found a way to check for a TCP/IP connection going away before sending
3770 the response to the final '.' that terminates a message, but only in the
3771 case where the client has not sent further data following the '.'
3772 (unfortunately, this is allowed). However, in many cases there won't be
3773 any further data because there won't be any more messages to send. A call
3774 to select() can be used: if it shows that the input is "ready", there is
3775 either input waiting, or the socket has been closed. An attempt to read
3776 the next input character can distinguish the two cases. Previously, Exim
3777 would have sent an OK response which the client would never have see.
3778 This could lead to message repetition. This fix should cure that, at
3779 least in a lot of common cases.
3781 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3782 advertised in response to EHLO.
3788 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3789 fixed by 4.65/MH/01 (is this a record?) are fixed:
3791 (i) An empty string was always treated as zero by the numeric comparison
3792 operators. This behaviour has been restored.
3794 (ii) It is documented that the numeric comparison operators always treat
3795 their arguments as decimal numbers. This was broken in that numbers
3796 starting with 0 were being interpreted as octal.
3798 While fixing these problems I realized that there was another issue that
3799 hadn't been noticed. Values of message_size_limit (both the global option
3800 and the transport option) were treated as octal if they started with 0.
3801 The documentation was vague. These values are now always treated as
3802 decimal, and I will make that clear in the documentation.
3808 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3809 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3812 MH/01 Don't check that the operands of numeric comparison operators are
3813 integers when their expansion is in "skipping" mode (fixes bug
3814 introduced by 4.64-PH/07).
3816 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3817 child addresses, Exim now panics and dies. Previously, because the count
3818 is held in a short int, deliveries were likely to be lost. As such a
3819 large number of recipients for a single message is ridiculous
3820 (performance will be very, very poor), I have chosen to impose a limit
3821 rather than extend the field.
3827 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3828 leftover -K file (the existence of which was triggered by #402).
3829 While we were at it, introduced process PID as part of the -K
3830 filename. This should rule out race conditions when creating
3833 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3834 processing considerably. Previous code took too long for large mails,
3835 triggering a timeout which in turn triggers #401.
3837 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3838 in the DK code in transports.c. sendfile() is not really portable,
3839 hence the _LINUX specificness.
3841 TF/01 In the add_headers option to the mail command in an Exim filter,
3842 there was a bug that Exim would claim a syntax error in any
3843 header after the first one which had an odd number of characters
3846 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3847 callout verification, Exim cached a "reject" for the entire domain. This
3848 is correct for most verifications, but it is not correct for a recipient
3849 verification with use_sender or use_postmaster set, because in that case
3850 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3851 case of MAIL FROM:<> rejection from other early rejections (e.g.
3852 rejection of HELO). When verifying a recipient using a non-null MAIL
3853 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3854 Whatever the result of the callout, the value of the domain cache is
3855 left unchanged (for any other kind of callout, getting as far as trying
3856 RCPT means that the domain itself is ok).
3858 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3861 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3862 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3863 session, the daemon ignores ECONNECT errors and logs others; it now
3864 ignores EPIPE as well.
3866 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3867 (quoted-printable decoding).
3869 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3870 later the small subsequent patch to fix an introduced bug.
3872 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3874 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3876 PH/08 An error is now given if message_size_limit is specified negative.
3878 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3879 to be given (somewhat) arbitrary names.
3881 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3884 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3885 miscellaneous code fixes
3887 PH/10 Added the log_reject_target ACL modifier to specify where to log
3890 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3891 hostname. This is wrong, because it relates to the incoming message (and
3892 probably the interface on which it is arriving) and not to the outgoing
3893 callout (which could be using a different interface). This has been
3894 changed to use the value of the helo_data option from the smtp transport
3895 instead - this is what is used when a message is actually being sent. If
3896 there is no remote transport (possible with a router that sets up host
3897 addresses), $smtp_active_hostname is used.
3899 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3900 tweaks were necessary in order to get it to work (see also 21 below):
3901 (a) The code assumed that strncpy() returns a negative number on buffer
3902 overflow, which isn't the case. Replaced with Exim's string_format()
3904 (b) There were several signed/unsigned issues. I just did the minimum
3905 hacking in of casts. There is scope for a larger refactoring.
3906 (c) The code used strcasecmp() which is not a standard C function.
3907 Replaced with Exim's strcmpic() function.
3908 (d) The code set only $1; it now sets $auth1 as well.
3909 (e) A simple test gave the error "authentication client didn't specify
3910 service in request". It would seem that Dovecot has changed its
3911 interface. Fortunately there's a specification; I followed it and
3912 changed what the client sends and it appears to be working now.
3914 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3917 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3918 address A is aliased to B and C, where B exists and C does not. Without
3919 -v the output is "A verified" because verification stops after a
3920 successful redirection if more than one address is generated. However,
3921 with -v the child addresses are also verified. Exim was outputting "A
3922 failed to verify" and then showing the successful verification for C,
3923 with its parentage. It now outputs "B failed to verify", showing B's
3924 parentage before showing the successful verification of C.
3926 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3927 look up a TXT record in a specific list after matching in a combined
3930 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3931 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3932 they consult the DNS. I had assumed they would set it the way they
3933 wanted; and indeed my experiments on Linux seem to show that in some
3934 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3935 To be on the safe side, however, I have now made the interface to
3936 host_find_byname() similar to host_find_bydns(), with an argument
3937 containing the DNS resolver options. The host_find_byname() function now
3938 sets these options at its start, just as host_find_bydns() does. The smtp
3939 transport options dns_qualify_single and dns_search_parents are passed to
3940 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3941 of host_find_byname() use the default settings of RES_DEFNAMES
3942 (qualify_single) but not RES_DNSRCH (search_parents).
3944 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3945 spool_read_header() do less string testing, by means of a preliminary
3946 switch on the second character of optional "-foo" lines. (This is
3947 overdue, caused by the large number of possibilities that now exist.
3948 Originally there were few.) While I was there, I also converted the
3949 str(n)cmp tests so they don't re-test the leading "-" and the first
3950 character, in the hope this might squeeze out yet more improvement.
3952 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3953 flag allowing group syntax was set by the header_syntax check but not
3954 turned off, possible causing trouble later; (2) The flag was not being
3955 set at all for the header_verify test, causing "group"-style headers to
3956 be rejected. I have now set it in this case, and also caused header_
3957 verify to ignore an empty address taken from a group. While doing this, I
3958 came across some other cases where the code for allowing group syntax
3959 while scanning a header line wasn't quite right (mostly, not resetting
3960 the flag correctly in the right place). These bugs could have caused
3961 trouble for malformed header lines. I hope it is now all correct.
3963 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3964 with the "reply" argument non-NULL. The code, however (which originally
3965 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3966 but it didn't always do it. This confused somebody who was copying the
3967 code for some other use. I have removed all the tests.
3969 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3970 feature that was used to support insecure browsers during the U.S. crypto
3971 embargo. It requires special client support, and Exim is probably the
3972 only MTA that supported it -- and would never use it because real RSA is
3973 always available. This code has been removed, because it had the bad
3974 effect of slowing Exim down by computing (never used) parameters for the
3975 RSA_EXPORT functionality.
3977 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3978 authenticator to fail if there's a tab character in the incoming data
3979 (there should never be unless someone is messing about, as it's supposed
3980 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3981 if the connection is using TLS or if the remote IP is the same as the
3982 local IP, and the "valid-client-cert option" if a client certificate has
3985 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3986 authenticators. This can be used for authorization after authentication
3987 succeeds. (In the case of plaintext, it servers for both authentication
3990 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3991 if any retry times were supplied.
3993 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3994 connection, where there is no client IP address to check. In this
3995 situation, the verify now always succeeds.
3997 PH/25 Applied John Jetmore's -Mset patch.
3999 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4001 PH/27 In a string expansion for a processed (not raw) header when multiple
4002 headers of the same name were present, leading whitespace was being
4003 removed from all of them, but trailing whitespace was being removed only
4004 from the last one. Now trailing whitespace is removed from each header
4005 before concatenation. Completely empty headers in a concatenation (as
4006 before) are ignored.
4008 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4009 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4011 PH/29 [Removed. This was a change that I later backed out, and forgot to
4012 correct the ChangeLog entry (that I had efficiently created) before
4013 committing the later change.]
4015 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4016 address errors (4xx response to RCPT) over the same connection as other
4017 messages routed to the same hosts. Such deliveries are always "forced",
4018 so retry times are not inspected. This resulted in far too many retries
4019 for the affected addresses. The effect occurred only when there were more
4020 hosts than the hosts_max_try setting in the smtp transport when it had
4021 the 4xx errors. Those hosts that it had tried were not added to the list
4022 of hosts for which the message was waiting, so if all were tried, there
4023 was no problem. Two fixes have been applied:
4025 (i) If there are any address or message errors in an SMTP delivery, none
4026 of the hosts (tried or untried) are now added to the list of hosts
4027 for which the message is waiting, so the message should not be a
4028 candidate for sending over the same connection that was used for a
4029 successful delivery of some other message. This seems entirely
4030 reasonable: after all the message is NOT "waiting for some host".
4031 This is so "obvious" that I'm not sure why it wasn't done
4032 previously. Hope I haven't missed anything, but it can't do any
4033 harm, as the worst effect is to miss an optimization.
4035 (ii) If, despite (i), such a delivery is accidentally attempted, the
4036 routing retry time is respected, so at least it doesn't keep
4037 hammering the server.
4039 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4040 in ${readsocket because some servers need this prod.
4042 PH/32 Added some extra debug output when updating a wait-xxx database.
4044 PH/33 The hint "could be header name not terminated by colon", which has been
4045 given for certain expansion errors for a long time, was not being given
4046 for the ${if def:h_colon_omitted{... case.
4048 PH/34 The spec says: "With one important exception, whenever a domain list is
4049 being scanned, $domain contains the subject domain." There was at least
4050 one case where this was not true.
4052 PH/35 The error "getsockname() failed: connection reset by peer" was being
4053 written to the panic log as well as the main log, but it isn't really
4054 panic-worthy as it just means the connection died rather early on. I have
4055 removed the panic log writing for the ECONNRESET error when getsockname()
4058 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4059 runs only) independently of the message's sender address. This meant
4060 that, if the 4xx error was in fact related to the sender, a different
4061 message to the same recipient with a different sender could confuse
4062 things. In particular, this can happen when sending to a greylisting
4063 server, but other circumstances could also provoke similar problems.
4064 I have changed the default so that the retry time for these errors is now
4065 based a combination of the sender and recipient addresses. This change
4066 can be overridden by setting address_retry_include_sender=false in the
4069 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4070 remote server are returned as part of bounce messages. This was not
4071 happening for LMTP over a pipe (the lmtp transport), but now it is the
4072 same for both kinds of LMTP.
4074 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4075 in header lines that were in fact CNAMEs. This is no longer the case.
4077 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4078 and queue runs started by the daemon processed all messages. This has
4079 been fixed so that -R and -S can now usefully be given with -q<time>.
4081 PH/40 Import PCRE release 6.7 (fixes some bugs).
4083 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4085 PH/42 Give an error if -q is specified more than once.
4087 PH/43 Renamed the variables $interface_address and $interface_port as
4088 $received_ip_address and $received_port, to make it clear that these
4089 values apply to message reception, and not to the outgoing interface when
4090 a message is delivered. (The old names remain recognized, of course.)
4092 PH/44 There was no timeout on the connect() call when using a Unix domain
4093 socket in the ${readsocket expansion. There now is.
4095 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4096 be meaningful with "accept".
4098 SC/01 Eximstats V1.43
4099 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4101 SC/02 Eximstats V1.44
4102 Use a glob alias rather than an array ref in the generated
4103 parser. This improves both readability and performance.
4105 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4106 Collect SpamAssassin and rejection statistics.
4107 Don't display local sender or destination tables unless
4108 there is data to show.
4109 Added average volumes into the top table text output.
4111 SC/04 Eximstats V1.46
4112 Collect data on the number of addresses (recipients)
4113 as well as the number of messages.
4115 SC/05 Eximstats V1.47
4116 Added 'Message too big' to the list of mail rejection
4117 reasons (thanks to Marco Gaiarin).
4119 SC/06 Eximstats V1.48
4120 Mainlog lines which have GMT offsets and are too short to
4121 have a flag are now skipped.
4123 SC/07 Eximstats V1.49 (Alain Williams)
4124 Added the -emptyok flag.
4126 SC/08 Eximstats V1.50
4127 Fixes for obtaining the IP address from reject messages.
4129 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4130 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4131 whitespace changes from 4.64-PH/27
4133 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4136 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4137 are found, allow negative numbers in numeric criteria)
4139 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4141 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4142 to match changes made in 4.64-PH/43
4144 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4146 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4147 "domain" is actually an IP address, and force a failure. This locks out
4148 those revolvers/nameservers that support "A-for-A" lookups, in
4149 contravention of the specifications.
4151 PH/48 When a host name was looked up from an IP address, and the subsequent
4152 forward lookup of the name timed out, the host name was left in
4153 $sender_host_name, contrary to the specification.
4155 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4156 restricted to single-key lookups, Exim was not diagnosing an error if
4157 * or *@ was used with a query-style lookup.
4159 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4161 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4162 long ago, but noone interested enough thought of it. Let's just say that
4163 the "1.1" means that there are some new functions that weren't there at
4164 some point in the past.
4166 PH/51 Error processing for expansion failure of helo_data from an smtp
4167 transport during callout processing was broken.
4169 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4170 tested/used via the -bh/-bhc/-bs options.
4172 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4173 bug, fixed in subsequent PCRE releases).
4175 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4176 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4178 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4184 SC/01 Use a glob alias rather than an array ref in eximstats generated
4185 parser. This improves both readability and performance.
4187 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4188 Don't display local sender or destination tables in eximstats unless
4189 there is data to show.
4190 Added average volumes into the eximstats top table text output.
4192 SC/03 Collect data on the number of addresses (recipients) as well
4193 as the number of messages in eximstats.
4195 TF/01 Correct an error in the documentation for the redirect router. Exim
4196 does (usually) call initgroups() when daemonizing.
4198 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4199 with consistent privilege compared to when running as a daemon.
4201 TF/03 Note in the spec that $authenticated_id is not set for local
4202 submissions from trusted users.
4204 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4205 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4207 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4208 by adding some example configuration directives to the default
4209 configuration file. A little bit of work is required to uncomment the
4210 directives and define how usernames and passwords are checked, but
4211 there is now a framework to start from.
4213 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4214 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4215 without this. I don't know how relevant this is to other LDAP libraries.
4217 PH/02 Add the verb name to the "unknown ACL verb" error.
4219 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4221 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4223 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4224 directory not expanded when it should be if an expanded home directory
4225 was set for the address (which is overridden by the transport).
4227 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4230 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4231 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4232 because it is too late at that time, and has no effect.
4234 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4235 security issue with \' (bugzilla #107). I could not use the
4236 PQescapeStringConn() function, because it needs a PGconn value as one of
4239 PH/08 When testing addresses using -bt, indicate those final addresses that
4240 are duplicates that would not cause an additional delivery. At least one
4241 person was confused, thinking that -bt output corresponded to deliveries.
4242 (Suppressing duplicates isn't a good idea as you lose the information
4243 about possibly different redirections that led to the duplicates.)
4245 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4246 systems where poll() doesn't work, in particular OS X.
4248 PH/10 Added more information to debugging output for retry time not reached.
4250 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4251 operations in malware.c.
4253 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4256 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4257 syslog_duplication set false, log lines that would normally be written
4258 both the the main log and to the reject log were not written to syslog at
4261 PH/14 In the default configuration, change the use of "message" in ACL warn
4262 statements to "add_header".
4264 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4265 not followed by a command (e.g. "seen endif").
4267 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4268 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4271 PH/17 Added extra conditions to the default value of delay_warning_condition
4275 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4276 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4277 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4280 The Auto-Submitted: and various List- headers are standardised, whereas I
4281 don't think Precedence: ever was.
4283 PH/18 Refactored debugging code in route_finduser() to show more information,
4284 in particular, the error code if getpwnam() issues one.
4286 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4287 This is apparently needed in addition to the PH/07 change above to avoid
4288 any possible encoding problems.
4290 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4291 but not after initializing Perl.
4293 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4294 output them only if debugging. By default they are written stderr,
4295 apparently, which is not desirable.
4297 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4300 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4303 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4305 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4306 authenticated or an ident call has been made. Suppress the default
4307 values for $authenticated_id and $authenticated_sender (but permit -oMai
4308 and -oMas) when testing with -bh.
4310 PH/24 Re-jigged the order of the tests in the default configuration so that the
4311 tests for valid domains and recipients precede the DNS black list and CSA
4312 tests, on the grounds that those ones are more expensive.
4314 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4315 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4316 This bug exists in every version of Exim that I still have, right back to
4319 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4320 However, an attempt to turn on case-sensitivity in a regex key by
4321 including (?-i) didn't work because the subject string was already
4322 lowercased, and the effects were non-intuitive. It turns out that a
4323 one-line patch can be used to allow (?-i) to work as expected.
4329 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4330 other effects) broke the use of negated acl sub-conditions.
4332 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4335 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4336 "Deny" causes Exim to reject the incoming connection with a 554 error.
4337 Unfortunately, if there is a major crisis, such as a disk failure,
4338 tcp-wrappers gives "deny", whereas what one would like would be some
4339 kind of temporary error. A kludge has been added to help with this.
4340 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4341 554 error is used if errno is still zero or contains ENOENT (which occurs
4342 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4345 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4347 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4348 errors. Otherwise a message that provokes a temporary error (when other
4349 messages do not) can cause a whole host to time out.
4351 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4352 addresses were routed directly to files or pipes from a redirect router.
4353 File deliveries just didn't batch; pipe deliveries might have suffered
4356 PH/06 A failure to get a lock for a hints database would erroneously always say
4357 "Failed to get write lock", even when it was really a read lock.
4359 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4360 of 0600. This has been changed to use the value of the lockfile_mode
4361 option (which defaults to 0600).
4363 PH/08 Applied small patch from the Sieve maintainer.
4365 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4366 folder from quota calculations, a direct delivery into this folder messed
4367 up the contents of the maildirsize file. This was because the regex was
4368 used only to exclude .Trash (or whatever) when the size of the mailbox
4369 was calculated. There was no check that a delivery was happening into an
4370 excluded directory. This bug has been fixed by ignoring all quota
4371 processing for deliveries into excluded directories.
