1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
43 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
44 Find and fix by Jasen Betts.
46 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
47 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
48 the systemwide configuration to override the Exim config.
50 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
51 API changes in libopendmarc.
53 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
54 pid 1, in the normal "background daemon" mode, having to drop process-
55 group leadership also lost track of needing to create listener sockets.
57 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
58 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
60 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
61 a capture group which obtained no text (eg. "(abc)*" matching zero
62 occurrences) could cause a segfault if the corresponding $<n> was
65 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
66 included a close-brace character (eg. it itself used an expansion) an
69 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
70 starting TLS. Previously it was after, meaning that attackers on such
71 ports had to be screened using the host_reject_connection main config
72 option. The new sequence aligns better with the STARTTLS behaviour, and
73 permits defences against crypto-processing load attacks, even though it
74 is strictly an incompatible change.
75 Also, avoid sending any SMTP fail response for either the connect ACL
76 or host_reject_connection, for TLS-on-connect ports.
78 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
79 Previously this was not permitted, but it makes reasonable sense.
80 While there, restore a restriction on using it from a connect ACL; given
81 the change JH/16 it could only return false (and before 4.91 was not
84 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
85 was exactly sized compared to the log buffer, a crash occurred with the
86 misleading message "bad memory reference; pool not found".
87 Found and traced by Jasen Betts.
89 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
90 dns_again_means_nonexist included an element causing a DNS lookup which
91 iteslf returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these lookups.
105 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
108 JH/23 Fix crash in string expansions. Previously, if an empty variable was
109 immediately followed by an expansion operator, a null-indirection read
110 was done, killing the process.
112 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
113 include an SMTP response string which is longer than that supported
114 by the delivering transport. Alleviate by wrapping such lines before
117 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
118 chars (RFC limit). Previously a limit of 12 items was made, which with
119 a not-impossible References: in the message being bounced could still
120 be over-large and get stopped in the transport.
122 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
123 close. Previously a bare socket close was done.
125 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
128 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
129 message being considered for delivery was already being handled by
130 another process, and having an SMTP connection already open, the function
131 to close it tried to use an uninitialized variable. This would afftect
132 high-volume sites more, especially when running mailing-list-style loads.
133 Pollution of logs was the major effect, as the other process delivered
134 the message. Found and partly investigated by Graeme Fowler.
136 JH/29 Change format of the internal ID used for message identification. The old
137 version only supported 31 bits for a PID element; the new 64 (on systems
138 which can use Base-62 encoding, which is all currently supported ones
139 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
140 and must use Base-36). The new ID is 23 characters rather than 16, and is
141 visible in various places - notably logs, message headers, and spool file
142 names. Various of the ancillary utilities also have to know the format.
143 As well as the expanded PID portion, the sub-second part of the time
144 recorded in the ID is expanded to support finer precision. Theoretically
145 this permits a receive rate from a single comms channel of better than the
147 The major timestamp part of the ID is not changed; at 6 characters it is
148 usable until about year 3700.
149 Updating from previously releases is fully supported: old-format spool
150 files are still usable, and the utilities support both formats. New
151 message will use the new format. The one hints-DB file type which uses
152 message-IDs (the transport wait- DB) will be discarded if an old-format ID
153 is seen; new ones will be built with only new-format IDs.
154 Optionally, a utility can be used to convert spool files from old to new,
155 but this is only an efficiency measure not a requirement for operation
156 Downgrading from new to old requires running a provided utility, having
157 first stopped all operations. This will convert any spool files from new
158 back to old (losing time-precision and PID information) and remove any
159 wait- hints databases.
161 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
162 we treated them as item separators when parsing for a list item, but they
163 need to be protected by the doublequotes. While there, add handling for
166 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
167 Found and fixed by Jasen Betts. No testcase for this as my usual text
168 editor insists on emitting only valid UTF-8.
170 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
171 a null-indirection SIGSEGV for the receive process.
177 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
178 after reception to before a subsequent reception. This should
179 mean slightly faster delivery, and also confirmation of reception
182 JH/02 Move from using the pcre library to pcre2. The former is no longer
183 being developed or supported (by the original developer).
185 JH/03 Constification work in the filters module required a major version
186 bump for the local-scan API. Specifically, the "headers_charset"
187 global which is visible via the API is now const and may therefore
188 not be modified by local-scan code.
190 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
191 sendfile() didi not account for the way the ClamAV driver code called it.
193 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
194 time check was being done for every character; replace that with one
197 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
198 sent was prefixed with a length byte.
200 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
201 RFC 5321. Previously Dovecot (at least) would log errors during
204 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
205 in 4.61 without a reason logged. Binaries will be bigger, which might
206 matter on diskspace-constrained systems, but debug is easier.
208 JH/09 Fix macro-definition during "-be" expansion testing. The move to
209 write-protected store for macros had not accounted for these runtime
210 additions; fix by removing this protection for "-be" mode.
212 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
213 handing out large-numbered file descriptors, violating the usual Unix
214 assumption (and required by Posix) that the lowest possible number will be
215 allocated by the kernel when a new one is needed. In the daemon, and any
216 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
217 are not useable for FD_SET() [and hence select()] and overwrite the stack.
218 Assorted crashes happen.
220 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
221 main-section options or in a connect ACL, the value from the first ever
222 connection was never replaced for subsequent connections. Found by
225 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
226 though only once PCRE2 was introduced: the memory accounting used under
227 debug offset allocations by an int, giving a hard trap in early startup.
228 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
230 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
231 with underbars is given. The write-protection of configuration introduced
232 in 4.95 trapped when normalisation was applied to an option not needing
235 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
237 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
238 failing to close the certificates directory, every hour or any time it
241 JH/16 Debugging initiated by an ACL control now continues through into routing
242 and transport processes. Previously debugging stopped any time Exim
243 re-execs, or for processing a queued message.
245 JH/17 The "expand" debug selector now gives more detail, specifically on the
246 result of expansion operators and items.
248 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
249 bad comparison between the option value and the name of the file to
250 be included was done, and a mismatch was wrongly identified.
251 4.88 to 4.95 are affected.
253 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
255 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
256 supplied as the db file-pair basename. Previously, if a directory
257 path was given, for example via the autoreply "once" option, the DB
258 file.pag and file.dir files would be created in that directory's
261 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
262 "taint" log_selector. These were previously deprecated.
264 JH/22 Fix static address-list lookups to properly return the matched item.
265 Previously only the domain part was returned.
267 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
268 the call into OpenSSL to send a TLS Close was being repeated; this
269 resulted in the library waiting for the peer's Close. If that was never
270 sent we waited forever. Fix by tracking send calls.
272 JH/24 The ${run} expansion item now expands its command string elements after
273 splitting. Previously it was before; the new ordering makes handling
274 zero-length arguments simpler. The old ordering can be obtained by
275 appending a new option "preexpand", after a comma, to the "run".
277 JH/25 Taint-check exec arguments for transport-initiated external processes.
278 Previously, tainted values could be used. This affects "pipe", "lmtp" and
279 "queryprogram" transport, transport-filter, and ETRN commands.
280 The ${run} expansion is also affected: in "preexpand" mode no part of
281 the command line may be tainted, in default mode the executable name
284 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
285 the the facility was not passed across execs, and only the first message
286 passed over a connection could use BDAT; any further ones using DATA.
288 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
289 uses $sending_ip_address and an interface is specified.
290 Previously any use of the local address in the EHLO name disabled
291 PIPECONNECT, the common case being to use the rDNS of it.
293 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
294 resumption. Previously verify failed because no certificate status is
295 passed on the wire for the restarted session. Fix by using the recorded
296 ocsp status of the stored session for the new connection.
298 JH/29 TLS resumption: the key for session lookup in the client now includes
299 more info that a server could potentially use in configuring a TLS
300 session, avoiding oferring mismatching sessions to such a server.
301 Previously only the server IP was used.
303 JH/30 Fix string_copyn() for limit greater than actual string length.
304 Previously the copied amount was the limit, which could result in a
305 overlapping memcpy for newly allocated destination soon after a
306 source string shorter than the limit. Found/investigated by KM.
308 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
309 close; it may be needed for a subsequent connection. This caused a
310 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
312 JH/32 Fix CHUNKING for a second message on a connection when the first was
313 rejected. Previously we did not reset the chunking-offered state, and
314 erroneously rejected the BDAT command. Investigation help from
317 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
318 an empty address. Previously the expansion returned an error.
320 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
321 proxy. Previously these were misparsed, leading to paniclog entries.
327 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
328 related applications. Previously an "H" was used where available info
329 says that "M" should be, so change to match.
331 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
332 as arguments, so an implementation trying to copy these into a local
333 buffer was taking a taint-enforcement trap. Fix by using dynamically
334 created buffers. Similar fix for radius expansion condition.
336 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
337 reasonable, eg. to count headers. Fix by using dynamically created
338 buffers rather than a local. Do similar fixes for ACL actions "dcc",
339 "log_reject_target", "malware" and "spam"; the arguments are expanded
340 so could be handling tainted values.
342 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
343 broken the (no-op) support for this sendmail command. Restore it
344 to doing nothing, silently, and returning good status.
346 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
347 record path was given (or the default used) without a leading directory
348 path, an error occurred on trying to open it. Use the transport's working
351 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
352 transport. Previously it was the name on the DNS A-record; use instead
353 the head of the CNAME chain leading there (if there is one). This seems
354 to align better with RFC 6125.
356 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
357 smtp_accept_max_per_host allocated resources which were not released
358 when the limit was exceeded. This eventually crashed the daemon. Fix
359 by adding a release action in that path.
361 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
362 expanded; previously using tainted values was rejected. Fix by using
363 dynamically-created buffers.
365 JH/09 Relax restrictions on ACL verify condition needing access to message
366 headers. Previously they were only permitted in data and non-smtp ACLs;
367 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
368 not_blind, header_sender and header_names_ascii verification.
370 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
371 Previously a macro used one argument twice; when called with the
372 argument as an expression having side-effects, incorrect operation
373 resulted. Use an inlineable function.
375 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
376 held open for a verify callout. Previously this wan not accounted for
377 and a corrupt onward SMTP conversation resulted.
379 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
380 Previously they were embedded naively in the constructed address; when
381 needed, strip the quoting and quote the entire local_part.
382 Also make the inbound_srs expansion condition handle quoting.
384 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
385 excluded, not matching the documentation.
387 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
388 was given for the sqlite_dbfile a trap resulted.
390 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
391 "name" argument resulted in a trap. There is no reason to disallow such;
392 this was a coding error.
394 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
395 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
396 spent suspended, ignoring the POSIX definition. Previously we assumed
397 it did and a constant offset from real time could be used as a correction.
398 Change to using the same clock source for the start-of-message and the
399 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
400 exists, just to get a clock slightly more aligned to reality.
402 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
403 RFC says it is optional some validators care. The missing char was not
404 intended but triggered by a line-wrap alignment. Discovery and fix by
405 Guillaume Outters, hacked on by JH.
407 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
408 name being quoted was tainted a trap would be taken. Fix by using
409 dynamically created buffers. The routine could have been called by a
410 rewrite with the "h" flag, by using the "-F" command-line option, or
411 by using a "name=" option on a control=submission ACL modifier.
413 JH/19 SPF: change the Authentication-Results expansion component to give
414 smtp.helo when the sender domain is empty. Previously it gave
417 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
418 not in 127.0.0.0/8 to help in spotting list domains taken over by a
419 domain-parking registrar.
421 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
422 Previously when a whitespace character was specified it was not inserted
423 after removing the newline.
425 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
426 the domain part of the recipient address. This overrides any tls_sni
427 option set, which was previously used.
429 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
432 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
433 is_tainted() had an off-by-one error in the overenthusiastic direction.
434 Find and fix by Gavan. Although NetBSD is not a supported platform for
435 4.94 this bug could affect other platforms.
437 PP/01 Fix default prime selection to be consistent.
438 One path used ike23 still, instead of exim.dev.20160529.3; now both
439 execution flows will use the same DH primes (currently
440 exim.dev.20160529.3).