4373 PH/10 Added the maildirfolder_create_regex option to appendfile.
4379 PH/01 The code for finding all the local interface addresses on a FreeBSD
4380 system running IPv6 was broken. This may well have applied to all BSD
4381 systems, as well as to others that have similar system calls. The broken
4382 code found IPv4 interfaces correctly, but gave incorrect values for the
4383 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4384 that it would not match correctly against @[] and not recognize the IPv6
4387 PH/02 The ipliteral router was not recognizing addresses of the form user@
4388 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4390 PH/03 Added disable_ipv6.
4392 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4393 decoded form, because it is most often used to construct To: headers
4394 lines in autoreplies, and the decoded form may well be syntactically
4395 invalid. However, $reply_address has leading white space removed, and all
4396 newlines turned into spaces so that the autoreply transport does not
4399 PH/05 If group was specified without a user on a router, and no group or user
4400 was specified on a transport, the group from the router was ignored.
4402 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4403 for visible compile-time settings that can be used to change these
4404 numbers, for those that want even more. Backwards compatibility with old
4405 spool files has been maintained. However, going back to a previous Exim
4406 release will lost any variables that are in spool files.
4408 PH/07 Two small changes when running in the test harness: increase delay when
4409 passing a TCP/IP connection to a new process, in case the original
4410 process has to generate a bounce, and remove special handling of
4411 127.0.0.2 (sic), which is no longer necessary.
4413 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4414 be the same on different OS.
4416 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4419 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4420 whether --show-vars was specified or not
4422 JJ/02 exipick: Added support for new ACL variable spool format introduced
4425 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4426 syntactically invalid From: or Reply-to: line, and a filter used this to
4427 generate an autoreply, and therefore failed to obtain an address for the
4428 autoreply, Exim could try to deliver to a non-existent relative file
4429 name, causing unrelated and misleading errors. What now happens is that
4430 it logs this as a hard delivery error, but does not attempt to create a
4433 PH/11 The exinext utility has a -C option for testing purposes, but although
4434 the given file was scanned by exinext itself; it wasn't being passed on
4435 when Exim was called.
4437 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4438 an end-of-file indication when reading a command response.
4440 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4441 compiled. In many other places in Exim, IPv6 addresses are always
4442 recognized, so I have changed this. It also means that IPv4 domain
4443 literals of the form [IPV4:n.n.n.n] are now always recognized.
4445 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4446 used if the router is not running as root, for example, when verifying at
4447 ACL time, or when using -bh. The debugging output from this situation was
4448 non-existent - all you got was a failure to exec. I have made two
4451 (a) Failures to set uid/gid, the current directory, or a process leader
4452 in a subprocess such as that created by queryprogram now generate
4453 suitable debugging output when -d is set.
4455 (b) The queryprogram router detects when it is not running as root,
4456 outputs suitable debugging information if -d is set, and then runs
4457 the subprocess without attempting to change uid/gid.
4459 PH/15 Minor change to Makefile for building test_host (undocumented testing
4462 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4463 additional section of a DNS packet that returns MX or SRV records.
4464 Instead, it always explicitly searches for A/AAAA records. This avoids
4465 major problems that occur when a DNS server includes only records of one
4466 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4467 fixed another bug: if SRV records were looked up and the corresponding
4468 address records were *not* found in the additional section, the port
4469 values from the SRV records were lost.
4471 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4472 using the correct key (the original address) when searching the retry
4473 rules in order to find which one to use for generating the retry hint.
4475 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4476 adding the default one. Similarly, if it contains a Reply-To: header, the
4477 errors_reply_to option, if set, is not used.
4479 PH/19 When calculating a retry time, Exim used to measure the "time since
4480 failure" by looking at the "first failed" field in the retry record. Now
4481 it does not use this if it is later than than the arrival time of the
4482 message. Instead it uses the arrival time. This makes for better
4483 behaviour in cases where some deliveries succeed, thus re-setting the
4484 "first failed" field. An example is a quota failure for a huge message
4485 when small messages continue to be delivered. Without this change, the
4486 "time since failure" will always be short, possible causing more frequent
4487 delivery attempts for the huge message than are intended.
4488 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4490 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4491 $1, $2, $3) because the numerical variables can be reset during some
4492 expansion items (e.g. "match"), thereby losing the authentication data.
4494 PH/21 Make -bV show the size of off_t variables so that the test suite can
4495 decide whether to run tests for quotas > 2G.
4497 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4498 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4499 filecount value is greater than 2G or if a quota value is greater than 2G
4500 on a system where the size of off_t is not greater than 4, a panic error
4503 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4504 never match. The debug and -bh output now contains an explicit error
4505 message indicating a malformed IPv4 address or mask.
4507 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4508 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4509 PH/23 above applies.
4511 PH/25 Do not write to syslog when running in the test harness. The only
4512 occasion when this arises is a failure to open the main or panic logs
4513 (for which there is an explicit test).
4515 PH/26 Added the /no_tell option to "control=freeze".
4517 PH/27 If a host name lookup failed very early in a connection, for example, if
4518 the IP address matched host_lookup and the reverse lookup yielded a name
4519 that did not have a forward lookup, an error message of the form "no IP
4520 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4521 could be logged. Now it outputs the IP address instead of "NULL".
4523 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4524 allows the sender and the authenticated sender to be set when
4525 submitting a message from within Exim. Since child_open_exim() is
4526 documented for local_scan(), the new function should be too.
4528 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4529 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4530 results in an empty string is now treated as unset.
4532 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4534 PH/31 Added sender_verify_fail logging option.
4536 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4537 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4538 tidied the source and removed it altogether.
4540 PH/33 When a queue run was abandoned because the load average was too high, a
4541 log line was always written; now it is written only if the queue_run log
4542 selector is set. In addition, the log line for abandonment now contains
4543 information about the queue run such as the pid. This is always present
4544 in "start" and "stop" lines but was omitted from the "abandon" line.
4546 PH/34 Omit spaces between a header name and the colon in the error message that
4547 is given when verify = headers_syntax fails (if there are lots of them,
4548 the message gets confusing).
4550 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4551 names, as there are now some PTR records that contain slashes. This check
4552 is only to protect against broken name servers that fall over on strange
4553 characters, so the fact that it applies to all lookups doesn't matter.
4555 PH/36 Now that the new test suite is complete, we can remove some of the
4556 special code in Exim that was needed for the old test suite. For example,
4557 sorting DNS records because real resolvers return them in an arbitrary
4558 order. The new test suite's fake resolver always returns records in the
4561 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4562 bounces) except when queue_only is set, to avoid logging races between
4563 the different processes.
4565 PH/38 Panic-die if .include specifies a non-absolute path.
4567 PH/39 A tweak to the "H" retry rule from its user.
4569 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4570 a label. They prevented compilation on older perls.
4572 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4573 a warning to be raised on newish perls.
4575 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4576 on queue. Changes to match documented behaviour of showing count of
4577 messages matching specified criteria.
4579 PH/40 Changed the default ident timeout from 30s to 5s.
4581 PH/41 Added support for the use of login_cap features, on those BSD systems
4582 that have them, for controlling the resources used by pipe deliveries.
4584 PH/42 The content-scanning code uses fopen() to create files in which to put
4585 message data. Previously it was not paying any attention to the mode of
4586 the files. Exim runs with umask(0) because the rest of the code creates
4587 files with open(), and sets the required mode explicitly. Thus, these
4588 files were ending up world-writeable. This was not a big issue, because,
4589 being within the spool directory, they were not world-accessible. I have
4590 created a function called modefopen, which takes an additional mode
4591 argument. It sets umask(777), creates the file, chmods it to the required
4592 mode, then resets the umask. All the relevant calls to fopen() in the
4593 content scanning code have been changed to use this function.
4595 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4596 to 24 hours. This avoids potential overflow problems when processing G
4597 and H retry rules. I suspect nobody ever tinkers with this value.
4599 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4601 PH/45 When the plaintext authenticator is running as a client, the server's
4602 challenges are checked to ensure they are valid base64 strings. By
4603 default, the authentication attempt is cancelled if an invalid string is
4604 received. Setting client_ignore_invalid_base64 true ignores these errors.
4605 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4606 they are received. Thus, the responses can be made to depend on the
4607 challenges. If an invalid string is ignored, an empty string is placed in
4610 PH/46 Messages that are created by the autoreply transport now contains a
4611 References: header, in accordance with RFCs 2822 and 3834.
4613 PH/47 Added authenticated_sender_force to the smtp transport.
4615 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4617 PH/49 Installed latest patch from the Sieve maintainer.
4619 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4620 was also set, the appendfile transport was unnecessarily scanning a
4621 directory of message files (e.g. for maildir delivery) to find the count
4622 of files (along with the size), even though it did not need this
4623 information. It now does the scan only if it needs to find either the
4624 size of the count of files.
4626 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4628 PH/52 Two bugs concerned with error handling when the smtp transport is
4631 (i) Exim was not creating retry information for temporary errors given
4632 for individual recipients after the DATA command when the smtp transport
4633 was used in LMTP mode. This meant that they could be retried too
4634 frequently, and not timed out correctly.
4636 (ii) Exim was setting the flag that allows error details to be returned
4637 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4638 recipients that were returned after the DATA command.
4640 PH/53 This is related to PH/52, but is more general: for any failing address,
4641 when detailed error information was permitted to be returned to the
4642 sender, but the error was temporary, then after the final timeout, only
4643 "retry timeout exceeded" was returned. Now it returns the full error as
4644 well as "retry timeout exceeded".
4646 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4647 do this, and (what is worse) MTAs that accept it.
4649 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4650 will now be deprecated.
4652 PH/56 New os.c-cygwin from the Cygwin maintainer.
4654 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4655 formats (previously only available in exim formats via -bpr, -bpru,
4656 and -bpra. Now also available in native and exiqgrep formats)
4658 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4659 with very large, slow to parse queues
4661 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4663 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4665 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4666 responses to authentication challenges, though it was showing the
4667 challenges; (ii) I've removed the CR characters from the debug output for
4670 PH/58 Allow for the insertion of a newline as well as a space when a string
4671 is turned into more than one encoded-word during RFC 2047 encoding. The
4672 Sieve code now uses this.
4674 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4675 data_4xx, lost_connection, tls_required.
4677 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4678 message was being sent as an SMTP response.
4680 PH/61 Add -l and -k options to exicyclog.
4682 PH/62 When verifying, if an address was redirected to one new address, so that
4683 verification continued, and the new address failed or deferred after
4684 having set something in $address_data, the value of $address_data was not
4685 passed back to the ACL. This was different to the case when no
4686 redirection occurred. The value is now passed back in both cases.
4688 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4689 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4690 use login_cap.h, so on its own it isn't the distinguishing feature. The
4691 new name refers directly to the setclassresources() function.
4693 PH/65 Added configuration files for NetBSD3.
4695 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4697 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4698 is preferred over IPv4.
4700 PH/68 The bounce_return_message and bounce_return_body options were not being
4701 honoured for bounces generated during the reception of non-SMTP messages.
4702 In particular, this applied to messages rejected by the ACL. This bug has
4703 been fixed. However, if bounce_return_message is true and bounce_return_
4704 body is false, the headers that are returned for a non-SMTP message
4705 include only those that have been read before the error was detected.
4706 (In the case of an ACL rejection, they have all been read.)
4708 PH/69 The HTML version of the specification is now built in a directory called
4709 spec_html instead of spec.html, because the latter looks like a path with
4710 a MIME-type, and this confuses some software.
4712 PH/70 Catch two compiler warnings in sieve.c.
4714 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4715 function verify_get_ident() calls ip_connect() to connect a socket, but
4716 if the "connect()" function timed out, ip_connect() used to close the
4717 socket. However, verify_get_ident() also closes the socket later, and in
4718 between Exim writes to the log, which may get opened at this point. When
4719 the socket was closed in ip_connect(), the log could get the same file
4720 descriptor number as the socket. This naturally causes chaos. The fix is
4721 not to close the socket in ip_connect(); the socket should be closed by
4722 the function that creates it. There was only one place in the code where
4723 this was missing, in the iplookup router, which I don't think anybody now
4724 uses, but I've fixed it anyway.
4726 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4727 well as to direct DNS lookups. Otherwise the handling of names in host
4728 lists is inconsistent and therefore confusing.
4734 PH/01 Two changes to the default runtime configuration:
4736 (1) Move the checks for relay_from_hosts and authenticated clients from
4737 after to before the (commented out) DNS black list checks.
4739 (2) Add control=submission to the relay_from_hosts and authenticated
4740 clients checks, on the grounds that messages accepted by these
4741 statements are most likely to be submissions.
4743 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4745 (1) Generate an error if the third argument for the ${prvs expansion is
4748 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4751 (3) Reset the variables that are obtained from the first argument of
4752 ${prvscheck and used in the second argument before leaving the code,
4753 because their memory is reclaimed, so using them afterwards may do
4756 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4757 one (it's much easier than Tom thought :-).
4759 (5) Because of (4), we can now allow for the use of $prvscheck_result
4760 inside the third argument.
4762 PH/03 For some reason, the default setting of PATH when running a command from
4763 a pipe transport was just "/usr/bin". I have changed it to
4766 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4767 anything to be listed in the output from -bV.
4769 PH/05 When a filter generated an autoreply, the entire To: header line was
4770 quoted in the delivery log line, like this:
4772 => >A.N.Other <ano@some.domain> <original@ddress> ...
4774 This has been changed so that it extracts the operative address. There
4775 may be more than one such address. If so, they are comma-separated, like
4778 => >ano@some.domain,ona@other.domain <original@ddress> ...
4780 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4781 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4782 not being looked up in the rDNS to get a host name, Exim was showing the
4783 IP address twice in Received: lines, even though the IP addresses were
4784 identical. For example:
4786 Received: from [1.2.3.4] (helo=[1.2.3.4])
4788 However, if the real host name was known, it was omitting the HELO data
4789 if it matched the actual IP address. This has been tidied up so that it
4790 doesn't show the same IP address twice.
4792 PH/07 When both +timestamp and +memory debugging was on, the value given by
4793 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4794 called by the debug printing, thereby overwriting the timestamp buffer.
4795 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4797 PH/08 When the original message was included in an autoreply transport, it
4798 always said "this is a copy of the message, including all the headers",
4799 even if body_only or headers_only was set. It now gives an appropriate
4802 PH/09 Applied a patch from the Sieve maintainer which:
4804 o fixes some comments
4805 o adds the (disabled) notify extension core
4806 o adds some debug output for the result of if/elsif tests
4807 o points to the current vacation draft in the documentation
4808 and documents the missing references header update
4812 o fixes a bug in processing the envelope test (when testing
4813 multiple envelope elements, the last element determined the
4816 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4817 Electronic Mail") by including:
4819 Auto-submitted: auto-generated
4821 in the messages that it generates (bounce messages and others, such as
4822 warnings). In the case of bounce messages for non-SMTP messages, there was
4823 also a typo: it was using "Auto_submitted" (underscore instead of
4824 hyphen). Since every message generated by Exim is necessarily in response
4825 to another message, thes have all been changed to:
4827 Auto-Submitted: auto-replied
4829 in accordance with these statements in the RFC:
4831 The auto-replied keyword:
4833 - SHOULD be used on messages sent in direct response to another
4834 message by an automatic process,
4836 - MUST NOT be used on manually-generated messages,
4838 - MAY be used on Delivery Status Notifications (DSNs) and Message
4839 Disposition Notifications (MDNs),
4841 - MUST NOT be used on messages generated by automatic or periodic
4842 processes, except for messages which are automatic responses to
4845 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4846 to the default Received: header definition.
4848 PH/12 Added log selector acl_warn_skipped (default on).
4850 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4851 variables because (a) they are in the wrong storage pool and (b) even if
4852 they were copied, it wouldn't work properly because of the caching.
4854 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4855 checking when decoding. Apparently there are clients that generate
4856 overlong encoded strings. Why am I not surprised?
4858 PH/15 If the first argument of "${if match_address" was not empty, but did not
4859 contain an "@" character, Exim crashed. Now it writes a panic log message
4860 and treats the condition as false.
4862 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4864 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4865 extension "envelope-auth". The code is finished and in agreement with
4866 other implementations, but there is no documentation so far and in fact,
4867 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4868 not changing the active code.
4870 Print executed "if" and "elsif" statements when debugging is used. This
4871 helps a great deal to understand what a filter does.
4873 Document more things not specified clearly in RFC3028. I had all this
4874 sorted out, when out of a sudden new issues came to my mind. Oops."
4876 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4879 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4880 canonical form (as documented). However, after a host name lookup from
4881 the IP address, check_host() was doing a simple string comparison with
4882 addresses acquired from the DNS when checking that the found name did
4883 have the original IP as one of its addresses. Since any found IPv6
4884 addresses are likely to be in abbreviated form, the comparison could
4885 fail. Luckily, there already exists a function for doing the comparison
4886 by converting both addresses to binary, so now that is used instead of
4887 the text comparison.
4889 PH/20 There was another similar case to PH/19, when a complete host name was
4890 given in a host list; looking up its IP address could give an abbreviated
4891 form, whereas the current host's name might or might not be abbreviated.
4892 The same fix has been applied.
4898 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4899 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4902 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4903 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4905 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4907 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4908 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4909 Also, for "Auto-Submitted", treat anything other than "no" as
4910 non-personal, in accordance with RFC 3834. (Previously it treated
4911 anything starting "auto-" as non-personal.)
4913 TF/01 The control=submission/name=... option had a problem with syntax
4914 errors if the name included a slash character. The /name= option
4915 now slurps the rest of the string, so it can include any characters
4916 but it must come last in the list of options (after /sender_retain
4919 PH/05 Some modifications to the interface to the fake nameserver for the new
4927 TK/01 Added the "success_on_redirect" address verification option. See
4928 NewStuff for rationale and an example.
4930 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4932 PH/02 Patch to exigrep to allow it to work on syslog lines.
4934 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4935 fread() to skip over the body file's header line, because in Cygwin the
4936 header line is locked and is inaccessible.
4938 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4939 co-exist for some time) to make it clear that it is the Exim ID that is
4940 referenced, not the Message-ID: header line.
4942 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4943 string_format() function, because snprintf() does not exist on all
4946 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4947 ${stat: expansion item.
4949 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4950 protocol synchronization error", to keep the pedants happy.
4952 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4953 well as for IRIX systems, when gcc is being used. See the host.c source
4956 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4958 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4961 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4962 transport and to the smtp transport in LMTP mode.
4964 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4966 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4967 run a filter in a subprocess. This could lead to confusion in subsequent
4968 lookups in the parent process. There should also be a search_tidyup() at
4969 the end of the subprocess.
4971 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4972 only if the host matched helo_try_verify_hosts, which caused the
4973 verification to occur when the EHLO/HELO command was issued. The ACL just
4974 tested the remembered result. Now, if a previous verification attempt has
4975 not happened, "verify = helo" does it there and then.
4977 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4979 TK/03 Fix log output including CR from clamd.
4981 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4982 exist provoked a memory error which could cause a segfault.
4984 PH/15 Installed PCRE 6.2
4986 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4988 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4989 of the problem. Specifically, suggested +O2 rather than +O1 for the
4992 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4994 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4995 record, the port was not correctly logged when the outgoing_port log
4996 selector was set (it logged the transort's default port).
4998 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4999 fallback_hosts, and "hosts" in the smtp transport.
5001 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5002 host errors such as "Connection refused".
5004 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5005 authentication with radiusclient 0.4.9:
5007 - Error returned from rc_read_config was caught wrongly
5008 - Username/password not passed on to radius server due to wrong length.
5010 The presumption is that some radiusclient API changes for 4.51/PH/17
5011 were not taken care of correctly. The code is still untested by me (my
5012 Linux distribution still has 0.3.2 of radiusclient), but it was
5013 contributed by a Radius user.
5015 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5016 expanding the "port" option of the smtp transport.
5018 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5019 while reading a MIME header. Thanks to Tom Hughes for a patch.
5021 PH/24 Include config.h inside local_scan.h so that configuration settings are
5024 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5025 that in an ACL for RCPT (for example), you can examine exactly what was
5028 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5029 commands, but it was not correctly comparing the address with the actual
5030 client host address. Thus, it would show the EHLO address in Received:
5031 header lines when this was not necessary.
5033 PH/27 Added the % operator to ${eval:}.
5035 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5036 it should be ignoring failures (because with -C, for example, it has lost
5037 privilege). It wasn't ignoring creation failures other than "already
5040 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5043 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5044 because an input line was too long, either on its own, or by virtue of
5045 too many continuations, the temporary file was not being removed, and the
5046 return code was incorrect.
5048 PH/31 Missing "BOOL" in function definition in filtertest.c.
5050 PH/32 Applied Sieve patches from the maintainer.
5052 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5054 PH/33 Added "verify = not_blind".
5056 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5057 Local/Makefile (with some defaults set). These are used in built scripts
5058 such as exicyclog, but they have never been used in the exim_install
5059 script (though there are many overriding facilities there). I have
5060 arranged that the exim_install script now takes note of these two
5063 PH/35 Installed configuration files for Dragonfly.
5065 PH/36 When a locally submitted message by a trusted user did not contain a
5066 From: header, and the sender address was obtained from -f or from an SMTP
5067 MAIL command, and the trusted user did not use -F to supply a sender
5068 name, $originator_name was incorrectly used when constructing a From:
5069 header. Furthermore, $originator_name was used for submission mode
5070 messages from external hosts without From: headers in a similar way,
5071 which is clearly wrong.
5073 PH/37 Added control=suppress_local_fixups.
5075 PH/38 When log_selector = +received_sender was set, and the addition of the
5076 sender made the log line's construction buffer exactly full, or one byte
5077 less than full, an overflow happened when the terminating "\n" was
5080 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5081 when the result of a list match is failure because a DNS lookup failed.
5083 PH/40 RM_COMMAND is now used in the building process.
5085 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5086 the "build-* directories that it finds.
5088 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5089 address in a domain literal was a prefix of an interface address.
5091 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5092 when verifying a sender address, unless rewrite_headers is false.
5094 PH/44 Wrote a long comment about why errors_to addresses are verified as
5095 recipients, not senders.
5097 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5098 the ratelimit ACL was added.
5100 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5102 PH/46 Added extra information about PostgreSQL errors to the error string.
5104 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5105 avoiding the need to install special zones in a real server. This is
5106 backwards compatible; if it can't find the fake resolver, it drops back.
5107 Thus, both old and new test suites can be run.
5109 TF/02 Added util/ratelimit.pl
5111 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5112 clock is set back in time.
5114 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5115 Candler <B.Candler@pobox.com>.
5117 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5118 OK for addresses that are the result of redirections.
5120 PH/48 A number of further additions for the benefit of the new test suite,
5121 including a fake gethostbyname() that interfaces to the fake DNS resolver
5124 TF/06 The fix for widen_domains has also been applied to qualify_single and
5125 search_parents which are the other dnslookup options that can cause
5128 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5131 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5133 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5134 service" if it encountered an MX record with an empty target hostname.
5135 The message is now "an MX or SRV record indicated no SMTP service".
5137 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5138 if the DNS of the sending site is misconfigured. This is quite a
5139 common situation. This change restores the behaviour of treating a
5140 helo verification defer as a failure.
5142 PH/51 If self=fail was set on a router, the bounce message did not include the
5143 actual error message.
5149 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5151 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5152 command itself ended in error, the underlying message about the transport
5153 filter timeout was being overwritten with the pipe command error. Now the
5154 underlying error message should be appended to the second error message.
5156 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5158 PH/02 Reduce the amount of output that "make" produces by default. Full output
5159 can still be requested.
5161 PH/03 The warning log line about a condition test deferring for a "warn" verb
5162 was being output only once per connection, rather than after each
5163 occurrence (because it was using the same function as for successful
5164 "warn" verbs). This seems wrong, so I have changed it.
5166 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5167 it should not have, which might have caused a crash in the right
5168 circumstances, but probably never did.
5170 PH/04 Installed a modified version of Tony Finch's patch to make submission
5171 mode fix the return path as well as the Sender: header line, and to
5172 add a /name= option so that you can make the user's friendly name appear
5175 TF/03 Added the control = fakedefer ACL modifier.
5177 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5178 Mark Lowes for thorough testing.
5180 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5182 TK/03 Merged latest SRS patch from Miles Wilton.
5184 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5185 with the definition in sysexits.h (which is #included earlier).
5186 Fortunately, Exim does not actually use EX_OK. The code used to try to
5187 preserve the sysexits.h value, by assuming that macro definitions were
5188 scanned for macro replacements. I have been disabused of this notion,
5189 so now the code just undefines EX_OK before #including unistd.h.
5191 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5192 in the smtp transport. When a block could not be written in a single
5193 write() function, the timeout was being re-applied to each part-write.
5194 This seems wrong - if the receiver was accepting one byte at a time it
5195 would take for ever. The timeout is now adjusted when this happens. It
5196 doesn't have to be particularly precise.
5198 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5199 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5201 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5202 without a domain if the check to <postmaster@domain> fails.
5204 SC/01 Eximstats: added -xls and the ability to specify output files
5205 (patch written by Frank Heydlauf).
5207 SC/02 Eximstats: use FileHandles for outputting results.
5209 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5211 SC/04 Eximstats: fixed display of large numbers with -nvr option
5213 SC/05 Eximstats: fixed merging of reports with empty tables.
5215 SC/06 Eximstats: added the -include_original_destination flag
5217 SC/07 Eximstats: removed tabs and trailing whitespace.
5219 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5221 TK/06 MBOX spool code: Add real "From " MBOX separator line
5222 so the .eml file is really in mbox format (even though
5223 most programs do not really care). Patch from Alex Miller.
5225 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5226 The latter is generated from $received_to and is only set if the
5227 message has one envelope recipient. SA can use these headers,
5228 obviously out-of-the-box. Patch from Alex Miller.
5230 PH/08 The ${def test on a variable was returning false if the variable's
5231 value was "0", contrary to what the specification has always said!
5232 The result should be true unless the variable is empty.
5234 PH/09 The syntax error of a character other than { following "${if
5235 def:variable_name" (after optional whitespace) was not being diagnosed.
5236 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5237 accidental colon was present, for example, could give incorrect results.
5239 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5240 result is used (not including appendfile, where other changes are about
5243 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5244 This involved changing a lot of size variables from int to off_t. It
5245 should work with maildirs and everything.
5247 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5248 spamd dying while we are connected to it.
5250 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5253 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5254 function for BDB 4.3.
5256 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5258 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5259 the sources for PCRE have been re-arranged and more files are now
5262 PH/15 The code I had for printing potentially long long variables in PH/11
5263 above was not the best (it lost precision). The length of off_t variables
5264 is now inspected at build time, and an appropriate printing format (%ld
5265 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5266 to be "long long int" or "long int". This is needed for the internal
5267 formatting function string_vformat().
5269 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5270 the configuration file to be ":syslog", then the script "guesses" where
5271 the logs files are, rather than using the compiled in default. In our
5272 case the guess is not the same as the compiled default, so the script
5273 suddenly stopped working when I started to use syslog. The patch checks
5274 to see if log_file_path is "". If so, it attempts to read it from exim
5275 with no configuration file to get the compiled in version, before it
5276 falls back to the previous guessing code."
5278 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5279 implementing BATV in an Exim configuration. See NewStuff for the gory
5282 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5283 Makefile that are specific to HP-UX.
5285 PH/18 If the "use_postmaster" option was set for a recipient callout together
5286 with the "random" option, the postmaster address was used as the MAIL
5287 FROM address for the random test, but not for the subsequent recipient
5288 test. It is now used for both.
5290 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5291 patch removes a few documentation additions to RFC 3028, because the
5292 latest draft now contains them. It adds the new en;ascii-case comparator
5293 and a new error check for 8bit text in MIME parts. Comparator and
5294 require names are now matched exactly. I enabled the subaddress
5295 extension, but it is not well tested yet (read: it works for me)."
5297 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5298 rework some of the code of TK/09 above to avoid the hardwired use of
5299 "%lld" and "long long". Replaced the call to snprintf() with a call to
5302 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5303 records point to non-existent hosts", "retry timeout exceeded", and
5304 "retry time not reached for any host after a long failure period".
5306 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5307 experimental DomainKeys support:
5309 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5310 (2) On an error such as an illegally used "control", the wrong name for
5311 the control was given.
5313 These problems did NOT occur unless DomainKeys support was compiled.
5315 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5317 PH/24 Added ${if match_ip condition.
5319 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5320 databases so that it will be absolutely obvious if a crash occurs in the
5321 DB library. This is a regular occurrence (often caused by mis-matched
5324 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5325 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5326 picked up on a user's system that detects such things. There doesn't seem
5327 to be a gcc warning option for this - only an attribute that has to be
5328 put on the function's prototype. It seems that in Fedora Core 4 they have
5329 set this on a number of new functions. No doubt there will be more in due
5332 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5333 specify a transport. However, if an address that was verified by such a
5334 router was the subject of a callout, Exim crashed because it tried to
5335 read the rcpt_include_affixes from the non-existent transport. Now it
5336 just assumes that the setting of that option is false. This bug was
5337 introduced by 4.51/PH/31.
5339 PH/28 Changed -d+all to exclude +memory, because that information is very
5340 rarely of interest, but it makes the output a lot bigger. People tend to
5341 do -d+all out of habit.
5343 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5344 code in os-type was giving problems when libc.so lives in lib64, like on
5347 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5348 aren't the modern standard, and it seems that some systems' include files
5349 don't always have them. Exim was already checking for some of the newer
5350 ones like T_AAAA, and defining it itself. I've added checks for all the
5351 record types that Exim uses.
5353 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5354 not automatically generating a new one, as it is supposed to. This
5355 prevented TLS from working. If the file did exist, but contained invalid
5356 data, a new version was generated, as expected. It was only the case of a
5357 non-existent file that was broken.
5359 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5360 with a change in libdomainkeys > 0.64.
5362 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5363 from DNS. If the selector record carries the flag, it now has
5364 precedence over the domain-wide flag.
5366 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5368 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5369 the use of a port name that isn't defined in /etc/services) occurred, the
5370 message was deferred as in a normal delivery, and thus remained on the
5371 spool, instead of being failed because of the mua_wrapper setting. This
5372 is now fixed, and I tidied up some of the mua_wrapper messages at the
5375 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5376 the messages in a hash of arrays rather than using individual hashes.
5377 This is a bit cleaner and results in dramatic memory savings, albeit
5378 at a slight CPU cost.
5380 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5381 as requested by Marc Sherman.
5383 SC/10 Eximstats: added histograms for user specified patterns as requested
5386 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5388 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5389 fopen() in the content-scanning modules that did not already have it.
5395 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5396 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5398 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5400 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5402 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5403 to test Sieve filters that use "vacation".
5405 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5406 that changes the way the GnuTLS parameters are stored in the cache file.
5407 The new format can be generated externally. For backward compatibility,
5408 if the data in the cache doesn't make sense, Exim assumes it has read an
5409 old-format file, and it generates new data and writes a new file. This
5410 means that you can't go back to an older release without removing the
5413 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5414 work if there are any delivery delays because "one_time" forces the
5415 parent to be marked "delivered", so its unseen clone is never tried
5416 again. For this reason, Exim now forbids the simultaneous setting of
5419 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5420 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5421 entry if you want to know the details. The fix, however, neglected to
5422 consider the case when local delivery batching is involved. The test for
5423 "previously delivered" was not happening when checking to see if an
5424 address could be batched with a previous (undelivered) one; under
5425 certain circumstances this could lead to multiple deliveries to the same
5428 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5429 in its include files, and this causes problems building Exim.
5431 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5432 header_syntax) but Exim was just ignoring anything given after a slash.
5433 In particular, this caused confusion with an attempt to use "verify =
5434 reverse_host_lookup/defer_ok". An error is now given when options are
5435 supplied for verify items that do not have them. (Maybe reverse_host_
5436 lookup should have a defer_ok option, but that's a different point.)
5438 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5439 defined by RFC 821) to 2048, because there were problems with some AUTH
5440 commands, and RFC 1869 says the size should be increased for extended
5441 SMTP commands that take arguments.
5443 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5446 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5447 "unknown" error; now it says that the functionality isn't in the binary.
5449 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5450 an address' error message when a string expansion fails (syntax or
5451 whatever). Otherwise the password may appear in the log. Following change
5452 PH/42 below, there is no longer a chance of it appearing in a bounce
5455 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5457 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5458 of its addresses was ever tried. (Bugzilla bug #2.)
5460 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5461 the result incorrectly in the debug output. (It correctly added a newline
5462 to what was transported.)
5464 TF/01 Added $received_time.
5466 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5467 commented out examples of how to interface to a virus scanner and to
5468 SpamAssassin. Also added commented examples of av_scanner and
5469 spamd_address settings.
5471 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5472 and controls are allowed in which ACLs. There were a couple of minor
5473 errors. Some of the entries in the conditions table (which is a table of
5474 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5475 negation of where the condition IS allowed.
5477 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5479 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5480 header file does not have a version number, so I've had to invent a new
5481 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5482 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5483 radiusclient), but it was contributed by a Radius user.
5485 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5486 files or directories, for OpenSSL.
5488 PH/19 When an Exim process that is running as root has to create an Exim log
5489 file, it does so in a subprocess that runs as exim:exim so as to get the
5490 ownership right at creation (otherwise, other Exim processes might see
5491 the file with the wrong ownership). There was no test for failure of this
5492 fork() call, which would lead to the process getting stuck as it waited
5493 for a non-existent subprocess. Forks do occasionally fail when resources
5494 run out. I reviewed all the other calls to fork(); they all seem to check
5497 PH/20 When checking for unexpected SMTP input at connect time (before writing
5498 the banner), Exim was not dealing correctly with a non-positive return
5499 from the read() function. If the client had disconnected by this time,
5500 the result was a log entry for a synchronization error with an empty
5501 string after "input=" when read() returned zero. If read() returned -1
5502 (an event I could not check), uninitialized data bytes were printed.
5503 There were reports of junk text (parts of files, etc) appearing after
5506 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5508 PH/22 Added support for macro redefinition, and (re)definition in between
5509 driver and ACL definitions.
5511 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5512 forgetting to use the resulting value; it was using the unexpanded value.
5514 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5515 hadn't been configured. The fix is from Juergen Kreileder, who
5516 understands it better than I do:
5518 "Here's what I see happening with three configured cyrus_sasl
5519 authenticators configured (plain, login, cram-md5):
5521 On startup auth_cyrus_sasl_init() gets called for each of these.
5522 This means three calls to sasl_listmech() without a specified mech_list.
5523 => SASL tests which mechs of all available mechs actually work
5524 => three warnings about OTP not working
5525 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5527 With the patch, sasl_listmech() also gets called three times. But now
5528 SASL's mech_list option is set to the server_mech specified in the the
5529 authenticator. Or in other words, the answer from sasl_listmech()
5530 gets limited to just the mech you're testing for (which is different
5532 => the return list contains just 'plain' or 'login', 'cram-md5' or
5533 nothing depending on the value of ob->server_mech.
5535 I've just tested the patch: Authentication still works fine,
5536 unavailable mechs specified in the exim configuration are still
5537 caught, and the auth.log warnings about OTP are gone."
5539 PH/25 When debugging is enabled, the contents of the command line are added
5540 to the debugging output, even when log_selector=+arguments is not
5543 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5544 answer is "GNU", and only if the return is "GNU/something" is the answer
5547 PH/27 $acl_verify_message is now set immediately after the failure of a
5548 verification in an ACL, and so is available in subsequent modifiers. In
5549 particular, the message can be preserved by coding like this:
5551 warn !verify = sender
5552 set acl_m0 = $acl_verify_message
5554 Previously, $acl_verify_message was set only while expanding "message"
5555 and "log_message" when a very denied access.
5557 PH/28 Modified OS/os.c-Linux with
5559 -#ifndef OS_LOAD_AVERAGE
5560 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5562 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5563 nomenclature these days.)
5565 PH/29 Installed patch from the Sieve maintainer that adds the options
5566 sieve_useraddress and sieve_subaddress to the redirect router.
5568 PH/30 In these circumstances:
5569 . Two addresses routed to the same list of hosts;
5570 . First host does not offer TLS;
5571 . First host accepts first address;
5572 . First host gives temporary error to second address;
5573 . Second host offers TLS and a TLS session is established;
5574 . Second host accepts second address.
5575 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5576 and peerdn, if requested) that were in fact used only for the second
5579 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5580 attention to any local part prefix or suffix that was matched by the
5581 router that accepted the address. It now behaves in the same way as it
5582 does for delivery: the affixes are removed from the local part unless
5583 rcpt_include_affixes is set on the transport.
5585 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5586 timeout during the DATA phase of an incoming message.
5588 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5589 applied a patch sanctioned by the Sieve maintainer.
5591 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5592 the uid or gid is negative. A case of a negative gid caused this to be
5593 noticed. The fix allows for either to be negative.
5595 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5596 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5599 PH/36 The change PH/12 above was broken. Fixed it.
5601 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5602 the grounds that routing the same address twice would always produce the
5603 same answer. This might have been true once, but it is certainly no
5604 longer true now. Routing a child address may depend on the previous
5605 routing that produced that child. Some complicated redirection strategies
5606 went wrong when messages had multiple recipients, and made Exim's
5607 behaviour dependent on the order in which the addresses were given.