442 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
443 option in smtp transport, to match the documentation. Previously
444 verification was not being done in some cases where it should have been.
446 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
447 than one server was defined and depending on the platform memory layout
448 details, an internal consistency trap could be hit while walking the list
451 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
452 files. The value is used by the authresults expansion item. Previously
453 if this was used in a router or transport, a crash could result.
455 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
456 have a sufficient incidence of EINTR returns from select that an
457 interaction with other operations done by the main daemon loop exposed
458 a bug in the error-handling. This was benign apart from the log
461 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
462 for consistency with delivered "=>" and failed "**" lines. While we're
463 there, handle PRX and TFO.
465 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
466 applied. This resulted, if any header-line rewrite rules were configured,
467 in a panic-log triggerable by sending a message with a long address in
468 a header. Fix by increasing the arbitrary limit to larger than a single
469 (dewrapped) 5322 header line maximum size.
471 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
472 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
473 RFC 6648 which deprecates X- options in protocols as a general practice.
474 Changeover between the implementations is handled by the mechanisms
477 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
478 a remote which uses callout/hold. Previously the local was queued.
480 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
483 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
484 log line. The asterisk is now consistently appended to the remote IP
485 (and port, if given), and will also be provided on defer and fail log
486 lines. Previously it could be placed on the local IP if that was being
487 logged, and was only provided on delivery lines.
489 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
491 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
492 authenticator client_send option. Previously the next char, after a pair
493 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
494 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
495 documentation. There is still no way to get a leading ^ immediately
496 after a NUL (ie. for the password of a PLAIN method authenticator.
498 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
499 files. For bad sizes read, delete the record and whine to paniclog.
501 JH/38 When logging an AUTH failure, as server, do not include sensitive
502 information. Previously, the credentials would be included if given
503 as part of the AUTH command line and an ACL denied authentication.
505 JH/39 Bug 2691: fix $local_part_data. When the matching list element
506 referred to a file, bad data was returned. This likely also affected
509 JH/40 The gsasl authenticator now supports caching of the salted password
510 generated by the client-side implementation. This required the addition
511 of a new variable: $auth4.
513 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
514 left undeleted; the attempt to re-create it then failed - resulting in
515 the usual "SIGHUP tp have daemon reload configuration" to not work.
516 This affected any platform not supporting "abstract" Unix-domain
517 sockets (i.e. not Linux).
519 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
520 recipients" response to RCPT in a later response, with a 250. The
521 previous coding assumed this would not happen, and under PIPELINING
522 would result in both lost and duplicate recipients for a message.
524 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
525 Previously the weighting was incorrectly applied. Similar fix for socks
526 proxies. Found and fixed by Heiko Schlichting.
528 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
529 not handle sub-lists included using the +namedlist syntax. While
530 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
531 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
534 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
535 Previously the permanent pool was used, so the sore could not be freed.
536 This meant a connection with many messages would use continually-growing
539 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
540 per-pool so as not to waste too much space. Previously a constant size
541 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
542 DOS attacks harder. The cost is wasted memory use in the larger blocks.
544 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
545 same space repeatedly, and a smaller process footprint.
547 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
548 for GnuTLS. Previously the name started "%s" which, while not a bug,
549 looked as if if might be one.
551 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
552 "continued connection") make the $proxy_* variables available. Previously
553 the information was not passed across the exec() call for subsequent
554 transport executions. This also mean that the log lines for the
555 messages can show the proxy information.
557 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
558 receive time. With modern systems the difference is significant.
559 The historical behaviour can be restored by disabling (a new) log_selector
560 "queue_time_exclusive".
562 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
563 content) were specifically tested for. Now, also cover expansions
564 resulting in ACL names and inline ACL content.
566 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
567 making it unusable in complex expressions.
569 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
570 fail with a taint-check on the spoolfile name, and leave the message
573 HS/01 Enforce absolute PID file path name.
575 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
577 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
579 PP/02 Bug 2643: Correct TLS DH constants.
580 A missing NUL termination in our code-generation tool had led to some
581 incorrect Diffie-Hellman constants in the Exim source.
582 Reported by kylon94, code-gen tool fix by Simon Arlott.
584 PP/03 Impose security length checks on various command-line options.
585 Fixes CVE-2020-SPRSS reported by Qualys.
587 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
588 better. Reported by Qualys.
590 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
591 providing a particularly obnoxious sender full name.
594 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
596 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
599 PP/08 Change default for recipients_max from unlimited to 50,000.
601 PP/09 Fix security issue with too many recipients on a message (to remove a
602 known security problem if someone does set recipients_max to unlimited,
603 or if local additions add to the recipient list).
604 Fixes CVE-2020-RCPTL reported by Qualys.
606 PP/10 Fix security issue in SMTP verb option parsing
607 Fixes CVE-2020-EXOPT reported by Qualys.
609 PP/11 Fix security issue in BDAT state confusion.
610 Ensure we reset known-good where we know we need to not be reading BDAT
611 data, as a general case fix, and move the places where we switch to BDAT
612 mode until after various protocol state checks.
613 Fixes CVE-2020-BDATA reported by Qualys.
615 HS/03 Die on "/../" in msglog file names
617 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
618 the Exim runtime user are allowed to create files.
620 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
623 QS/03 When reading the output from interpreted forward files we do not
624 pass the pipe between the parent and the interpreting process to
625 executed child processes (if any).
627 QS/04 Always die if requested from internal logging, even is logging is
630 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
631 the API; compilation noo longer completes with DMARC support included.
632 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
633 been reported on other platforms.
635 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
637 JH/56 Make the majority of info read from config files readonly, for defence-in-
638 depth against exploits. Suggestion by Qualys.
639 Not supported on Solaris 10.
641 JH/57 Fix control=fakreject for a custom message containing tainted data.
642 Previously this resulted in a log complaint, due to a re-expansion present
643 since fakereject was originally introduced.
645 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
646 modifier was given, a loop resulted.
648 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
649 pipelined SMTP command followed the BDAT LAST then it would be
650 incorrectly treated as part of the message body, causing a verification
653 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
654 require looser validation rules than those for 821-level addresses,
655 which only permit IP addresses.