5609 I have moved the duplicate checking until after the routing is complete.
5610 Exim scans the addresses that are assigned to local and remote
5611 transports, and removes any duplicates. This means that more work will be
5612 done, as duplicates will always all be routed, but duplicates are
5613 presumably rare, so I don't expect this is of any significance.
5615 For deliveries to pipes, files, and autoreplies, the duplicate checking
5616 still happens during the routing process, since they are not going to be
5619 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5620 It corrects a timeout issue with spamd. This is Ian's comment: "The
5621 background is that sometimes spamd either never reads data from a
5622 connection it has accepted, or it never writes response data. The exiscan
5623 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5624 blindly assumes that writes won't block so it may never time out."
5626 PH/39 Allow G after quota size as well as K and M.
5628 PH/40 The value set for $authenticated_id in an authenticator may not contain
5629 binary zeroes or newlines because the value is written to log lines and
5630 to spool files. There was no check on this. Now the value is run through
5631 the string_printing() function so that such characters are converted to
5632 printable escape sequences.
5634 PH/41 $message_linecount is a new variable that contains the total number of
5635 lines in the message. Compare $body_linecount, which is the count for the
5638 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5639 bounce and delay warning messages, except in certain special cases, which
5642 (a) An SMTP error message from a remote host;
5643 (b) A message specified in a :fail: redirection;
5644 (c) A message specified in a "fail" command in a system filter;
5645 (d) A message specified in a FAIL return from the queryprogram router;
5646 (e) A message specified by the cannot_route_message router option.
5648 In these cases only, Exim does include the error details in bounce and
5649 warning messages. There are also a few cases where bland messages such
5650 as "unrouteable address" or "local delivery error" are given.
5652 PH/43 $value is now also set for the "else" part of a ${run expansion.
5654 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5655 being worked on, but at least Exim now implements the latest version to
5658 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5659 process to complete was treated as a delivery failure, a timeout while
5660 writing the message to the pipe was logged, but erroneously treated as a
5661 successful delivery. Such timeouts include transport filter timeouts. For
5662 consistency with the overall process timeout, these timeouts are now
5663 treated as errors, giving rise to delivery failures by default. However,
5664 there is now a new Boolean option for the pipe transport called
5665 timeout_defer, which, if set TRUE, converts the failures into defers for
5666 both kinds of timeout. A transport filter timeout is now identified in
5669 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5670 systems where "make" and "gmake" are different, calling "gmake" at top
5671 level broke things. I've arranged for the value of $(MAKE) to be passed
5672 from the Makefile to this script so that it can call the same version of
5676 A note about Exim versions 4.44 and 4.50
5677 ----------------------------------------
5679 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5680 changes of various kinds. As a consequence, a big documentation update was
5681 needed. This delayed the release for rather longer than seemed good, especially
5682 in the light of a couple of (minor) security issues. Therefore, the changes
5683 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5684 release. So 4.44 and 4.50 are in effect two different branches that both start
5687 I have left the 4.50 change log unchanged; it contains all the changes since
5688 4.43. The change log for 4.44 is below; many of its items are identical to
5689 those for 4.50. This seems to be the most sensible way to preserve the
5690 historical information.
5696 1. Minor wording change to the doc/README.SIEVE file.
5698 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5699 computation of the current number of files was incorrect.
5701 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5702 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5705 4. Give more explanation in the error message when the command for a transport
5706 filter fails to execute.
5708 5. There are several places where Exim runs a non-Exim command in a
5709 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5710 was being done only for the command run by the queryprogram router. It is
5711 now done for all such subprocesses. The other cases are: ${run, transport
5712 filters, and the commands run by the lmtp and pipe transports.
5714 6. Added CONFIGURE_GROUP build-time option.
5716 7. Some older OS have a limit of 256 on the maximum number of file
5717 descriptors. Exim was using setrlimit() to set 1000 as a large value
5718 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5719 systems. I've change it so that if it can't get 1000, it tries for 256.
5721 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5722 was an oversight, and furthermore, ever since the addition of extra
5723 controls (e.g. 4.43/32), the checks on when to allow different forms of
5724 "control" were broken. There should now be diagnostics for all cases when a
5725 control that does not make sense is encountered.
5727 9. Added the /retain_sender option to "control=submission".
5729 10. $recipients is now available in the predata ACL (oversight).
5731 11. Tidy the search cache before the fork to do a delivery from a message
5732 received from the command line. Otherwise the child will trigger a lookup
5733 failure and thereby defer the delivery if it tries to use (for example) a
5734 cached ldap connection that the parent has called unbind on.
5736 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5737 of $address_data from the recipient verification was clobbered by the
5738 sender verification.
5740 13. The value of address_data from a sender verification is now available in
5741 $sender_address_data in subsequent conditions in the ACL statement.
5743 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5745 15. Added a new option "connect=<time>" to callout options, to set a different
5748 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5749 was its contents. (It was OK if the option was not defined at all.)
5751 17. A "Completed" log line is now written for messages that are removed from
5752 the spool by the -Mrm option.
5754 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5755 information about exactly what failed.
5757 19. Added -dd to debug only the daemon process.
5759 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5760 handles timeouts, both on the server side and network timeouts. Renamed the
5761 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5763 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5764 It is now set to "smtps".
5766 22. $host_address is now set to the target address during the checking of
5767 ignore_target_hosts.
5769 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5770 being passed; this would have caused $sender_host_name to have been used if
5771 matching the list had actually called for a host name (not very likely,
5772 since this list is usually IP addresses). A host name is now passed as
5775 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5776 code that specifies a non-restarting handler (typically sigaction() in
5777 modern systems) in an attempt to fix a rare and obscure crash bug.
5779 25. Narrowed the window for a race in the daemon that could cause it to ignore
5780 SIGCHLD signals. This is not a major problem, because they are used only to
5781 wake it up if nothing else does.
5783 26. A malformed maildirsize file could cause Exim to calculate negative values
5784 for the mailbox size or file count. Odd effects could occur as a result.
5785 The maildirsize information is now recalculated if the size or filecount
5788 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5789 support for a long time. Removed HAVE_SYS_VFS_H.
5791 28. Installed the latest version of exipick from John Jetmore.
5793 29. In an address list, if the pattern was not a regular expression, an empty
5794 subject address (from a bounce message) matched only if the pattern was an
5795 empty string. Non-empty patterns were not even tested. This was the wrong
5796 because it is perfectly reasonable to use an empty address as part of a
5797 database query. An empty address is now tested by patterns that are
5798 lookups. However, all the other forms of pattern expect the subject to
5799 contain a local part and a domain, and therefore, for them, an empty
5800 address still always fails if the pattern is not itself empty.
5802 30. Exim went into a mad DNS loop when attempting to do a callout where the
5803 host was specified on an smtp transport, and looking it up yielded more
5804 than one IP address.
5806 31. Re-factored the code for checking spool and log partition space into a
5807 function that finds that data and another that does the check. The former
5808 is then used to implement four new variables: $spool_space, $log_space,
5809 $spool_inodes, and $log_inodes.
5811 32. The RFC2047 encoding function was originally intended for short strings
5812 such as real names; it was not keeping to the 75-character limit for
5813 encoded words that the RFC imposes. It now respects the limit, and
5814 generates multiple encoded words if necessary. To be on the safe side, I
5815 have increased the buffer size for the ${rfc2047: expansion operator from
5818 33. It is now permitted to omit both strings after an "if" condition; if the
5819 condition is true, the result is "true". As before, when the second string
5820 is omitted, a false condition yields an empty string. This makes it less
5821 cumbersome to write custom ACL and router conditions.
5823 34. Failure to deliver a bounce message always caused it to be frozen, even if
5824 there was an errors_to setting on the router. The errors_to setting is now
5827 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5828 canonical form (fully expanded) before being placed in
5829 $sender_host_address.
5831 36. The table in the code that translates DNS record types into text (T_A to
5832 "A" for instance) was missing entries for NS and CNAME. It is just possible
5833 that this could have caused confusion if both these types were looked up
5834 for the same domain, because the text type is used as part of Exim's
5835 per-process caching. But the chance of anyone hitting this buglet seems
5838 37. The dnsdb lookup has been extended in a number of ways.
5840 (1) There is a new type, "zns", which walks up the domain tree until it
5841 finds some nameserver records. It should be used with care.
5843 (2) There is a new type, "mxh", which is like "mx" except that it returns
5844 just the host names, not the priorities.
5846 (3) It is now possible to give a list of domains (or IP addresses) to be
5847 looked up. The behaviour when one of the lookups defers can be
5848 controlled by a keyword.
5850 (4) It is now possible to specify the separator character for use when
5851 multiple records are returned.
5853 38. The dnslists ACL condition has been extended: it is now possible to supply
5854 a list of IP addresses and/or domains to be looked up in a particular DNS
5857 39. Added log_selector=+queue_time_overall.
5859 40. When running the queue in the test harness, wait just a tad after forking a
5860 delivery process, to get repeatability of debugging output.
5862 41. Include certificate and key file names in error message when GnuTLS fails
5863 to set them up, because the GnuTLS error message doesn't include the name
5864 of the failing file when there is a problem reading it.
5866 42. Allow both -bf and -bF in the same test run.
5868 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5870 44. The "Exiscan patch" is now merged into the mainline Exim source.
5872 45. Sometimes the final signoff response after QUIT could fail to get
5873 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5874 < 0 before doing a fflush(). This bug looks as though it goes back to the
5875 introduction of TLS in release 3.20, but "sometimes" must have been rare
5876 because the tests only now provoked it.
5878 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5879 (this can affect the format of dates).
5881 47. exim_tidydb, when checking for the continued existence of a message for
5882 which it has found a message-specific retry record, was not finding
5883 messages that were in split spool directories. Consequently, it was
5884 deleting retry records that should have stayed in existence.
5886 48. Steve fixed some bugs in eximstats.
5888 49. The SPA authentication driver was not abandoning authentication and moving
5889 on to the next authenticator when an expansion was forced to fail,
5890 contradicting the general specification for all authenticators. Instead it
5891 was generating a temporary error. It now behaves as specified.
5893 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5894 (the order specifies the preference for clients). The order is now AES256,
5895 AES128, 3DES, ARCFOUR128.
5897 51. Small patch to Sieve code - explicitly set From: when generating an
5900 52. Exim crashed if a remote delivery caused a very long error message to be
5901 recorded - for instance if somebody sent an entire SpamAssassin report back
5902 as a large number of 550 error lines. This bug was coincidentally fixed by
5903 increasing the size of one of Exim's internal buffers (big_buffer) that
5904 happened as part of the Exiscan merge. However, to be on the safe side, I
5905 have made the code more robust (and fixed the comments that describe what
5908 53. Now that there can be additional text after "Completed" in log lines (if
5909 the queue_time_overall log selector is set), a one-byte patch to exigrep
5910 was needed to allow it to recognize "Completed" as not the last thing in
5913 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5914 patch that reportedly fixes this has been added. I am not expert enough to
5915 create a test for it. This is what the patch creator wrote:
5917 "I found a little strange behaviour of ldap code when working with
5918 Windows 2003 AD Domain, where users was placed in more than one
5919 Organization Units. When I tried to give exim partial DN, the exit code
5920 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5921 But simultaneously result of request was absolutely normal ldap result,
5922 so I produce this patch..."
5924 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5925 REFERENCE, so I have modified the code to exclude the patch when that macro
5928 55. Some experimental protocols are using DNS PTR records for new purposes. The
5929 keys for these records are domain names, not reversed IP addresses. The
5930 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5931 leaves it alone. Component reversal etc. now happens only for IP addresses.
5934 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5936 57. Double the size of the debug message buffer (to 2048) so that more of very
5937 long debug lines gets shown.
5939 58. The exicyclog utility now does better if the number of log files to keep
5940 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5942 59. Two changes related to the smtp_active_hostname option:
5944 (1) $smtp_active_hostname is now available as a variable.
5945 (2) The default for smtp_banner uses $smtp_active_hostname instead
5946 of $primary_hostname.
5948 60. The host_aton() function is supposed to be passed a string that is known
5949 to be a valid IP address. However, in the case of IPv6 addresses, it was
5950 not checking this. This is a hostage to fortune. Exim now panics and dies
5951 if the condition is not met. A case was found where this could be provoked
5952 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5953 components; fortuitously, this particular loophole had already been fixed
5954 by change 4.50/55 above.
5956 If there are any other similar loopholes, the new check in host_aton()
5957 itself should stop them being exploited. The report I received stated that
5958 data on the command line could provoke the exploit when Exim was running as
5959 exim, but did not say which command line option was involved. All I could
5960 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5961 running as the user.
5964 61. There was a buffer overflow vulnerability in the SPA authentication code
5965 (which came originally from the Samba project). I have added a test to the
5966 spa_base64_to_bits() function which I hope fixes it.
5969 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5970 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5972 63. The daemon start-up calls getloadavg() while still root for those OS that
5973 need the first call to be done as root, but it missed one case: when
5974 deliver_queue_load_max is set with deliver_drop_privilege. This is
5975 necessary for the benefit of the queue runner, because there is no re-exec
5976 when deliver_drop_privilege is set.
5978 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5979 This has been fixed.
5981 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5982 was in use, was not putting the data itself into the right store pool;
5983 consequently, it could be overwritten for a subsequent message in the same
5984 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5987 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5989 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5990 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5991 the function were treating the return as a boolean value, which happened to
5992 work because 0=false and not-0=true, but is not correct code.
5994 68. The host_aton() function was not handling scoped IPv6 addresses (those
5995 with, for example, "%eth0" on the end) correctly.
5997 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5998 negated items (that is, ~something) in unsigned ints. Some compilers
5999 apparently mutter when there is no cast.
6001 70. If an address verification called from an ACL failed, and did not produce a
6002 user-specific message (i.e. there was only a "system" message), nothing was
6003 put in $acl_verify_message. In this situation, it now puts the system
6006 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6007 session; change 4.31/43 added the unwanted input to the log line - except
6008 that it did not do this in the start of session case. It now does.
6010 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6011 This is wrong and can cause the other end to generate a synchronization
6012 error if it is another Exim or anything else that does the synchronization
6013 check. A QUIT command is no longer sent after a timeout.
6015 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6016 during host lookups.
6018 74. The defer_ok option of callout verification was not working if it was used
6019 when verifying addresses in header lines, that is, for this case:
6021 verify = header_sender/callout=defer_ok
6023 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6024 those file descriptors could be used for SMTP connections. If anything
6025 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6026 could be sent to the SMTP client, causing chaos. The daemon now opens
6027 stdin, stdout, and stderr to /dev/null when it puts itself into the
6030 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6031 log by default. The user can override this with suitable Perl magic.
6033 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6034 the log message when discard triggers, was not working for the DATA ACL or
6035 for the non-SMTP ACL.
6037 78. Error message wording change in sieve.c.
6039 79. If smtp_accept_max_per_host was set, the number of connections could be
6040 restricted to fewer than expected, because the daemon was trying to set up
6041 a new connection before checking whether the processes handling previous
6042 connections had finished. The check for completed processes is now done
6043 earlier. On busy systems, this bug wouldn't be noticed because something
6044 else would have woken the daemon, and it would have reaped the completed
6047 80. If a message was submitted locally by a user whose login name contained one
6048 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6049 It caused a spool format error. I have fixed the spool reading code. A
6050 related problem was that the "from" clause in the Received: line became
6051 illegal because of the space(s). It is now covered by ${quote_local_part.
6053 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6056 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6058 83. Patch from Sieve maintainer to fix unterminated string problem in
6059 "vacation" handling.
6061 84. Some minor changes to the Linux configuration files to help with other
6062 OS variants using glibc.
6064 85. One more patch for Sieve to update vacation handling to latest spec.
6067 ----------------------------------------------------
6068 See the note above about the 4.44 and 4.50 releases.
6069 ----------------------------------------------------
6075 1. Change 4.43/35 introduced a bug that caused file counts to be
6076 incorrectly computed when quota_filecount was set in an appendfile
6079 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6080 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6083 3. Give more explanation in the error message when the command for a transport
6084 filter fails to execute.
6086 4. There are several places where Exim runs a non-Exim command in a
6087 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6088 was being done only for the command run by the queryprogram router. It is
6089 now done for all such subprocesses. The other cases are: ${run, transport
6090 filters, and the commands run by the lmtp and pipe transports.
6092 5. Some older OS have a limit of 256 on the maximum number of file
6093 descriptors. Exim was using setrlimit() to set 1000 as a large value
6094 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6095 systems. I've change it so that if it can't get 1000, it tries for 256.
6097 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6098 was an oversight, and furthermore, ever since the addition of extra
6099 controls (e.g. 4.43/32), the checks on when to allow different forms of
6100 "control" were broken. There should now be diagnostics for all cases when a
6101 control that does not make sense is encountered.
6103 7. $recipients is now available in the predata ACL (oversight).
6105 8. Tidy the search cache before the fork to do a delivery from a message
6106 received from the command line. Otherwise the child will trigger a lookup
6107 failure and thereby defer the delivery if it tries to use (for example) a
6108 cached ldap connection that the parent has called unbind on.
6110 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6111 of $address_data from the recipient verification was clobbered by the
6112 sender verification.
6114 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6115 was its contents. (It was OK if the option was not defined at all.)
6117 11. A "Completed" log line is now written for messages that are removed from
6118 the spool by the -Mrm option.
6120 12. $host_address is now set to the target address during the checking of
6121 ignore_target_hosts.
6123 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6124 being passed; this would have caused $sender_host_name to have been used if
6125 matching the list had actually called for a host name (not very likely,
6126 since this list is usually IP addresses). A host name is now passed as
6129 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6130 code that specifies a non-restarting handler (typically sigaction() in
6131 modern systems) in an attempt to fix a rare and obscure crash bug.
6133 15. Narrowed the window for a race in the daemon that could cause it to ignore
6134 SIGCHLD signals. This is not a major problem, because they are used only to
6135 wake it up if nothing else does.
6137 16. A malformed maildirsize file could cause Exim to calculate negative values
6138 for the mailbox size or file count. Odd effects could occur as a result.
6139 The maildirsize information is now recalculated if the size or filecount
6142 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6143 support for a long time. Removed HAVE_SYS_VFS_H.
6145 18. Updated exipick to current release from John Jetmore.
6147 19. Allow an empty sender to be matched against a lookup in an address list.
6148 Previously the only cases considered were a regular expression, or an
6151 20. Exim went into a mad DNS lookup loop when doing a callout where the
6152 host was specified on the transport, if the DNS lookup yielded more than
6155 21. The RFC2047 encoding function was originally intended for short strings
6156 such as real names; it was not keeping to the 75-character limit for
6157 encoded words that the RFC imposes. It now respects the limit, and
6158 generates multiple encoded words if necessary. To be on the safe side, I
6159 have increased the buffer size for the ${rfc2047: expansion operator from
6162 22. Failure to deliver a bounce message always caused it to be frozen, even if
6163 there was an errors_to setting on the router. The errors_to setting is now
6166 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6167 canonical form (fully expanded) before being placed in
6168 $sender_host_address.
6170 24. Updated eximstats to version 1.33
6172 25. Include certificate and key file names in error message when GnuTLS fails
6173 to set them up, because the GnuTLS error message doesn't include the name
6174 of the failing file when there is a problem reading it.
6176 26. Expand error message when OpenSSL has problems setting up cert/key files.
6179 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6180 (this can affect the format of dates).
6182 28. exim_tidydb, when checking for the continued existence of a message for
6183 which it has found a message-specific retry record, was not finding
6184 messages that were in split spool directories. Consequently, it was
6185 deleting retry records that should have stayed in existence.
6187 29. eximstats updated to version 1.35
6188 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6189 1.35 - bugfix such that pie charts by volume are generated correctly
6191 30. The SPA authentication driver was not abandoning authentication and moving
6192 on to the next authenticator when an expansion was forced to fail,
6193 contradicting the general specification for all authenticators. Instead it
6194 was generating a temporary error. It now behaves as specified.
6196 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6197 (the order specifies the preference for clients). The order is now AES256,
6198 AES128, 3DES, ARCFOUR128.
6200 31. Small patch to Sieve code - explicitly set From: when generating an
6203 32. Exim crashed if a remote delivery caused a very long error message to be
6204 recorded - for instance if somebody sent an entire SpamAssassin report back
6205 as a large number of 550 error lines. This bug was coincidentally fixed by
6206 increasing the size of one of Exim's internal buffers (big_buffer) that
6207 happened as part of the Exiscan merge. However, to be on the safe side, I
6208 have made the code more robust (and fixed the comments that describe what
6211 33. Some experimental protocols are using DNS PTR records for new purposes. The
6212 keys for these records are domain names, not reversed IP addresses. The
6213 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6214 leaves it alone. Component reversal etc. now happens only for IP addresses.
6217 34. The host_aton() function is supposed to be passed a string that is known
6218 to be a valid IP address. However, in the case of IPv6 addresses, it was
6219 not checking this. This is a hostage to fortune. Exim now panics and dies
6220 if the condition is not met. A case was found where this could be provoked
6221 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6222 components; fortuitously, this particular loophole had already been fixed
6223 by change 4.50/55 or 4.44/33 above.
6225 If there are any other similar loopholes, the new check in host_aton()
6226 itself should stop them being exploited. The report I received stated that
6227 data on the command line could provoke the exploit when Exim was running as
6228 exim, but did not say which command line option was involved. All I could
6229 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6230 running as the user.
6233 35. There was a buffer overflow vulnerability in the SPA authentication code
6234 (which came originally from the Samba project). I have added a test to the
6235 spa_base64_to_bits() function which I hope fixes it.
6238 36. The daemon start-up calls getloadavg() while still root for those OS that
6239 need the first call to be done as root, but it missed one case: when
6240 deliver_queue_load_max is set with deliver_drop_privilege. This is
6241 necessary for the benefit of the queue runner, because there is no re-exec
6242 when deliver_drop_privilege is set.
6244 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6245 was in use, was not putting the data itself into the right store pool;
6246 consequently, it could be overwritten for a subsequent message in the same
6247 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6250 38. Sometimes the final signoff response after QUIT could fail to get
6251 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6252 < 0 before doing a fflush(). This bug looks as though it goes back to the
6253 introduction of TLS in release 3.20, but "sometimes" must have been rare
6254 because the tests only now provoked it.
6260 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6261 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6262 Now it returns an integer. A number of calls were still expecting a T/F
6263 return. Fortuitously, in all cases, the tests worked in OK situations,
6264 which is the norm. However, things would have gone wrong on any write
6265 failures on the smtp file descriptor. This function is used when sending
6266 messages over SMTP and also when doing verify callouts.
6268 2. When Exim is called to do synchronous delivery of a locally submitted
6269 message (the -odf or -odi options), it no longer closes stderr before doing
6272 3. Implemented the mua_wrapper option.
6274 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6276 5. Implemented the functions header_remove(), header_testname(),
6277 header_add_at_position(), and receive_remove_recipient(), and exported them
6280 6. If an ACL "warn" statement specified the addition of headers, Exim already
6281 inserted X-ACL-Warn: at the start if there was no header name. However, it
6282 was not making this test for the second and subsequent header lines if
6283 there were newlines in the string. This meant that an invalid header could
6284 be inserted if Exim was badly configured.
6286 7. Allow an ACL "warn" statement to add header lines at the start or after all
6287 the Received: headers, as well as at the end.
6289 8. Added the rcpt_4xx retry error code.
6291 9. Added postmaster_mailfrom=xxx to callout verification option.
6293 10. Added mailfrom=xxxx to the callout verification option, for verify=
6296 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6297 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6299 12. Inserted some casts to stop certain compilers warning when using pointer
6300 differences as field lengths or precisions in printf-type calls (mostly
6301 affecting debugging statements).
6303 13. Added optional readline() support for -be (dynamically loaded).
6305 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6306 same clock tick as a message's arrival, so that its received time was the
6307 same as the "first fail" time on the retry record, and that message
6308 remained on the queue past the ultimate address timeout, every queue runner
6309 would try a delivery (because it was past the ultimate address timeout) but
6310 after another failure, the ultimate address timeout, which should have then
6311 bounced the address, did not kick in. This was a "< instead of <=" error;
6312 in most cases the first failure would have been in the next clock tick
6313 after the received time, and all would be well.
6315 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6316 being recognized when the domain list was tested by the match_domain
6317 condition in an expansion string.
6319 16. Added the ${str2b64: operator.
6321 17. Exim was always calling setrlimit() to set a large limit for the number of
6322 processes, without checking whether the existing limit was already
6323 adequate. (It did check for the limit on file descriptors.) Furthermore,
6324 errors from getrlimit() and setrlimit() were being ignored. Now they are
6325 logged to the main and panic logs, but Exim does carry on, to try to do its
6326 job under whatever limits there are.
6328 18. Imported PCRE 5.0.
6330 19. Trivial typo in log message " temporarily refused connection" (the leading
6333 20. If the log selector return_path_on_delivery was set and an address was
6334 redirected to /dev/null, the delivery process crashed because it assumed
6335 that a return path would always be set for a "successful" delivery. In this
6336 case, the whole delivery is bypassed as an optimization, and therefore no
6339 21. Internal re-arrangement: the function for sending a challenge and reading
6340 a response while authentication was assuming a zero-terminated challenge
6341 string. It's now changed to take a pointer and a length, to allow for
6342 binary data in such strings.
6344 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6346 23. Exim was not respecting finduser_retries when seeking the login of the
6347 uid under which it was called; it was always trying 10 times. (The default
6348 setting of finduser_retries is zero.) Also, it was sleeping after the final
6349 failure, which is pointless.
6351 24. Implemented tls_on_connect_ports.
6353 25. Implemented acl_smtp_predata.
6355 26. If the domain in control=submission is set empty, Exim assumes that the
6356 authenticated id is a complete email address when it generates From: or
6357 Sender: header lines.
6359 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6360 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6361 chown and chgrp in /bin and hostname in /usr/bin.
6363 28. Exim was keeping the "process log" file open after each use, just as it
6364 does for the main log. This opens the possibility of it remaining open for
6365 long periods when the USR1 signal hits a daemon. Occasional processlog
6366 errors were reported, that could have been caused by this. Anyway, it seems
6367 much more sensible not to leave this file open at all, so that is what now
6370 29. The long-running daemon process does not normally write to the log once it
6371 has entered its main loop, and it closes the log before doing so. This is
6372 so that log files can straightforwardly be renamed and moved. However,
6373 there are a couple of unusual error situations where the daemon does write
6374 log entries, and I had neglected to close the log afterwards.
6376 30. The text of an SMTP error response that was received during a remote
6377 delivery was being truncated at 512 bytes. This is too short for some of
6378 the long messages that one sometimes sees. I've increased the limit to
6381 31. It is now possible to make retry rules that apply only when a message has a
6382 specific sender, in particular, an empty sender.
6384 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6385 it possible to be selective about when SMTP synchronization is enforced.
6387 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6389 32. Implemented hosts_connection_nolog.
6391 33. Added an ACL for QUIT.
6393 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6396 35. Added mailbox_size and mailbox_filecount to appendfile.
6398 36. Added control = no_multiline_responses to ACLs.
6400 37. There was a bug in the logic of the code that waits for the clock to tick
6401 in the case where the clock went backwards by a substantial amount such
6402 that the microsecond fraction of "now" was more than the microsecond
6403 fraction of "then" (but the whole seconds number was less).
6405 38. Added support for the libradius Radius client library this is found on
6406 FreeBSD (previously only the radiusclient library was supported).
6412 1. When certain lookups returned multiple values in the form name=value, the
6413 quoting of the values was not always being done properly. Specifically:
6414 (a) If the value started with a double quote, but contained no whitespace,
6416 (b) If the value contained whitespace other than a space character (i.e.
6417 tabs or newlines or carriage returns) it was not quoted.
6418 This fix has been applied to the mysql and pgsql lookups by writing a
6419 separate quoting function and calling it from the lookup code. The fix
6420 should probably also be applied to nisplus, ibase and oracle lookups, but
6421 since I cannot test any of those, I have not disturbed their existing code.
6423 2. A hit in the callout cache for a specific address caused a log line with no
6424 reason for rejecting RCPT. Now it says "Previous (cached) callout
6425 verification failure".
6427 3. There was an off-by-one bug in the queryprogram router. An over-long
6428 return line was truncated at 256 instead of 255 characters, thereby
6429 overflowing its buffer with the terminating zero. As well as fixing this, I
6430 have increased the buffer size to 1024 (and made a note to document this).
6432 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6433 when Exim is waiting for an SMTP response from a remote server, Exim
6434 restarts its select() call on the socket, thereby resetting its timeout.
6435 This is not a problem when such interrupts are rare. Somebody set up a cron
6436 job to run exiwhat every 2 minutes, which is less than the normal select()
6437 timeout (5 or 10 minutes). This meant that the select() timeout never
6438 kicked in because it was always reset. I have fixed this by comparing the
6439 time when an interrupt arrives with the time at the start of the first call
6440 to select(). If more time than the timeout has elapsed, the interrupt is
6441 treated as a timeout.
6443 5. Some internal re-factoring in preparation for the addition of Sieve
6444 extensions (by MH). In particular, the "personal" test is moved to a
6445 separate function, and given an option for scanning Cc: and Bcc: (which is
6446 not set for Exim filters).
6448 6. When Exim created an email address using the login of the caller as the
6449 local part (e.g. when creating a From: or Sender: header line), it was not
6450 quoting the local part when it contained special characters such as @.
6452 7. Installed new OpenBSD configuration files.
6454 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6455 try to make them clearer.
6457 9. Callout options, other than the timeout value, were being ignored when
6458 verifying sender addresses in header lines. For example, when using
6460 verify = header_sender/callout=no_cache
6462 the cache was (incorrectly) being used.
6464 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6465 only the Cygwin environment.
6467 11. When return_path_on_delivery was set as a log selector, if different remote
6468 addresses in the same message used different return paths and parallel
6469 remote delivery occurred, the wrong values would sometimes be logged.
6470 (Whenever a remote delivery process finished, the return path value from
6471 the most recently started remote delivery process was logged.)
6473 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6474 lines when AUTH and/or TLS are in use. This is the "received protocol"
6475 field. Exim used to use "asmtp" for authenticated SMTP, without any
6476 indication (in the protocol name) for TLS use. Now it follows the RFC and
6477 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6478 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6479 names appear in log lines as well as in Received: header lines.
6481 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6482 extensions, and comparison tests, and to fix some bugs.
6484 14. Changes to the "personal" filter test:
6486 (1) The test was buggy in that it was just doing the equivalent of
6487 "contains" tests on header lines. For example, if a user's address was
6488 anne@some.where, the "personal" test would incorrectly be true for
6490 To: susanne@some.where
6492 This test is now done by extracting each address from the header in turn,
6493 and checking the entire address. Other tests that are part of "personal"
6494 are now done using regular expressions (for example, to check local parts
6495 of addresses in From: header lines).
6497 (2) The list of non-personal local parts in From: addresses has been
6498 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6499 taken from the Sieve specification recommendations.
6501 (3) If the message contains any header line starting with "List-" it is
6502 treated as non-personal.
6504 (4) The test for "circular" in the Subject: header line has been removed
6505 because it now seems ill-conceived.
6507 15. Minor typos in src/EDITME comments corrected.
6509 16. Installed latest exipick from John Jetmore.
6511 17. If headers_add on a router specified a text string that was too long for
6512 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6513 of string_sprintf() is now avoided.
6515 18. $message_body_size was not set (it was always zero) when running the DATA
6516 ACL and the local_scan() function.
6518 19. For the "mail" command in an Exim filter, no default was being set for
6519 the once_repeat time, causing a random time value to be used if "once" was
6520 specified. (If the value happened to be <= 0, no repeat happened.) The
6521 default is now 0s, meaning "never repeat". The "vacation" command was OK
6522 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6523 (I found it when inspecting the code).
6525 20. There is now an overall timeout for performing a callout verification. It
6526 defaults to 4 times the callout timeout, which applies to individual SMTP
6527 commands during the callout. The overall timeout applies when there is more
6528 than one host that can be tried. The timeout is checked before trying the
6529 next host. This prevents very long delays if there are a large number of
6530 hosts and all are timing out (e.g. when the network connections are timing
6531 out). The value of the overall timeout can be changed by specifying an
6532 additional sub-option for "callout", called "maxwait". For example:
6534 verify = sender/callout=5s,maxwait=20s
6536 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6537 to the end before writing, but this should make it even safer).
6539 22. Exim was forgetting that it had advertised PIPELINING for the second and
6540 subsequent messages on an SMTP connection. It was also not resetting its
6541 memory on STARTTLS and an internal HELO.
6543 23. When Exim logs an SMTP synchronization error within a session, it now
6544 records whether PIPELINING has been advertised or not.
6546 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6547 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6548 rather than long int.
6550 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6552 26. Added the never_mail option to autoreply.
6558 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6559 crash if the getsockname() call failed; this can happen if a connection is
6560 closed very soon after it is established. The problem was simply in the
6561 order in which certain operations were done, causing Exim to try to write
6562 to the SMTP stream before it had set up the file descriptor. The bug has
6563 been fixed by making things happen in the correct order.
6569 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6570 before the connection was closed, thus losing the rejection response.
6572 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6573 some early Solaris releases, but causes trouble in current releases where
6574 socklen_t is defined.
6576 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6579 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6582 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6583 that releasing the top of it at the end releases what was used for sub-
6584 expansions (unless the block got too big). However, discard this block if
6585 the first thing is a variable or header, so that we can use its block when
6586 it is dynamic (useful for very large $message_headers, for example).
6588 6. Lookups now cache *every* query, not just the most recent. A new, separate
6589 store pool is used for this. It can be recovered when all lookup caches are
6590 flushed. Lookups now release memory at the end of their result strings.
6591 This has involved some general refactoring of the lookup sources.
6593 7. Some code has been added to the store_xxx() functions to reduce the amount
6594 of flapping under certain conditions.
6596 8. log_incoming_interface used to affect only the <= reception log lines. Now
6597 it causes the local interface and port to be added to several more SMTP log
6598 lines, for example "SMTP connection from", and rejection lines.
6600 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6602 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6604 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6605 could be overwritten at the end of the current message (or the start of a
6606 new message if it was set in a HELO ACL). The value is now preserved for
6607 the duration of the SMTP connection.
6609 12. If a transport had a headers_rewrite setting, and a matching header line
6610 contained an unqualified address, that address was qualified, even if it
6611 did not match any rewriting rules. The underlying bug was that the values
6612 of the flags that permit the existence of unqualified sender and recipient
6613 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6614 non-local messages, and by -bnq for local messages) were not being
6615 preserved with the message after it was received.
6617 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6618 "next input=" as part of the text comprising the host identity instead of
6619 the correct text. The code was using the same buffer for two different
6620 strings. However, depending on which order the printing function evaluated
6621 its arguments, the bug did not always show up. Under Linux, for example, my
6622 test suite worked just fine.
6624 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6625 doesn't work with some older versions of Perl. It has been changed to "my",
6626 which in any case is probably the better facility to use.
6628 15. A really picky compiler found some instances of statements for creating
6629 error messages that either had too many or two few arguments for the format
6632 16. The size of the buffer for calls to the DNS resolver has been increased
6633 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6634 for addresses that have a lot of PTR records. This alleviates a problem; it
6635 does not fully solve it.
6637 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6638 buffer now truncates the list and logs the incident, which is the same
6639 action as happens when Exim is looking up a host name and its aliases.
6640 Previously in this situation something unpredictable would happen;
6641 sometimes it was "internal error: store_reset failed".
6643 18. If a server dropped the connection unexpectedly when an Exim client was
6644 using GnuTLS and trying to read a response, the client delivery process
6645 crashed while trying to generate an error log message.
6647 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6648 string, for example:
6650 warn message = H1: something\nH2: something
6652 the text was added as a single header line from Exim's point of view
6653 though it ended up OK in the delivered message. However, searching for the
6654 second and subsequent header lines using $h_h2: did not work. This has been
6655 fixed. Similarly, if a system filter added multiple headers in this way,
6656 the routers could not see them.
6658 20. Expanded the error message when iplsearch is called with an invalid key to
6659 suggest using net-iplsearch in a host list.
6661 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6662 ACLs are no longer actually imposed (and a message to that effect is
6665 22. If a "gecos" field in a passwd entry contained escaped characters, in
6666 particular, if it contained a \" sequence, Exim got it wrong when building
6667 a From: or a Sender: header from that name. A second bug also caused
6668 incorrect handling when an unquoted " was present following a character
6669 that needed quoting.
6671 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6672 was not being matched caselessly.
6674 24. Arranged for all hyphens in the exim.8 source to be escaped with
6677 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6678 properly. Recipient callout cache records were still being keyed to include
6679 the sender, even when use_sender was set false. This led to far more
6680 callouts that were necessary. The sender is no longer included in the key
6681 when use_sender is false.
6683 26. Added "control = submission" modifier to ACLs.