661 JH/01 Avoid costly startup code when not strictly needed. This reduces time
662 for some exim process initialisations. It does mean that the logging
663 of TLS configuration problems is only done for the daemon startup.
665 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
667 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
668 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
671 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
672 DKIM signing is being done. Previously a transport_filter would always
673 disable CHUNKING, falling back to traditional DATA.
675 JH/05 Regard command-line recipients as tainted.
677 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
679 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
680 PAM library frees one of the arguments given to it, despite the
681 documentation. Therefore a plain malloc must be used.
683 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
684 on-stack buffers were used, resulting in a taint trap when DSN information
685 copied from a received message was written into the buffer.
687 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
688 the ordering of its ARC headers. This caused a crash.
690 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
691 a new record was being constructed with information from the peer, a trap
694 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
695 installation would get error messages from DMARC verify, when it hit the
696 nonexistent file indicated by the default. Distros wanting DMARC enabled
697 should both provide the file and set the option.
698 Also enforce no DMARC verification for command-line sourced messages.
700 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
701 could, depending on the platform, hang at the STARTTLS response.
703 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
704 message on a connection. Previously if one message had ARC headers and
705 the following one did not, a crash could result when adding an
706 Authentication-Results: header.
708 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
709 Exim main code and Exim-related utities. The introduction of taint
710 tracking also did many adjustments to string handling. Since then, eximon
711 frequently terminated with an assert failure.
713 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
714 check for 452 responses. This slightly helps the inefficieny of doing
715 a large alias-expansion into a recipient-limited target. The max_rcpt
716 transport option still applies (and at the current default, will override
717 the new feature). The check is done for either cause of synch, and forces
718 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
719 connection. The new facility is not tunable at this time.
721 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
722 library live data was being used, so the results became garbage. Make
723 copies while it is still usable.
725 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
726 on delivery deferred (==) and failed (**) lines (if a delivery was
727 attemtped). Previously it was only on completion (=>) lines.
729 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
730 for the expansion of the server_scram_iter and server_scram_salt options.
732 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
733 are now specifically given a NO_DATA response without hitting the system
734 resolver. The library goes on to do the now-standard TXT lookup.
735 Use of dnsdb lookups is not affected.
737 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
738 only retrieve the errormessage once. Previously two calls to dlerror()
739 were used, and the second one (for mainlog/paniclog) retrieved null
742 JH/20 Taint checking: disallow use of tainted data for
743 - the appendfile transport file and directory options
744 - the pipe transport command
745 - the autoreply transport file, log and once options
746 - file names used by the redirect router (including filter files)
748 - paths used by single-key lookups
749 Previously this was permitted.
751 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
752 adjusted the size of a major service buffer; this failed because the
753 buffer was in use at the time. Change to a compile-time increase in the
754 buffer size, when this authenticator is compiled into exim.
756 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
757 previous fast-mode was untenable in the face of glibs using mmap to
758 support larger malloc requests.
760 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
761 New values supported, if defined on system where compiled:
762 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
763 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
765 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
766 running a limited number of proceses in parallel, a benefit is gained. The
767 amount varies with the platform hardware and load. The use of the option
768 queue_run_in_order means we cannot do this, as ordering becomes
771 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
772 had introduced a string-copy (for ensuring NUL-termination) which was not
773 appropriate for that case, which can include embedded NUL bytes in the
774 block of data. Investigation showed the copy to actually be needless, the
775 data being length-specified.
777 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
778 done during a receiving connection, and both used TLS, global info was
779 used rather than per-connection info for tracking the state of data
780 queued for transmission. This could result in a connection hang.
782 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
783 Previously, when delivering serveral messages down a single connection
784 only the first would provide a SIZE. This was due to the size information
785 not being properly tracked.
787 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
788 TAI (at 37 seconds currently), pretend to be in UTC for time-related
789 expansion and logging. Previously, spurious values such as a future
790 minute could be seen.
792 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
793 it could crash from a null-deref. This could also affect the
794 ${addresses: } operator and ${readsock } item.
796 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
797 message following a mime one, the variable was not reset.
799 JH/30 When an pipelined-connect fails at the first response, assume incorrect
800 cached capability (perhaps the peer reneged?) and immediately retry in
803 JH/31 Fix spurious detection of timeout while writing to transport filter.
805 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
806 an attempt to copy the string was made before checking it.
808 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
809 taint of the lookup key was maintained; we now regard the presence in the
810 filesystem as sufficient validation.
812 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
813 argument is supplied.
815 JH/35 The dsearch lookup now requires that the directory is an absolute path.
816 Previously this was not checked, and nonempty relative paths made an
817 access under Exim's current working directory.
819 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
820 Previously no event was raised.
822 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
823 parameter supplied by the sender MAIL FROM command. Previously it was
824 ignored, and only the check_spool_space option value for the required
827 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
828 the size of the signing public-key. Previously it was instead giving
829 the size of the signature hash.
831 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
832 the default. See the (new) dkim_verify_min_keysizes option.
834 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
835 and an ACL use a lookup for checking either the local_part or domain,
836 stale data could be accessed. Ensure that variable references are
837 dropped between messages.
839 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
840 by the client was not checked as pointing within response data before
841 being used. A malicious client could thus cause an out-of-bounds read and
842 possibly gain authentication. Fix by adding the check.
844 JH/42 Internationalisation: change the default for downconversion in the smtp
845 transport to be "if needed". Previously it was "as previously set" for
846 the message, which usually meant "if needed" for message-submission but
847 "no" for everything else. However, MTAs have been seen using SMTPUTF8
848 even when the envelope addresses did not need it, resulting in forwarding
849 failures to non-supporting MTAs. A downconvert in such cases will be
850 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
851 transport. The change does mean that addresses needing conversion will
852 be converted when previously a delivery failure would occur.
854 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
855 response was received it would be used unchecked in a fail-DSN, violating
856 standards on line-length limits. Truncate if needed.
858 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
859 doesn't work. (Additionally add a new main config option to configure the
866 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
867 side, to decode a TLS 1.3 packet capture.