6685 27. Added the ${base62d: operator to decode base 62 numbers.
6687 28. dnsdb lookups can now access SRV records.
6689 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6690 the configuration file.
6692 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6693 (-v) mode. This makes the output for a verbose queue run more intelligible.
6695 31. Added a use_postmaster feature to recipient callouts.
6697 32. Added the $body_zerocount variable, containing the number of binary zero
6698 bytes in the message body.
6700 33. The time of last modification of the "new" subdirectory is now used as the
6701 "mailbox time last read" when there is a quota error for a maildir
6704 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6706 35. Added +ignore_unknown as a special item in host lists.
6708 36. Code for decoding IPv6 addresses in host lists is now included, even if
6709 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6710 address was recognized as an IP address, but was then not correctly decoded
6711 into binary, causing unexpected and incorrect effects when compared with
6718 1. Very minor rewording of debugging text in manualroute to say "list of
6719 hosts" instead of "hostlist".
6721 2. If verify=header_syntax was set, and a header line with an unqualified
6722 address (no domain) and a large number of spaces between the end of the
6723 name and the colon was received, the reception process suffered a buffer
6724 overflow, and (when I tested it) crashed. This was caused by some obsolete
6725 code that should have been removed. The fix is to remove it!
6727 3. When running in the test harness, delay a bit after writing a bounce
6728 message to get a bit more predictability in the log output.
6730 4. Added a call to search_tidyup() just before forking a reception process. In
6731 theory, someone could use a lookup in the expansion of smtp_accept_max_
6732 per_host which, without the tidyup, could leave open a database connection.
6734 5. Added the variables $recipient_data and $sender_data which get set from a
6735 lookup success in an ACL "recipients" or "senders" condition, or a router
6736 "senders" option, similar to $domain_data and $local_part_data.
6738 6. Moved the writing of debug_print from before to after the "senders" test
6741 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6742 problems for message scanning, either using a data ACL, or using
6743 local_scan() because the Received: header was not generated till after they
6744 were called (in order to set the time as the time of reception completion).
6745 I have revised the way this works. The header is now generated after the
6746 body is received, but before the ACL or local_scan() are called. After they
6747 are run, the timestamp in the header is updated.
6753 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6754 before starting a queue runner without re-exec. This happened only when
6755 deliver_drop_privilege was set or when the Exim user was set to root. The
6756 effect of the bug was that timeouts during subsequent deliveries caused
6757 crashes instead of being properly handled. The handler is now left at its
6758 default (and expected) setting.
6760 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6761 message, again when deliver_drop_privilege is set or Exim is run as root.
6762 The bug described in (1) was not present in this case, but the tidying up
6763 of the other signals was missing. I have made the two cases consistent.
6765 3. The ignore_target_hosts setting on a manualroute router was being ignored
6766 for hosts that were looked up using the /MX notation.
6768 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6771 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6772 operated on the sender address. After changing the $sender_address to <>
6773 for the sender address verify, Exim was re-instated it as the original
6774 (before rewriting) address, but remembering that it had rewritten it, so it
6775 wasn't rewriting it again. This bug also had the effect of breaking the
6776 sender address verification caching when the sender address was rewritten.
6778 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6779 This has been changed so that if the ip literal address matches
6780 ignore_target_hosts, the router declines.
6782 7. Added expansion conditions match_domain, match_address, and match_local_
6783 part (NOT match_host).
6785 8. The placeholder for the Received: header didn't have a length field set.
6787 9. Added code to Exim itself and to exim_lock to test for a specific race
6788 condition that could lead to file corruption when using MBX delivery. The
6789 issue is with the lockfile that is created in /tmp. If this file is removed
6790 after a process has opened it but before that process has acquired a lock,
6791 there is the potential for a second process to recreate the file and also
6792 acquire a lock. This could lead to two Exim processes writing to the file
6793 at the same time. The added code performs the same test as UW imapd; it
6794 checks after acquiring the lock that its file descriptor still refers to
6795 the same named file.
6797 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6798 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6801 11. Added the smtp_active_hostname option. If used, this will typically be made
6802 to depend on the incoming interface address. Because $interface_address is
6803 not set up until the daemon has forked a reception process, error responses
6804 that can happen earlier (such as "too many connections") no longer contain
6807 12. If an expansion in a condition on a "warn" statement fails because a lookup
6808 defers, the "warn" statement is abandoned, and the next ACL statement is
6809 processed. Previously this caused the whole ACL to be aborted.
6811 13. Added the iplsearch lookup type.
6813 14. Added ident_timeout as a log selector.
6815 15. Added tls_certificate_verified as a log selector.
6817 16. Added a global option tls_require_ciphers (compare the smtp transport
6818 option of the same name). This controls incoming TLS connections.
6820 17. I finally figured out how to make tls_require_ciphers do a similar thing
6821 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6822 before starting the TLS session.
6824 18. Tabs are now shown as \t in -bP output.
6826 19. If the log selector return_path_on_delivery was set, Exim crashed when
6827 bouncing a message because it had too many Received: header lines.
6829 20. If two routers both had headers_remove settings, and the first one included
6830 a superfluous trailing colon, the final name in the first list and the
6831 first name in the second list were incorrectly joined into one item (with a
6832 colon in the middle).
6838 1. Added -C and -D options to the exinext utility, mainly to make it easier
6839 to include in the automated testing, but these could be helpful when
6840 multiple configurations are in use.
6842 2. The exinext utility was not formatting the output nicely when there was
6843 an alternate port involved in the retry record key, nor when there was a
6844 message id as well (for retries that were specific to a specific message
6845 and a specific host). It was also confused by IPv6 addresses, because of
6846 the additional colons they contain. I have fixed the IPv4 problem, and
6847 patched it up to do a reasonable job for IPv6.
6849 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6850 delivery, the log line now contains "pipelined" if PIPELINING was used.
6852 4. An SMTP transport process used to panic and die if the bind() call to set
6853 an explicit outgoing interface failed. This has been changed; it is now
6854 treated in the same way as a connect() failure.
6856 5. A reference to $sender_host_name in the part of a conditional expansion
6857 that was being skipped was still causing a DNS lookup. This no longer
6860 6. The def: expansion condition was not recognizing references to header lines
6861 that used bh_ and bheader_.
6863 7. Added the _cache feature to named lists.
6865 8. The code for checking quota_filecount in the appendfile transport was
6866 allowing one more file than it should have been.
6868 9. For compatibility with Sendmail, the command line option
6876 and sets the incoming protocol and host name (for trusted callers). The
6877 host name and its colon can be omitted when only the protocol is to be set.
6878 Note the Exim already has two private options, -pd and -ps, that refer to
6879 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6880 "s", but I don't think that's a major issue.
6882 10. A number of refactoring changes to the code, none of which should affect
6885 (a) The number of logging options was getting close to filling up the
6886 32-bit word that was used as a bit map. I have split them into two classes:
6887 those that are passed in the argument to log_write(), and those that are
6888 only ever tested independently outside of that function. These are now in
6889 separate 32-bit words, so there is plenty of room for expansion again.
6890 There is no change in the user interface or the logging behaviour.
6892 (b) When building, for example, log lines, the code previously used a
6893 macro that called string_cat() twice, in order to add two strings. This is
6894 not really sufficiently general. Furthermore, there was one instance where
6895 it was actually wrong because one of the argument was used twice, and in
6896 one call a function was used. (As it happened, calling the function twice
6897 did not affect the overall behaviour.) The macro has been replaced by a
6898 function that can join an arbitrary number of extra strings onto a growing
6901 (c) The code for expansion conditions now uses a table and a binary chop
6902 instead of a serial search (which was left over from when there were very
6903 few conditions). Also, it now recognizes conditions like "pam" even when
6904 the relevant support is not compiled in: a suitably worded error message is
6905 given if an attempt is made to use such a condition.
6907 11. Added ${time_interval:xxxxx}.
6909 12. A bug was causing one of the ddress fields not to be passed back correctly
6910 from remote delivery subprocesses. The field in question was not being
6911 subsequently used, so this caused to problems in practice.
6913 13. Added new log selectors queue_time and deliver_time.
6915 14. Might have fixed a bug in maildirsizefile handling that threw up
6916 "unexpected character" debug warnings, and recalculated the data
6917 unnecessarily. In any case, I expanded the warning message to give more
6920 15. Added the message "Restricted characters in address" to the statements in
6921 the default ACL that block characters like @ and % in local parts.
6923 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6924 Three changes have been made:
6926 (a) There was a serious bug; a negative response to MAIL caused the whole
6927 recipient domain to be cached as invalid, thereby blocking all messages
6928 to all local parts at the same domain, from all senders. This bug has
6929 been fixed. The domain is no longer cached after a negative response to
6930 MAIL if the sender used is not empty.
6932 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6935 (c) A new callout option, "use_sender" has been added for people who want
6936 the modified behaviour.
6942 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6945 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6946 indeed breaks things for older releases.
6948 3. Added additional logging to the case where there is a problem reading data
6949 from a filter that is running in a subprocess using a pipe, in order to
6950 try to track down a specific problem.
6952 4. Testing facility fudge: when running in the test harness and attempting
6953 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6954 getting "No route to host". Convert this to a timeout.
6956 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6959 6. Some OS don't have socklen_t but use size_t instead. This affects the
6960 fifth argument of getsockopt() amongst other things. This is now
6961 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6962 can be set for individual OS. I have set it for SunOS5, OSF1, and
6963 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6964 some earlier ones do not.
6966 7. Change 4.30/15 was not doing the test caselessly.
6968 8. The standard form for an IPv6 address literal was being rejected by address
6969 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6970 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6971 this, as well as the form without the "IPv6" on the front (but only when
6972 address literals are enabled, of course).
6974 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6976 10. Exim crashed if a message with an empty sender address specified by -f
6977 encountered a router with an errors_to setting. This could be provoked only
6978 by a command such as
6982 where an empty string was supplied; "<>" did not hit this bug.
6984 11. Installed PCRE release 4.5.
6986 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6987 remained set. It is now erased.
6989 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6990 times from message ids (which are base 36 rather than the normal 62).
6992 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6993 were being counted as actual protocol errors, and logged if the log
6994 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6995 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6996 and DATA following a set of rejected RCPTs do not count as protocol errors.
6997 In other words, Exim assumes they were pipelined, though this may not
6998 actually be the case. Of course, in all cases the client gets an
6999 appropriate error code.
7001 15. If a lookup fails in an ACL condition, a message about the failure may
7002 be available; it is used if testing the ACL cannot continue, because most
7003 such messages specify what the cause of the deferral is. However, some
7004 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7005 that caused an old message to be retained and used if a later statement
7006 caused a defer, replacing the real cause of the deferral.
7008 16. If an IP address had so many PTR records that the DNS lookup buffer
7009 was not large enough to hold them, Exim could crash while trying to process
7010 the truncated data. It now detects and logs this case.
7012 17. Further to 4.21/58, another change has been made: if (and only if) the
7013 first line of a message (the first header line) ends with CRLF, a bare LF
7014 in a subsequent header line has a space inserted after it, so as not to
7015 terminate the header.
7017 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7018 unnecessarily, used atoi() instead of strtol(), and didn't check the
7019 termination when getting file sizes from file names by regex.
7021 19. Completely re-implemented the support for maildirsize files, in the light
7022 of a number of problems with the previous contributed implementation
7023 (4.30/29). In particular:
7025 . If the quota is zero, the maildirsize file is maintained, but no quota is
7028 . If the maildir directory does not exist, it is created before any attempt
7029 to write a maildirsize file.
7031 . The quota value in the file is just a cache; if the quota is changed in
7032 the transport, the new value overrides.
7034 . A regular expression is available for excluding directories from the
7037 20. The autoreply transport checks the characters in options that define the
7038 message's headers; it allows continued headers, but it was checking with
7039 isspace() after an embedded newline instead of explicitly looking for a
7042 21. If all the "regular" hosts to which an address was routed had passed their
7043 expiry times, and had not reached their retry times, the address was
7044 bounced, even if fallback hosts were defined. Now Exim should go on to try
7047 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7048 equivalent code in the SMTP transport. Some hosts send humungous responses
7049 to HELO/EHLO, more than 1024 it seems.
7051 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7052 gives compiler warnings in some environments. I've now done it "properly",
7055 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7056 (because of problems with the built-in one) was declared to return uschar *
7057 instead of char *, causing compiler failure.
7059 25. Fixed a file descriptor leak when processing alias/forward files.
7061 26. Fixed a minor format string issue in dbfn.c.
7063 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7065 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7066 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7069 29. When a sender address is verified, it is cached, to save repeating the test
7070 when there is more than one recipient in a message. However, when the
7071 verification involves a callout, it is possible for different callout
7072 options to be set for different recipients. It is too complicated to keep
7073 track of this in the cache, so now Exim always runs a verification when a
7074 callout is required, relying on the callout cache for the optimization.
7075 The overhead is duplication of the address routing, but this should not be
7078 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7079 to be verified with callout=postmaster, and the main callout worked but the
7080 postmaster check failed, the verification correctly failed. However, if a
7081 subsequent RCPT command asked for sender verification *without* the
7082 postmaster check, incorrect caching caused this verification also to fail,
7085 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7086 it was not caching the DNS options (qualify_single, search_parents) that
7087 were used when the lookup failed. A subsequent lookup with different
7088 options therefore always gave the same answer, though there were cases
7089 where it should not have. (Example: a "domains = !$mx_any" option on a
7090 dnslookup router: the "domains" option is always processed without any
7091 widening, but the router might have qualify_single set.) Now Exim uses the
7092 cached value only when the same options are set.
7094 32. Added John Jetmore's "exipick" utility to the distribution.
7096 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7097 than a timeout (e.g. a certificate is required, and is not provided), an
7098 Exim server now closes the connection immediately. Previously it waited for
7099 the client to close - but if the client is SSL, it seems that they each
7100 wait for each other, leading to a delay before one of them times out.
7102 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7103 maintained 0.8.x compatibility because I don't think many are using it, and
7104 it is clearly obsolete.
7106 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7109 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7110 client certificate was expired. A simple patch fixes this, though I don't
7111 understand the full logic of why the verify callback is called multiple
7114 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7115 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7116 which causes problems with some clients (such as the Certicom SSL Plus
7117 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7118 disables the coutermeasure allowing Eudora to connect."
7120 38. Exim was not checking that a write() to a log file succeeded. This could
7121 lead to Bad Things if a log got too big, in particular if it hit a file
7122 size limit. Exim now panics and dies if it cannot write to a log file, just
7123 as it does if it cannot open a log file.
7125 39. Modified OS/Makefile-Linux so that it now contains
7127 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7129 The two -D definitions ensure that Exim is compiled with large file
7130 support, which makes it possible to handle log files that are bigger than
7133 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7134 instance) a domain was checked against a named list that involved a lookup,
7135 causing $domain_data to be set, then another domain was checked against the
7136 same list, then the first domain was re-checked, the value of $domain_data
7137 after the final check could be wrong. In particular, if the second check
7138 failed, it could be set empty. This bug probably also applied to
7141 41. The strip_trailing_dot option was not being applied to the address given
7142 with the -f command-line option.
7144 42. The code for reading a message's header from the spool was incrementing
7145 $received_count, but never initializing it. This meant that the value was
7146 incorrect (doubled) while delivering a message in the same process in which
7147 it was received. In the most common configuration of Exim, this never
7148 happens - a fresh exec is done - but it can happen when
7149 deliver_drop_privilege is set.
7151 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7152 it now includes up to 150 characters of the unexpected data in the log
7155 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7156 and building data strings. The size of both of these buffers was 10 000
7157 bytes - far larger than anybody would *ever* want, thought I. Needless to
7158 say, somebody hit the limit. I have increased the maximum line length to
7159 20 000 and the maximum data length of concatenated lines to 100 000. I have
7160 also fixed two bugs, because there was no checking on these buffers. Tsk,
7161 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7162 buffer is too small.
7164 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7165 lsearch lookups. Now it does.
7167 46. When parsing a route_list item in a manualroute router, a fixed-length
7168 buffer was used for the list of hosts. I made this 1024 bytes long,
7169 thinking that nobody would ever have a list of hosts that long. Wrong.
7170 Somebody had a whole pile of complicated expansion conditions, and the
7171 string was silently truncated, leading to an expansion error. It turns out
7172 that it is easier to change to an unlimited length (owing to other changes
7173 that have happened since this code was originally written) than to build
7174 structure for giving a limitation error. The length of the item that
7175 expands into the list of hosts is now unlimited.
7177 47. The lsearch lookup could not handle data where the length of text line was
7178 more than 4095 characters. Such lines were truncated, leading to shortened
7179 data being returned. It should now handle lines of any length.
7181 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7182 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7185 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7186 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7187 no longer happens in comment lines. A list of replacements is now placed
7188 at the head of all of the source files, except those whose only change is
7189 to replace PERL_COMMAND in the very first #! line.
7191 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7192 messages on the queue, with a bottom-up merge sort, using code contributed
7193 by Michael Haardt. This should make operations like -bp somewhat faster on
7194 large queues. It won't affect queue runners, except when queue_run_in_order
7197 51. Installed eximstats 1.31 in the distribution.
7199 52. Added support for SRV lookups to the dnslookup router.
7201 53. If an ACL referred to $message_body or $message_body_end, the value was not
7202 reset for any messages that followed in the same SMTP session.
7204 54. The store-handling optimization for building very long strings was not
7205 differentiating between the different store pools. I don't think this
7206 actually made any difference in practice, but I've tidied it.
7208 55. While running the routers to verify a sender address, $sender_address
7209 was still set to the sender address. This is wrong, because when routing to
7210 send a bounce to the sender, it would be empty. Therefore, I have changed
7211 it so that, while verifying a sender address, $sender_address is set to <>.
7212 (There is no change to what happens when verifying a recipient address.)
7214 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7215 target A or AAAA records (if not already returned) without resetting the
7216 qualify_single or search_parents options of the DNS resolver. These are
7217 inappropriate in this case because the targets of MX and SRV records must
7218 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7219 target that, when qualified, matched something in the local domain. These
7220 two options are now turned off when doing these lookups.
7222 57. It seems that at least some releases of Reiserfs (which does not have the
7223 concept of a fixed number of inodes) returns zero and not -1 for the
7224 number of available inodes. This interacted badly with check_spool_inodes,
7225 which assumed that -1 was the "no such thing" setting. What I have done is
7226 to check that the total number of inodes is greater than zero before doing
7227 the test of how many are available.
7229 58. When a "warn" ACL statement has a log_message modifier, the message is
7230 remembered, and not repeated. This is to avoid a lot of repetition when a
7231 message has many recipients that cause the same warning to be written.
7232 However, Exim was preserving the list of already written lines for an
7233 entire SMTP session, which doesn't seem right. The memory is now reset if a
7234 new message is started.
7236 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7237 debugging output unless log_rewrite was also set.
7239 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7240 of (int)(handle) when we know that handle contains (void *)(-1).
7242 61. The Exim daemon panic-logs an error return when it closes the incoming
7243 connection. However "connection reset by peer" seems to be common, and
7244 isn't really an error worthy of noting specially, so that particular error
7247 62. When Exim is trying to find all the local interfaces, it used to panic and
7248 die if the ioctl to get the interface flags failed. However, it seems that
7249 on at least one OS (Solaris 9) it is possible to have an interface that is
7250 included in the list of interfaces, but for which you get a failure error
7251 for this call. This happens when the interface is not "plumbed" into a
7252 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7253 failure of the "get flags" call assumes that the interface is down.
7255 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7256 makes life easier for people who are doing arithmetic on fields extracted
7257 from dates, where you often get leading zeros that should not be
7258 interpreted as octal.
7260 64. Added qualify_domain to the redirect router, to override the global
7263 65. If a pathologically long header line contained very many addresses (the
7264 report of this problem mentioned 10 000) and each of them was rewritten,
7265 Exim could use up a very large amount of memory. (It kept on making new
7266 copies of the header line as it rewrote, and never released the old ones.)
7267 At the expense of a bit more processing, the header rewriting function has
7268 been changed so that it no longer eats memory in this way.
7270 66. The generation of the Received: header has been moved from the time that a
7271 message starts to be received, to the time that it finishes. The timestamp
7272 in the Received: header should now be very close to that of the <= log
7273 line. There are two side-effects of this change:
7275 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7276 logged header lines no longer include the local Received: line, because
7277 it has not yet been created. The same applies to a copy of the message
7278 that is returned to a non-SMTP sender when a message is rejected.
7280 (b) When a filter file is tested using -bf, no additional Received: header
7281 is added to the test message. After some thought, I decided that this
7284 This change does not affect the value of $received_for. It is still set
7285 after address rewriting, but before local_scan() is called.
7287 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7289 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7290 gave an unhelpful panic error message, and a defer error. I have managed to
7291 change this behaviour so that it now rejects any supplied certificate,
7292 which seems right, as the list of acceptable certificates is empty.
7294 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7295 gave an unhelpful defer error. I have not managed to make this reject any
7296 supplied certificates, but the error message it gives is "no certificate
7297 supplied", which is not helpful.
7299 70. exigrep's output now also includes lines that are not associated with any
7300 message, but which match the given pattern. Implemented by a patch from
7301 Martin Sluka, which also tidied up the Perl a bit.
7303 71. Recipient callout verification, like sender verification, was using <> in
7304 the MAIL FROM command. This isn't really the right thing, since the actual
7305 sender may affect whether the remote host accepts the recipient or not. I
7306 have changed it to use the actual sender in the callout; this means that
7307 the cache record is now keyed on a recipient/sender pair, not just the
7308 recipient address. There doesn't seem to be a real danger of callout loops,
7309 since a callout by the remote host to check the sender would use <>.
7310 [SEE ABOVE: changed after hitting problems.]
7312 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7313 temporary errors. However, in the case of such a code being given after
7314 the end of a data transmission (i.e. after ".") Exim was failing to write
7315 a retry record for the message. (Yes, there was some broken host that was
7316 actually sending 8xx at this point.)
7318 73. An unknown lookup type in a host list could cause Exim to panic-die when
7319 the list was checked. (An example that provoked this was putting <; in the
7320 middle of a list instead of at the start.) If this happened during a DATA
7321 ACL check, a -D file could be left lying around. This kind of configuration
7322 error no longer causes Exim to die; instead it causes a defer error. The
7323 incident is still logged to the main and panic logs.
7325 74. Buglet left over from Exim 3 conversion. The message "too many messages
7326 in one connection" was written to the rejectlog but not the mainlog, except
7327 when address rewriting (yes!) was being logged.
7329 75. Added write_rejectlog option.
7331 76. When a system filter was run not as root (that is, when system_filter_user
7332 was set), the values of the $n variables were not being returned to the
7333 main process; thus, they were not subsequently available in the $sn
7336 77. Added +return_path_on_delivery log selector.
7338 78. A connection timeout was being treated differently from recipients deferred
7339 when testing hosts_max_try with a message that was older than the host's
7340 retry timeout. (The host should not be counted, thus allowing all hosts to
7341 be tried at least once before bouncing.) This may have been the cause of an
7342 occasionally reported bug whereby a message would remain on the queue
7343 longer than the retry timeout, but would be bounced if a delivery was
7344 forced. I say "may" because I never totally pinned down the problem;
7345 setting up timeout/retry tests is difficult. See also the next item.
7347 79. The ultimate address timeout was not being applied to errors that involved
7348 a combination of host plus message (for example, a timeout on a MAIL
7349 command). When an address resolved to a number of possible hosts, and they
7350 were not all tried for each delivery (e.g. because of hosts_max_try), a
7351 message could remain on the queue longer than the retry timeout.
7353 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7356 81. Fixed an obscure SMTP outgoing bug which required at least the following
7357 conditions: (a) there was another message waiting for the same server;
7358 (b) the server returned 5xx to all RCPT commands in the first message so
7359 that the message was not completed; (c) the server dropped the connection
7360 or gave a negative response to the RSET that Exim sends to abort the
7361 transaction. The observed case was a dropped connection after DATA that had
7362 been sent in pipelining mode. That is, the server had advertised PIPELINING
7363 but was not implementing it correctly. The effect of the bug was incorrect
7364 behaviour, such as trying another host, and this could lead to a crash.
7370 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7371 and daemon.c were passed as pointers to ints; they should have been
7372 pointers to socklen_t variables (which are typically unsigned ints).
7374 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7377 3. Fixed a really odd bug that affected only the testing scheme; patching a
7378 certain fixed string in the binary changed the value of another string that
7379 happened to be identical to the end of the original first string.
7381 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7382 name", it returns that address as the IP address. On some operating
7383 systems (e.g. Solaris), it also passes back the IP address string as the
7384 "host name". However, on others (e.g. Linux), it passes back an empty
7385 string. Exim wasn't checking for this, and was changing the host name to an
7386 empty string, assuming it had been canonicalized.
7388 5. Although rare, it is permitted to have more than one PTR record for a given
7389 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7390 all the names associated with an address, because they do in Solaris.
7391 However, it seems that they do not in Linux for data that comes from the
7392 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7393 I found this out when I moved to a new Linux workstation and tried to run
7394 the Exim test suite.
7396 To get round this problem I have changed the code so that it now does its
7397 own call to the DNS to look up PTR records when searching for a host name.
7398 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7399 addresses that are only in /etc/hosts are still found.
7401 This behaviour is, however, controlled by an option called host_lookup_
7402 order, which defaults to "bydns:byaddr". If people want to use the other
7403 order, or indeed, just use one or the other means of lookup, they can
7404 specify it in this variable.
7406 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7407 some operating systems, this comes back from gethostbyaddr() as an empty
7408 string, and this is what Exim used to test for. However, it seems that in
7409 other systems, "." is yielded. Exim now tests for this case too.
7411 7. The values of check_spool_space and check_log_space are now held internally
7412 as a number of kilobytes instead of an absolute number of bytes. If a
7413 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7414 kilobyte. This means that much larger values can be stored.
7416 8. Exim monitor: an attempt to get the action menu when not actually pointing
7417 at a message produces an empty menu entitled "No message selected". This
7418 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7419 no entries in it ("Shell widget menu has zero width and/or height"). So I
7420 have added a single, blank menu entry in this case.
7422 9. Added ${quote_local_part.
7424 10. MIME decoding is now applied to the contents of Subject: header lines when
7427 11. Now that a reference to $sender_host_address automatically causes a reverse
7428 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7429 host lookup before query-style lookups in lists that might use this
7430 variable. This has therefore been abolished, and the "net-" prefix is no
7431 longer necessary for query-style lookups.
7433 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7434 have been a typo for LFLAGS, so it has been changed.
7436 13. The install script calls Exim with "-C /dev/null" in order to find the
7437 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7438 to be output. However, since Exim outputs its version number before the
7439 error, it didn't break the script. It just looked ugly. I fixed this by
7440 always allowing "-C /dev/null" if the caller is root.
7442 14. Ignore overlarge ACL variable number when reading spool file - insurance
7443 against a later release with more variables having written the file.
7445 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7446 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7447 this, as well as the form without the "IPv6" on the front.
7449 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7450 OS/Makefile-Darwin file.
7452 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7453 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7455 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7456 variable. However, some OS use ints for inodes. I've added cast to long int
7457 to get rid of the compiler warning.
7459 19. I had forgotten to lock out "/../" in configuration file names when
7460 ALT_CONFIG_PREFIX was set.
7462 20. Routers used for verification do not need to specify transports. However,
7463 if such a router generated a host list, and callout was configured, Exim
7464 crashed, because it could not find a port number from the (non-existent)
7465 transport. It now assumes port 25 in this circumstance.
7467 21. Added the -t option to exigrep.
7469 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7470 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7471 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7472 them is set, LOOKUP_LSEARCH is forced.
7474 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7476 24. Added sender and host information to the "rejected by local_scan()" log
7477 line; previously there was no indication of these.
7479 25. Added .include_if_exists.
7481 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7482 Linux. It turns out that not all file systems support this. Apparently some
7483 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7484 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7485 not supported on the file descriptor, is now ignored when Exim is trying to
7486 sync a directory. This applies only to Linux.
7488 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7490 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7491 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7494 29. Added support for maildirsize files from supplied patch (modified a bit).
7496 30. The use of :fail: followed by an empty string could lead Exim to respond to
7497 sender verification failures with (e.g.):
7499 550 Verification failed for <xxx>
7500 550 Sender verify failed
7502 where the first response line was missing the '-' that indicates it is not
7503 the final line of the response.
7505 31. The loop for finding the name of the user that called Exim had a hardwired
7506 limit of 10; it now uses the value of finduser_retries, which is used for
7507 all other user lookups.
7509 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7512 33. Exim was neglecting to zero errno before one call of strtol() when
7513 expanding a string and expecting an integer value. On some systems this
7514 resulted in spurious "integer overflow" errors. Also, it was casting the
7515 result into an int without checking.
7517 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7518 did not work. The code looks as if it has *never* worked, though it appears
7519 to have been documented since at least release 1.62. I have made it work.
7521 35. The "timeout_DNS" error in retry rules, also documented since at least
7522 1.62, also never worked. As it isn't clear exactly what this means, and
7523 clearly it isn't a major issue, I have abolished the feature by treating it
7524 as "timeout", and writing a warning to the main and panic logs.
7526 36. The display of retry rules for -brt wasn't always showing the error code
7529 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7530 timeout_connect_A, timeout_connect_MX.
7532 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7533 to the empty sender.
7535 39. The daemon was not analysing the content of -oX till after it had closed
7536 stderr and disconnected from the controlling terminal. This meant that any
7537 syntax errors were only noted on the panic log, and the return code from
7538 the command was 0. By re-arranging the code a little, I've made the
7539 decoding happen first, so such errors now appear on stderr, and the return
7540 code is 1. However, the actual setting up of the sockets still happens in
7541 the disconnected process, so errors there are still only recorded on the
7544 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7545 connections (as happens on some IP stacks) was logged at start up time as
7546 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7547 it from "IPv6 and IPv4", which means that two separate sockets are being
7550 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7551 says whether AF_INET or AF_INET6 was passed as an argument.
7553 42. Exiwhat output was messed up when time zones were included in log
7556 43. Exiwhat now gives more information about the daemon's listening ports,
7557 and whether -tls-on-connect was used.
7559 44. The "port" option of the smtp transport is now expanded.
7561 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7562 silently ignored. Now an error message is written to the main and panic
7565 46. There's a new ACL modifier called "logwrite" which writes to a log file
7566 as soon as it is encountered.
7568 47. Added $local_user_uid and $local_user_gid at routing time.
7570 48. Exim crashed when trying to verify a sender address that was being
7573 49. Exim was recognizing only a space character after ".include". It now also
7574 recognizes a tab character.
7576 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7577 extracting the relevant information from the specification. The man page no
7578 longer contains scrambled data for the -d option, and I've added a section
7579 at the front about calling Exim under different names.
7581 51. Added "extra_headers" argument to the "mail" command in filter files.
7583 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7586 53. Installed eximstats 1.29.
7588 54. Added transport_filter_timeout as a generic transport option.
7590 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7591 Cc: header lines. This was required by RFC 822, but it not required by RFC
7594 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7595 incoming messages that did not have them. Now it does so only if the
7596 message originates locally, that is, if there is no associated remote host
7597 address. When Resent- header lines are present, this applies to the Resent-
7598 lines rather than the non-Resent- lines.
7600 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7601 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7603 58. Messages for configuration errors now include the name of the main
7604 configuration file - useful now that there may be more than one file in a
7605 list (.included file names were always shown).
7607 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7608 for those rare installations that do not start the daemon as root or run it
7609 setuid root. I've cut out the call to initgroups() if the daemon is not
7612 60. The Exim user and group can now be bound into the binary as text strings
7613 that are looked up at the start of Exim's processing.
7615 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7617 62. Added $mailstore_basename variable.
7619 63. Installed patch to sieve.c from Michael Haardt.
7621 64. When Exim failed to open the panic log after failing to open the main log,
7622 the original message it was trying to log was written to stderr and debug
7623 output, but if they were not available (the usual case in production), it
7624 was lost. Now it is written to syslog before the two lines that record the
7625 failures to open the logs.
7627 65. Users' Exim filters run in subprocesses under the user's uid. It is
7628 possible for a "deliver" command or an alias in a "personal" command to
7629 provoke an address rewrite. If logging of address rewriting is configured,
7630 this fails because the process is not running as root or exim. There may be
7631 a better way of dealing with this, but for the moment (because 4.30 needs
7632 to be released), I have disabled address rewrite logging when running a
7633 filter in a non-root, non-exim process.
7639 1. The buildconfig auxiliary program wasn't quoting the value set for
7640 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7641 not defined. This bug was masked in 4.22 by the effect that was fixed in
7644 2. Some messages that were rejected after a message id was allocated were
7645 shown as "incomplete" by exigrep. It no longer does this for messages that
7646 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7648 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7649 have allow_domain_literals set, the ID did not get logged in the <= line.
7650 Domain literals are now always recognized in Message-ID: header lines.
7652 4. The first argument for a ${extract expansion item is the key name or field
7653 number. Leading and trailing spaces in this item were not being ignored,
7654 causing some misleading effects.
7656 5. When deliver_drop_privilege was set, single queue runner processes started
7657 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7658 same command in the process it spins off) were not dropping privilege.
7660 6. When the daemon running as "exim" started a queue runner, it always
7661 re-executed Exim in the spun-off process. This is a waste of effort when
7662 deliver_drop_privilege is set. The new process now just calls the
7663 queue-runner function directly.
7669 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7672 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7673 ignored. Though the use of -oP was forcing the writing of a pid file, it
7674 was always written to the default place.
7676 3. If the message "no IP address found for host xxxx" is generated during
7677 incoming verification, it is now followed by identification of the incoming
7678 connection (so you can more easily find what provoked it).
7680 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7682 5. Added some features to "harden" Exim a bit more against certain attacks:
7684 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7685 be put in Local/Makefile. This is like the never_users runtime option,
7686 but it cannot be overridden. The default setting is "root".
7688 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7689 prefix string with which any file named in a -C command line option
7692 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7693 is retained for -C and -D only if the caller of Exim is root. Without
7694 it, the exim user may also use -C and -D and retain privilege.
7696 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7697 command line option is disabled.
7699 6. Macro names set by the -D option must start with an upper case letter, just
7700 like macro names defined in the configuration file.
7702 7. Added "dereference=" facility to LDAP.
7704 8. Two instances of the typo "uknown" in the source files are fixed.
7706 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7707 the Configure-Makefile script screwed up while processing it.
7709 10. Incorporated PCRE 4.4.
7711 11. The SMTP synchronization check was not operating right at the start of an
7712 SMTP session. For example, it could not catch a HELO sent before the client
7713 waited for the greeting. There is now a check for outstanding input at the
7714 point when the greeting is written. Because of the duplex, asynchronous
7715 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7716 way, but not yet received, when the check is performed.
7718 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7719 on TCP/IP sockets, because this apparently causes some broken clients to
7722 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7723 unchanged) from the Cygwin maintainer.
7725 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7726 was supported instead of testing for mbx. Effectively a typo.
7728 15. The spa authenticator server code was not checking that the input it
7729 received was valid base64.
7731 16. The debug output line for the "set" modifier in ACLs was not showing the
7732 name of the variable that was being set.
7734 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7736 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7737 lookup on its own. Something else was needed to trigger the lookup. For
7738 example, a match in host_lookup or the need for a host name in a host list.
7739 Now, if $sender_host_name is referenced and the host name has not yet been
7740 looked up, a lookup is performed. If the lookup fails, the variable remains
7741 empty, and $host_lookup_failed is set to "1".
7743 19. Added "eqi" as a case-independent comparison operator.
7745 20. The saslauthd authentication condition could segfault if neither service
7746 nor realm was specified.
7748 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7749 error message that was logged was misleading, and incoming SMTP
7750 connections were dropped. The message is now more accurate, and temporary
7751 errors are given to SMTP connections.
7753 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7754 (except RSET) with a 421 error, until QUIT is received. However, it was
7755 failing to send a response to QUIT.
7757 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7758 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7759 ACL is now run after the helo_try_verify_hosts code.
7761 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7762 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7763 case-independent, so other case variants are also recognized). Apparently
7764 some systems use these upper case variants.
7766 25. If more than two messages were waiting for the same host, and a transport
7767 filter was specified for the transport, Exim sent two messages over the
7768 same TCP/IP connection, and then failed with "socket operation on non-
7769 socket" when it tried to send the third.
7771 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7773 27. The extern definition of crypt16() in expand.c was not being excluded when
7774 the OS had its own crypt16() function.
7776 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7777 as a preferred synonym for return_size_limit, both as an option and as an
7780 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7782 30. Changed the default configuration ACL to relax the local part checking rule
7783 for addresses that are not in any local domains. For these addresses,
7784 slashes and pipe symbols are allowed within local parts, but the sequence
7785 /../ is explicitly forbidden.
7787 31. SPA server authentication was not clearing the challenge buffer before
7790 32. log_message in a "warn" ACL statement was writing to the reject log as
7791 well as to the main log, which contradicts the documentation and doesn't
7792 seem right (because no rejection is happening). So I have stopped it.