869 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
870 Previously the default library behaviour applied, sending two, each in
873 JH/03 Debug output for ACL now gives the config file name and line number for
876 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
878 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
880 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
881 buffer overrun for (non-chunking) other transports.
883 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
884 TLS1.3, means that a server rejecting a client certificate is not visible
885 to the client until the first read of encrypted data (typically the
886 response to EHLO). Add detection for that case and treat it as a failed
887 TLS connection attempt, so that the normal retry-in-clear can work (if
888 suitably configured).
890 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
891 and/or domain. Found and fixed by Jason Betts.
893 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
894 configuration). If a CNAME target was not a wellformed name pattern, a
897 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
898 the OS reports them interleaved with other addresses.
900 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
901 used both for input and for a verify callout, both encrypted, SMTP
902 responses being sent by the server could be lost. This resulted in
903 dropped connections and sometimes bounces generated by a peer sending
906 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
907 string. Previously it was possible to cause undefined behaviour in a
908 library routine (usually a crash). Found by "zerons".
910 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
913 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
914 API was removed, so update to use the newer ones.
916 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
917 any timeout set, is taking a long time. Previously we would hang on to a
918 rotated logfile "forever" if the input was arriving with long gaps
919 (a previous attempt to fix addressed lack, for a long time, of initial
922 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
923 shared (NFS) environment. The length of the tempfile name is now
924 4 + 16 ("hdr.$message_exim_id") which might break on file
925 systems which restrict the file name length to lower values.
928 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
929 shared (NFS) environment.
931 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
932 did for all versions <4.90). Notably -M, -m, --invert, -I may be
935 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
936 on some platforms for bit 31.
938 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
939 to changes apparently associated with TLS1.3 handling some of the APIs
940 previously used were either nonfunctional or inappropriate. Strings
941 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
942 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
943 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
944 This affects log line X= elements, the $tls_{in,out}_cipher variables,
945 and the use of specific cipher names in the encrypted= ACL condition.
947 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
949 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
950 verification result was not updated unless hosts_require_ocsp applied.
952 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
953 queue_list_requires_admin set to false, non-admin users were denied the
956 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
957 directory-of-certs mode. Previously they were advertised despite the
960 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
961 A single TCP connection by a client will now hold a TLS connection open
962 for multiple message deliveries, by default. Previously the default was to
965 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
966 default. If built with the facility, DANE will be used. The facility
967 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
969 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
970 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
971 must be defined and you must still, unless you define DISABLE_TLS, manage
972 the the include-dir and library-file requirements that go with that
973 choice. Non-TLS builds are still supported.
975 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
978 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
979 default. If the platform supports and has the facility enabled, it will
980 be requested on all coneections.
982 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
983 controlled by the build-time option SUPPORT_PIPE_CONNECT.
985 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
987 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
988 success-DSN messages. Previously the From: header was always the default
989 one for these; the option was ignored.
991 JH/28 Fix the timeout on smtp response to apply to the whole response.
992 Previously it was reset for every read, so a teergrubing peer sending
993 single bytes within the time limit could extend the connection for a
994 long time. Credit to Qualsys Security Advisory Team for the discovery.
996 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
997 delivery address, which leaked information of the results of local
998 forwarding. Change to the original envelope recipient address, per
1001 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1002 requested. Previously not bounce was generated and a log entry of
1003 error ignored was made.
1005 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1007 JH/32 Introduce a general tainting mechanism for values read from the input
1008 channel, and values derived from them. Refuse to expand any tainted
1009 values, to catch one form of exploit.
1011 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1012 was unused and the unexpanded text used for the test. Found and
1013 fixed by Ruben Jenster.
1015 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1016 an attempt to use a TLS library read routine dereffed a nul pointer,
1019 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1020 them in our smtp response.
1022 JH/36 Have the general router option retry_use_local_part default to true when
1023 any of the restrictive preconditions are set (to anything). Previously it
1024 was only for check_local user. The change removes one item of manual
1025 configuration which is required for proper retries when a remote router
1026 handles a subset of addresses for a domain.
1028 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1029 link count into consideration.
1031 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1032 caused the extension of big_buffer, the following lines were ignored.
1034 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1035 accordance with RFC 2308. Previously there was no expiry, so a longlived
1036 receive process (eg. due to ACL delays) versus a short SOA value could
1039 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1041 JH/39 Promote DMARC support to mainline.
1043 JH/40 Bug 2452: Add a References: header to DSNs.
1045 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1046 parameters. The relevant library call is documented as "Deprecated: This
1047 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1048 3.6.0, DH parameters are negotiated following RFC7919."
1050 HS/06 Change the default of dnssec_request_domains to "*"
1052 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1053 carried on and emitted a BDAT command, even when PIPELINING was not
1056 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1057 buffer was used for the filename, resulting in a trap when tainted
1058 arguments (eg. $domain) were used.
1060 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1061 recommended to avoid a possible server-load attack. The feature can be
1062 re-enabled via the openssl_options main cofiguration option.
1064 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1065 for version 4.90 - adding a "more data" boolean to the arguments.
1066 Bumped the ABI version number also, this having been missed previously;
1067 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1068 respect of usage of smtp_printf() by either local_scan code or libraries
1069 accessed via the ${dlfunc } expansion item. Both will need coding
1070 adjustment for any calls to smtp_printf() to match the new function
1071 signature; a FALSE value for the new argument is always safe.
1073 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1074 the file-offset (which the Linux syscall does, and exim expects); this
1075 resulted in an indefinite loop.
1077 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1078 to do ARC verification. The Authentication-Results: header line added
1079 by the configuration then had no ARC item.
1081 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1082 an incorrect mode was used for reading the result, resulting in it being
1089 JH/01 Remove code calling the customisable local_scan function, unless a new
1090 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1092 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1093 non-signal-safe functions being used.
1095 JH/03 Bug 2269: When presented with a received message having a stupidly large
1096 number of DKIM-Signature headers, disable DKIM verification to avoid
1097 a resource-consumption attack. The limit is set at twenty.
1099 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1100 report of oldest_pass in ${authres } in consequence, and separate out
1101 some descriptions of reasons for verification fail.
1103 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1104 files in the spool were present and unlocked. A queue-runner could spot
1105 them, resulting in a duplicate delivery. Fix that by doing the unlock
1106 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1107 add more error-checking on spoolfile handling while that code is being
1110 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1111 No known attacks, no CVE, this is defensive hardening.
1113 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1114 a queue-runner could start a delivery while other operations were ongoing.
1115 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1116 Found and investigated by Tim Stewart. Fix by using the open message data
1117 file handle rather than opening another, and not locally closing it (which
1118 releases a lock) for that case, while creating the temporary .eml format
1119 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1121 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1122 $sender_verify_failure/$recipient_verify_failure to "random".
1124 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1127 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1128 Previously this would segfault.
1130 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1133 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1134 like zero, since the resolver should be doing this for us, But we need one
1135 as a CNAME but no MX presence gets the CNAME returned; we need to check
1136 that doesn't point to an MX to declare it "no MX returned" rather than
1137 "error, loop". A new main option is added so the older capability of
1138 following some limited number of chain links is maintained.
1140 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1142 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1143 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1144 now-deprecated earlier definitions used only specified the range up to TLS
1145 1.2 (in the older-version library docs).
1147 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1149 JH/15 Rework TLS client-side context management. Stop using a global, and
1150 explicitly pass a context around. This enables future use of TLS for
1151 connections to service-daemons (eg. malware scanning) while a client smtp
1152 connection is using TLS; with cutthrough connections this is quite likely.
1154 JH/16 Fix ARC verification to do AS checks in reverse order.
1156 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1158 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1159 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1160 Previously the "utf8" would be re-prepended for every additional message.
1162 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1163 Previously thery were accepted, resulting in issues when attempting to
1164 forward messages to a non-supporting MTA.
1166 PP/02 Let -n work with printing macros too, not just options.
1168 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1169 one parent address was copied, and bogus data was used at delivery-logging
1170 time. Either a crash (after delivery) or bogus log data could result.
1171 Discovery and analysis by Tim Stewart.
1173 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1174 Previously if the string ended mid-character, we did not insert the
1175 promised '?' replacement.
1177 PP/04 Documentation: current string operators work on bytes, not codepoints.
1179 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1180 should pack well giving a smaller memory footprint so better caching and
1181 therefore performance. Group the declarations where this can't be done so
1182 that the byte-sized flag variables are not interspersed among pointer
1183 variables, giving a better chance of good packing by the compiler.
1185 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1186 non-null, to avoid issues with sites running BATV. Previously reports were
1187 sent with an empty envelope sender so looked like bounces.
1189 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1190 The ignore_error flag wasn't being returned from the filter subprocess so
1191 was not set for later routers. Investigation and fix by Matthias Kurz.
1193 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1194 and a msg:complete for the whole, when a message is manually removed using
1195 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1197 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1198 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1199 hope that is portable enough.
1201 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1202 requiring a known-CA anchor certificate; make it now rely entirely on the
1203 TLSA as an anchor. Checking the name on the leaf cert against the name
1204 on the A-record for the host is still done for TA (but not for EE mode).
1206 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1207 included in delivery lines for non-proxied connections, when compiled with
1208 SUPPORT_SOCKS and running with proxy logging enabled.
1210 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1211 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1212 move the existing event to fire before the normal logging of message
1213 failure so that custom logging is bracketed by normal logging.
1215 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1216 msg:fail:internal event. Developement by Matthias Kurz.
1218 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1219 far too small for todays use of crypto signatures stored there. Go all
1220 the way to the max DNS message size of 64kB, even though this might be
1221 overmuch for IOT constrained device use.
1223 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1224 copy a string over itself. The library routine is documented as not
1225 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1227 JH/32 For main options check_spool_space and check_inode_space, where the
1228 platform supports 64b integers, support more than the previous 2^31 kB
1229 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1230 the previous G, M, k.
1232 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1233 $authenticated_fail_id variable on authentication failure. Previously
1236 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1237 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1238 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1239 GNUTLS_SEC_PARAM_MEDIUM.
1241 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1242 side. Previously we would continue as if no SNI had been received.
1244 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1245 "<" character, which should be a list-separator specification, we walked
1246 off past the nul-terimation.
1248 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1249 causes) even when the retry time is not yet met. Previously they were
1250 not, meaning that when (say) an account was over-quota and temp-rejecting,
1251 and multiple senders' messages were queued, only one sender would get
1252 notified on each configured delay_warning cycle.
1254 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1256 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1257 cipher-suites, an error can be left on the stack even for a succeeding
1258 accept; this results in impossible error messages when a later operation
1261 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1262 return error codes indicating retry. Under TLS1.3 this becomes required.
1264 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1265 it only wrote the new authenticators, resulting in a lack of tracking of
1266 peer changes of ESMTP extensions until the next cache flush.
1268 JH/41 Fix the loop reading a message header line to check for integer overflow,
1269 and more-often against header_maxsize. Previously a crafted message could
1270 induce a crash of the recive process; now the message is cleanly rejected.
1272 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1273 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1279 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1280 When redis_servers is set to a list of > 1 element, and the Redis servers
1281 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1282 case of MOVED into a DEFER case instead, thus moving the query onto the
1283 next server in the list. For a cluster of N elements, all N servers must
1284 be defined in redis_servers.
1286 GF/02 Catch and remove uninitialized value warning in exiqsumm
1287 Check for existence of @ARGV before looking at $ARGV[0]
1289 JH/01 Replace the store_release() internal interface with store_newblock(),
1290 which internalises the check required to safely use the old one, plus
1291 the allocate and data copy operations duplicated in both (!) of the
1292 extant use locations.
1294 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1295 modifier. This matches the restriction on the commandline.
1297 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1298 Previously only the last row was returned.
1300 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1301 we assumed that tags in the header were well-formed, and parsed the
1302 element content after inspecting only the first char of the tag.
1303 Assumptions at that stage could crash the receive process on malformed
1306 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1307 While running the DKIM ACL we operate on the Permanent memory pool so that
1308 variables created with "set" persist to the DATA ACL. Also (at any time)
1309 DNS lookups that fail create cache records using the Permanent pool. But
1310 expansions release any allocations made on the current pool - so a dnsdb
1311 lookup expansion done in the DKIM ACL releases the memory used for the
1312 DNS negative-cache, and bad things result. Solution is to switch to the
1313 Main pool for expansions.
1314 While we're in that code, add checks on the DNS cache during store_reset,
1315 active in the testsuite.
1316 Problem spotted, and debugging aided, by Wolfgang Breyha.
1318 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1319 When none of the hosts presented to a transport match an already-open
1320 connection, close it and proceed with the list. Previously we would
1321 queue the message. Spotted by Lena with Yahoo, probably involving
1324 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1325 Previously a spurious "250 OK id=" response was appended to the proper
1328 JH/08 The "support for" informational output now, which built with Content
1329 Scanning support, has a line for the malware scanner interfaces compiled
1330 in. Interface can be individually included or not at build time.
1332 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1333 by the template makefile "src/EDITME". The "STREAM" support for an older
1334 ClamAV interface method is removed.
1336 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1337 rows affected is given instead).
1339 JH/11 The runtime Berkeley DB library version is now additionally output by
1340 "exim -d -bV". Previously only the compile-time version was shown.
1342 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1343 SMTP connection. Previously, when one had more recipients than the
1344 first, an abortive onward connection was made. Move to full support for
1345 multiple onward connections in sequence, handling cutthrough connection
1346 for all multi-message initiating connections.
1348 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1349 routers. Previously, a multi-recipient message would fail to match the
1350 onward-connection opened for the first recipient, and cause its closure.
1352 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1353 a timeout on read on a GnuTLS initiating connection, resulting in the
1354 initiating connection being dropped. This mattered most when the callout
1355 was marked defer_ok. Fix to keep the two timeout-detection methods
1358 JH/15 Relax results from ACL control request to enable cutthrough, in
1359 unsupported situations, from error to silently (except under debug)
1360 ignoring. This covers use with PRDR, frozen messages, queue-only and
1363 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1365 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1366 metadata, resulting in a crash in free().
1368 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1369 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1370 Broken also in d185889f4, with init system revamp.
1372 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1373 Previously we abruptly closed the connection after reading a malware-
1374 found indication; now we go on to read the "scan ok" response line,
1377 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1378 ACL. Previously, a crash would result.
1380 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1381 macro text after a replacement (previously it was only once per line) and
1382 by skipping builtin macros when searching for an uppercase lead character.
1384 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1385 for the build is renamed.
1387 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1388 was allocated for every new TLS startup, meaning one per message. Fix
1389 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1391 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1392 reported the original. Fix to report (as far as possible) the ACL
1393 result replacing the original.
1395 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1396 OpenSSL. Certificate information is loaded for every new TLS startup,
1397 and the resources needed to be freed.
1399 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1401 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1404 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1405 DATA response info to the (existing) per-recipient response info for
1406 the "C=" log element. It can have useful tracking info from the
1407 destination system. Patch from Simon Arlott.
1409 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1410 length value. Previously this would segfault.
1412 HS/02 Support Avast multiline protoocol, this allows passing flags to
1413 newer versions of the scanner.
1415 JH/28 Ensure that variables possibly set during message acceptance are marked
1416 dead before release of memory in the daemon loop. This stops complaints
1417 about them when the debug_store option is enabled. Discovered specifically
1418 for sender_rate_period, but applies to a whole set of variables.
1419 Do the same for the queue-runner and queue-list loops, for variables set
1420 from spool message files. Do the same for the SMTP per-message loop, for
1421 certain variables indirectly set in ACL operations.
1423 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1424 as a multi-recipient message from a mailinglist manager). The coding had
1425 an arbitrary cutoff number of characters while checking for more input;
1426 enforced by writing a NUL into the buffer. This corrupted long / fast
1427 input. The problem was exposed more widely when more pipelineing of SMTP
1428 responses was introduced, and one Exim system was feeding another.
1429 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1430 receiving system, and refused recipients seen by the sending system
1431 (propating to people being dropped from mailing lists).
1432 Discovered and pinpointed by David Carter.
1434 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1435 replaced by the ${authresults } expansion.
1437 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1439 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1440 allows proper process termination in container environments.
1442 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1443 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1445 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1446 and "err_perm", deprecated since 4.83 when the RFC-defined words
1447 "temperror" and "permerror" were introduced.
1449 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1450 transport-filters or DKIM-signing. The restriction was lost in the
1451 consolidation of verify-callout and delivery SMTP handling.
1452 Extend the restriction to also cover ARC-signing.
1454 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1455 in defer=pass mode supply a 450 to the initiator. Previously the message
1458 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1459 tls_require_ciphers is used as before.
1461 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1462 "pass_unscanned". Only tmpfails from the scanner are written to
1463 the paniclog, as they may require admin intervention (permission
1464 denied, license issues). Other scanner errors (like decompression
1465 bombs) do not cause a paniclog entry.
1467 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1468 Previously it was possible to log spurious information in receive log
1471 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1472 triggered odd behaviour from Outlook Express clients.
1474 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1477 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1478 since the IETF WG has not yet settled on that versus the original
1479 "bare" representation.
1481 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1482 Previously the millisecond value corrupted the output.
1483 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1484 corrupted the output.
1490 JH/01 Rework error string handling in TLS interface so that the caller in
1491 more cases is responsible for logging. This permits library-sourced
1492 string to be attached to addresses during delivery, and collapses
1493 pairs of long lines into single ones.
1495 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1496 during configuration. Wildcards are allowed and expanded.
1498 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1499 This permits better logging.
1501 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1502 do not close it down and have the child transport start it up again on
1503 the passed-on TCP connection. Instead, proxy the child (and any
1504 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1505 affected: the continued delivery log lines do not have any DNSSEC, TLS
1506 Certificate or OCSP information. TLS cipher information is still logged.
1508 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1509 identical IP addresses on different listening ports. Will also affect
1512 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1513 add noisy ifdef guards to special-case this sillyness.
1514 Patch from Bernd Kuhls.
1516 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1517 than 255 are no longer allowed.
1519 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1520 Disable the session-cache too, which might reduce our load. Since we
1521 currrectly use a new context for every connection, both as server and
1522 client, there is no benefit for these.
1523 GnuTLS appears to not support tickets server-side by default (we don't
1524 call gnutls_session_ticket_enable_server()) but client side is enabled
1525 by default on recent versions (3.1.3 +) unless the PFS priority string
1528 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1529 <https://reproducible-builds.org/specs/source-date-epoch/>.
1531 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1532 the check for any unsuccessful recipients did not notice the limit, and
1533 erroneously found still-pending ones.
1535 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1536 MSG_MORE. Only in-clear (not on TLS connections).
1538 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1539 transport-filter is involved we can buffer the headers in memory for
1540 creating the signature, and read the spool data file once for the
1541 signature and again for transmission.
1543 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1544 4.77 as the kernel support then wasn't solid, having issues in 64bit
1545 mode. Now, it's been long enough. Add support for FreeBSD also.
1547 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1548 case where the routing stage had gathered several addresses to send to
1549 a host before calling the transport for the first, we previously failed
1550 to close down TLS in the old transport process before passing the TCP
1551 connection to the new process. The new one sent a STARTTLS command
1552 which naturally failed, giving a failed delivery and bloating the retry
1553 database. Investigation and fix prototype from Wolfgang Breyha.
1555 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1556 false-negatives in the check that the sender had not preempted a response
1557 or prompt from Exim (running as a server), due to that code's lack of
1558 awareness of the SMTP input buffering.
1560 PP/04 Add commandline_checks_require_admin option.
1561 Exim drops privileges sanely, various checks such as -be aren't a
1562 security problem, as long as you trust local users with access to their
1563 own account. When invoked by services which pass untrusted data to
1564 Exim, this might be an issue. Set this option in main configuration
1565 AND make fixes to the calling application, such as using `--` to stop
1568 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1569 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1570 responses to those, into a single TLS record each way (this usually means
1571 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1574 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1575 affects you only if you're dancing at the edge of the param size limits.
1576 If you are, and this message makes sense to you, then: raise the
1577 configured limit or use OpenSSL 1.1. Nothing we can do for older
1580 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1581 cmdline element to get the documented default one. Previously it was
1584 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1585 are unsuable for non-message-specific reasons - eg. connection timeout,
1586 banner-time rejection.
1588 JH/16 Fix logging of delivery remote port, when specified by router, under
1591 PP/06 Repair manualroute's ability to take options in any order, even if one
1592 is the name of a transport.
1595 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1597 JH/17 Change the list-building routines interface to use the expanding-string
1598 triplet model, for better allocation and copying behaviour.
1600 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1601 Previously it was constructed the first time a possibly-matching string
1602 was met in the configuration file input during startup; now it is done
1605 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1606 compatible one, to avoid the (poorly documented) possibility of a config
1607 file in the working directory redirecting the DB files, possibly correpting
1608 some existing file. CVE-2017-10140 assigned for BDB.
1610 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1611 cache-hot. Previously, although the result was properly cached, the
1612 initial verify call returned a defer.
1614 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1615 the main verify for receipient in uncached-mode.
1617 JH/22 Retire historical build files to an "unsupported" subdir. These are
1618 defined as "ones for which we have no current evidence of testing".
1620 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1621 if present. Previously it was ignored.
1623 JH/24 Start using specified-initialisers in C structure init coding. This is
1624 a C99 feature (it's 2017, so now considered safe).
1626 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1627 if was a fixed-sized field and bitmask ops via macros; it is now more
1630 PP/07 GitHub PR 56: Apply MariaDB build fix.
1631 Patch provided by Jaroslav Škarvada.
1633 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1634 during Coverity cleanups [4.87 JH/47]
1635 Diagnosis and fix provided by Michael Fischer v. Mollard.
1637 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1638 the right size to place the terminating semicolon on its own folded
1639 line, the header hash was calculated to an incorrect value thanks to
1640 the (relaxed) space the fold became.
1642 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1643 and confused the parent.
1645 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1646 which could crash as a result. This could lead to undeliverable messages.
1648 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1651 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1652 matters on fast-turnover and PID-randomising systems, which were getting
1653 out-of-order delivery.
1655 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1656 a possibly-overlapping copy. The symptom was that "Remote host closed
1657 connection in response to HELO" was logged instead of the actual 4xx
1660 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1661 Previously only that bufferd was discarded, resulting in SYMTP command
1664 JH/32 DKIM: when a message has multiple signatures matching an identity given
1665 in dkim_verify_signers, run the dkim acl once for each. Previously only
1666 one run was done. Bug 2189.
1668 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1669 "panic the current process" to "deliberately defer". The panic log is
1670 still written with the problem list name; the mail and reject logs now
1671 get a temp-reject line for the message that was being handled, saying
1672 something like "domains check lookup or other defer". The SMTP 451
1673 message is still "Temporary local problem".
1675 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1676 A crafted sequence of BDAT commands could result in in-use memory beeing
1677 freed. CVE-2017-16943.
1679 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1680 from SMTP input. Previously it was always done; now only done for DATA
1681 and not BDAT commands. CVE-2017-16944.
1683 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1684 to the message (such as an overlong header line). Previously this was
1685 not done and we did not exit BDAT mode. Followon from the previous item
1686 though a different problem.
1692 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1693 than -2003 did; needs libidn2 in addition to libidn.
1695 JH/02 The path option on a pipe transport is now expanded before use.
1697 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1698 Patch provided by "Björn", documentation fix added too.
1700 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1701 missing a wire-to-host endian conversion.
1703 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1704 close after a BDAT command line could be taken as a following command,
1705 giving a synch failure. Fix by only checking for synch immediately
1706 before acknowledging the chunk.
1708 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1709 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1710 macro. Patches provided by Josh Soref.
1712 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1713 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1714 by IANA as a service extension. However, John Klensin suggests that we
1717 