7794 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7795 However, I am unable to do any testing of this.
7797 34. Fixed an infelicity in the appendfile transport. When checking directories
7798 for a mailbox, to see if any needed to be created, it was accidentally
7799 using path names with one or more superfluous leading slashes; tracing
7800 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7802 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7803 contents are added to the log line that is written for every discarded
7804 recipient. (Previously a log_message setting was ignored.)
7806 36. The ${quote: operator now quotes the string if it is empty.
7808 37. The install script runs exim in order to find its version number. If for
7809 some reason other than non-existence or emptiness, which it checks, it
7810 could not run './exim', it was installing it with an empty version number,
7811 i.e. as "exim-". This error state is now caught, and the installation is
7814 38. An argument was missing from the function that creates an error message
7815 when Exim fails to connect to the socket for saslauthd authentication.
7816 This could cause Exim to crash, or give a corrupted message.
7818 39. Added isip, isip4, and isip6 to ${if conditions.
7820 40. The ACL variables $acl_xx are now saved with the message, and can be
7821 accessed later in routers, transports, and filters.
7823 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7824 strings in the file are not string-expanded.
7826 42. If a MAIL command specified a SIZE value that was too large to fit into an
7827 int variable, the check against message_size_limit failed. Such values are
7828 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7829 day this will have to be increased, but I don't think I want to be around
7830 when emails are that large.
7837 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7838 iconv() is not standard in FreeBSD.
7840 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7841 IPv6 enabled. The observed symptom was a segmentation fault on return from
7842 the function os_common_find_running_interfaces() in src/os.c.
7844 3. In the check_special_case() function in daemon.c I had used "errno" as an
7845 argument name, which causes warnings on some systems. This was basically a
7846 typo, since it was named "eno" in the comments!
7848 4. The code that waits for the clock to tick (at a resolution of some fraction
7849 of a second) so as to ensure message-id uniqueness was always waiting for
7850 at least one whole tick, when it could have waited for less. [This is
7851 almost certainly not relevant at current processor speeds, where it is
7852 unlikely to ever wait at all. But we try to future-proof.]
7854 5. The function that sleeps for a time interval that includes fractions of a
7855 second contained a race. It did not block SIGALRM between setting the
7856 timer, and suspending (a couple of lines later). If the interval was short
7857 and the sigsuspend() was delayed until after it had expired, the suspension
7858 never ended. On busy systems this could lead to processes getting stuck for
7861 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7862 process, before it forks any delivery processes. The open lookup caching
7863 mechanism meant that the open file or database connection was passed into
7864 the delivery process. The problem was that delivery processes always tidy
7865 up cached lookup data. This could cause a problem for the next delivery
7866 process started by the queue runner, because the external queue runner
7867 process does not know about the closure. So the next delivery process
7868 still has data in the lookup cache. In the case of a file lookup, there was
7869 no problem because closing a file descriptor in a subprocess doesn't affect
7870 the parent. However, if the lookup was caching a connection to a database,
7871 the connection was closed, and the second delivery process was likely to
7872 see errors such as "PGSQL: query failed: server closed the connection
7873 unexpectedly". The problem has been fixed by closing all cached lookups
7874 in a queue runner before running a delivery process.
7876 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7877 seem to have the "const" qualifier which it has on other OS. I've
7880 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7881 *of the same type* (client or server) with the same public name that an
7882 error should be diagnosed.
7884 9. When Exim looked up a host name for an IP address, but failed to find the
7885 original IP address when looking up the host name (a safety check), it
7886 output the message "<ip address> does not match any IP for NULL", which was
7887 confusing, to say the least. The bug was that the host name should have
7888 appeared instead of "NULL".
7890 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7891 user that is built into the binary, and the -C or -D options is used, root
7892 privilege is dropped before the configuration file is read. In addition,
7893 logging is switched to stderr instead of the normal log files. If the
7894 configuration then re-defines the Exim user, the unprivileged environment
7895 is probably not what is expected, so Exim logs a panic warning message (but
7898 However, if deliver_drop_privilege is set, the unprivileged state may well
7899 be exactly what is intended, so the warning has been cut out in that case,
7900 and Exim is allowed to try to write to its normal log files.
7906 1. smtp_return_error_details was not giving details for temporary sender
7907 or receiver verification errors.
7909 2. Diagnose a configuration error if two authenticators have the same public
7912 3. Exim used not to create the message log file for a message until the first
7913 delivery attempt. This could be confusing when incoming messages were held
7914 for policy or load reasons. The message log file is now created at the time
7915 the message is received, and an initial "Received" line is written to it.
7917 4. The automatically generated man page for command line options had a minor
7918 bug that caused no ill effects; however, a more serious problem was that
7919 the procedure for building the man page automatically didn't always
7920 operate. Consequently, release 4.20 contains an out-of-date version. This
7921 shouldn't happen again.
7923 5. When building Exim with embedded Perl support, the script that builds the
7924 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7925 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7927 6. The freeze_tell option was not being used for messages that were frozen on
7928 arrival, either by an ACL or by local_scan().
7930 7. Added the smtp_incomplete_transaction log selector.
7932 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7933 was accepting AUTH without a new EHLO.
7935 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7936 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7939 10. Logging of TCP/IP connections (when configured) now happens in the main
7940 daemon process instead of the child process, so that the TCP/IP connection
7941 count is more accurate (but it can never be perfect).
7943 11. The use of "drop" in a nested ACL was not being handled correctly in the
7944 outer ACL. Now, if condition failure induced by the nested "drop" causes
7945 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7946 or "require"), the connection is dropped.
7948 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7949 that yield "discard" can now be used with an "accept" or a "discard" verb,
7950 but an error is generated for any others (because I can't see a useful way
7951 to define what should happen).
7953 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7954 are now processed in the same way as lines in the Exim configuration file.
7955 In particular, continuation lines are supported.
7957 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7959 15. Added -ti meaning -t -i.
7961 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7962 domains, and warn by logging if others are found.
7964 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7965 returned from GIFCONF when Exim is trying to find the list of interfaces on
7966 a host. The code in os.c has been modified to copy each ifreq to an aligned
7967 structure in all cases.
7969 Also, in some cases, the returned ifreq's were being copied to a 'struct
7970 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7971 means the last couple of bytes of an IPv6 address could be chopped if the
7972 ifreq contained only a normal sockaddr (14 bytes storage).
7974 18. Named domain lists were not supported in the hosts_treat_as_local option.
7975 An entry such as +xxxx was not recognized, and was treated as a literal
7978 19. Ensure that header lines added by a DATA ACL are included in the reject log
7979 if the ACL subsequently rejects the message.
7981 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7982 MD5 (which is deprecated).
7984 21. When testing a filter file using -bf, Exim was writing a message when it
7985 took the sender from a "From " line in the message, but it was not doing so
7986 when it took $return_path from a Return-Path: header line. It now does.
7988 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7989 with a valid header line field name (a series of printing characters
7990 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7992 23. Changed "disc" in the source to "disk" to conform to the documentation and
7993 the book and for uniformity.
7995 24. Ignore Sendmail's -Ooption=value command line item.
7997 25. When execve() failed while trying to run a command in a pipe transport,
7998 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7999 could be confused with a return value of 69 from the command itself. This
8000 has been changed to 127, the value the shell returns if it is asked to run
8001 a non-existent command. The wording for the related log line suggests a
8002 non-existent command as the problem.
8004 26. If received_header_text expands to an empty string, do not add a Received:
8005 header line to the message. (Well, it adds a token one on the spool, but
8006 marks it "old" so that it doesn't get used or transmitted.)
8008 27. Installed eximstats 1.28 (addition of -nt option).
8010 28. There was no check for failure on the call to getsockname() in the daemon
8011 code. This can fail if there is a shortage of resources on the system, with
8012 ENOMEM, for example. A temporary error is now given on failure.
8014 29. Contrary to the C standard, it seems that in some environments, the
8015 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8016 program. Exim now does this explicitly; it affects the formatting of
8017 timestamps using strftime().
8019 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8020 complaints. I've now initialized all the variables, to avoid this.
8022 32. Header lines added by a system filter were not being "seen" during
8023 transport-time rewrites.
8025 33. The info_callback() function passed to OpenSSL is set up with type void
8026 (*)(SSL *, int, int), as described somewhere. However, when calling the
8027 function (actually a macro) that sets it up, the type void(*)() is
8028 expected. I've put in a cast to prevent warnings from picky compilers.
8030 34. If a DNS black list lookup found a CNAME record, but there were no A
8031 records associated with the domain it pointed at, Exim crashed.
8033 35. If a DNS black list lookup returned more than one A record, Exim ignored
8034 all but the first. It now scans all returned addresses if a particular IP
8035 value is being sought. In this situation, the contents of the
8036 $dnslist_value variable are a list of all the addresses, separated by a
8039 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8040 to accept a host name and all its aliases if the forward lookup for any of
8041 them yielded the IP address of the incoming connection. Now it accepts only
8042 those names whose forward lookup yields the correct IP address. Any other
8043 names are discarded. This closes a loophole whereby a rogue DNS
8044 administrator could create reverse DNS records to break through a
8045 wildcarded host restriction in an ACL.
8047 37. If a user filter or a system filter that ran in a subprocess used any of
8048 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8049 the wrong values were passed to the pipe command ($thisaddress had the
8050 value of $0, $0 had the value of $1, etc). This bug was introduced by
8051 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8053 38. Improved the line breaking for long SMTP error messages from ACLs.
8054 Previously, if there was no break point between 40 and 75 characters, Exim
8055 left the rest of the message alone. Two changes have been made: (a) I've
8056 reduced the minimum length to 35 characters; (b) if it can't find a break
8057 point between 35 and 75 characters, it looks ahead and uses the first one
8058 that it finds. This may give the occasional overlong line, but at least the
8059 remaining text gets split now.
8061 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8062 file descriptors might be low, and that setting 1000 would always raise it.
8063 It turns out that in some environments, the limit is already over 1000 and
8064 that lowering it causes trouble. So now Exim takes care not to decrease it.
8066 40. When delivering a message, the value of $return_path is set to $sender_
8067 address at the start of routing (routers may change the value). By an
8068 oversight, this default was not being set up when an address was tested by
8069 -bt or -bv, which affected the outcome if any router or filter referred to
8072 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8073 log only when "message" or "log_message" are set. However, if one of the
8074 conditions was an address verification, or a call to a nested ACL, the
8075 messages generated by the underlying test were being passed through. This
8076 no longer happens. The underlying message is available in $acl_verify_
8077 message for both "message" and "log_message" expansions, so it can be
8078 passed through if needed.
8080 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8081 new expansion $bh_ to give the encoded byte string without charset
8082 translation. Translation happens only if iconv() is available; HAVE_ICONV
8083 indicates this at build time. HEADERS_CHARSET gives the charset to
8084 translate to; headers_charset can change it in the configuration, and
8085 "headers charset" can change it in an individual filter file.
8087 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8088 that creates RFC 2047 encoded "words" labels them as that charset instead
8089 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8090 expansion operator; (ii) when Exim creates a From: line for a local
8091 message; (iii) when a header line is rewritten to include a "phrase" part.
8093 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8094 buggy, causing it to skip the first lines of messages whose message ID
8095 ended in 'D'. This would not have bitten before Exim release 4.14, because
8096 message IDs were unlikely to end in 'D' before then. The effect was to have
8097 incorrect size information for certain domains.
8099 45. #include "config.h" was missing at the start of the crypt16.c module. This
8100 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8103 46. If there was a timeout during a "random" callout check, Exim treated it as
8104 a failure of the random address, and carried on sending RSET and the real
8105 address. If the delay was just some slowness somewhere, the response to the
8106 original RCPT would be taken as a response to RSET and so on, causing
8107 mayhem of various kinds.
8109 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8110 when I implemented it. It didn't allow for the fact that some option values
8111 may legitimately be negative (e.g. size_addition), and it didn't even do
8112 the right test for positive values.
8114 48. Domain names in DNS records are case-independent. Exim always looks them up
8115 in lower case. Some resolvers return domain names in exactly the case they
8116 appear in the zone file, that is, they may contain uppercase letters. Not
8117 all resolvers do this - some return always lower case. Exim was treating a
8118 change of case by a resolver as a change of domain, similar to a widening
8119 of a domain abbreviation. This triggered its re-routing code and so it was
8120 trying to route what was effectively the same domain again. This normally
8121 caused routing to fail (because the router wouldn't handle the domain
8122 twice). Now Exim checks for this case specially, and just changes the
8123 casing of the domain that it ultimately uses when it transmits the message
8126 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8129 50. If a filter generated a file delivery with a non-absolute name (possible if
8130 no home directory exists for the router), the forbid_file option was not
8133 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8134 the existing equality matching.
8136 52. Exim was using ints instead of ino_t variables in some places where it was
8137 dealing with inode numbers.
8139 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8140 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8141 called TMPDIR, and if it finds it is different, it changes its value.
8143 54. The smtp_printf() function is now made available to local_scan() so
8144 additional output lines can be written before returning. There is also an
8145 smtp_fflush() function to enable the detection of a dropped connection.
8146 The variables smtp_input and smtp_batched_input are exported to
8149 55. Changed the default runtime configuration: the message "Unknown user"
8150 has been removed from the ACL, and instead placed on the localuser router,
8151 using the cannot_route_message feature. This means that any verification
8152 failures that generate their own messages won't get overridden. Similarly,
8153 the "Unrouteable address" message that was in the ACL for unverifiable
8154 relay addresses has also been removed.
8156 56. Added hosts_avoid_esmtp to the smtp transport.
8158 57. The exicyclog script was not checking for the esoteric option
8159 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8160 will work only if exicyclog is run under the appropriate euid.
8162 58. Following a discussion on the list, the rules by which Exim recognises line
8163 endings on incoming messages have been changed. The -dropcr and drop_cr
8164 options are now no-ops, retained only for backwards compatibility. The
8165 following line terminators are recognized: LF CRLF CR. However, special
8166 processing applies to CR:
8168 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8169 nor a local message in the state where . is a terminator.
8171 (ii) If a bare CR is encountered in a header line, an extra space is added
8172 after the line terminator so as not to end the header. The reasoning
8173 behind this is that bare CRs in header lines are most likely either
8174 to be mistakes, or people trying to play silly games.
8176 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8177 was being incorrectly given as 18 bytes larger than it should have been.
8178 This is a VOB (very old bug).
8180 60. This may never have affected anything current, but just in case it has:
8181 When the local host is found other than at the start of a list of hosts,
8182 the local host, those with the same MX, and any that follow, are discarded.
8183 When the list in question was part of a longer list of hosts, the following
8184 hosts (not currently being processed) were also being discarded. This no
8185 longer happens. I'm not sure if this situation could ever has previously
8188 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8191 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8192 header. This is something that is recommended in a new Internet Draft, and
8193 is something that is documented as being done by Sendmail. There are two
8194 possible values. For messages generated by the autoreply transport, Exim
8197 Auto-Submitted: auto-replied
8199 whereas for all other generated messages (e.g. bounces) it adds
8201 Auto-Submitted: auto-generated
8203 63. The "personal" condition in filters now includes a test for the
8204 Auto-Submitted: header. If it contains the string "auto-" the message it
8205 not considered personal.
8207 64. Added rcpt_include_affixes as a generic transport option.
8209 65. Added queue_only_override (default true).
8211 66. Added the syslog_duplication option.
8213 67. If what should have been the first header line of a message consisted of
8214 a space followed by a colon, Exim was mis-interpreting it as a header line.
8215 It isn't of course - it is syntactically invalid and should therefore be
8216 treated as the start of the message body. The misbehaviour could have
8217 caused a number of strange effects, including loss of data in subsequent
8218 header lines, and spool format errors.
8220 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8221 client host had authenticated. This control can now be exercised by an ACL
8222 for more flexibility.
8224 69. By default, callouts do not happen when testing with -bh. There is now a
8225 variant, -bhc, which does actually run the callout code, including
8226 consulting and updating the callout cache.
8228 70. Added support for saslauthd authentication, courtesy of Alexander
8231 71. If statvfs() failed on the spool or log directories while checking their
8232 size for availability, Exim confusingly gave the error "space shortage".
8233 Furthermore, in debugging mode it crashed with a floating point exception.
8234 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8235 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8236 serious problem, Exim now writes to the main and panic logs when this
8237 happens, with details of the failure. It then refuses to accept the
8238 incoming message, giving the message "spool directory problem" or "log
8239 directory problem" with a 421 code for SMTP messages.
8241 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8242 0, 1, and 2 exist, because some OS complain for execs without them (see
8243 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8244 descriptors. However, the code omitted to check that the open succeeded,
8245 causing mysterious errors if for some reason the permissions on /dev/null
8246 got screwed. Now Exim writes a message to the main and panic logs, and
8247 bombs out if it can't open /dev/null.
8249 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8250 interact so that it is all more flexible. It is supposed to remain
8251 backwards compatible. Also added extra_local_interfaces.
8253 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8254 to bomb out with an assertion failure - to the client this appears as a
8255 connection drop. This problem occurs in the part of the code that was taken
8256 from the Samba project. Fortunately, the assertion is in a very simple
8257 function, so I have fixed this by reproducing the function inline in the
8258 one place where it is called, and arranging for authentication to fail
8259 instead of killing the process with assert().
8261 75. The SPA client code was not working when the server requested OEM rather
8262 than Unicode encoding.
8264 76. Added code to make require_files with a specific uid setting more usable in
8265 the case where statting the file as root fails - usually a non-root-mounted
8266 NFS file system. When this happens and the failure is EACCES, Exim now
8267 forks a subprocess and does the per-uid checking as the relevant uid.
8269 77. Added process_log_path.
8271 78. If log_file_path was not explicitly set, a setting of check_log_space or
8272 check_log_inodes was ignored.
8274 79. If a space check for the spool or log partitions fails, the incident is now
8275 logged. Of course, in the latter case the data may get lost...
8277 80. Added the %p formatting code to string_format() so that it can be used to
8278 print addresses in debug_print(). Adjusted all the address printing in the
8279 debugging in store.c to use %p rather than %d.
8281 81. There was a concern that a line of code in smtp_in.c could overflow a
8282 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8283 initially expressed, the concern was not well-founded, because trailing
8284 spaces are removed early. However, if the trailing spaces were followed by
8285 a NULL, they did not get removed, so the overflow was possible. Two fixes
8288 (a) I re-wrote the offending code in a cleaner fashion.
8289 (b) If an incoming SMTP command contains a NULL character, it is rejected
8292 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8293 runs initgroups(), so that if the Exim user is in any additional groups,
8294 they will be used during message reception.
8300 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob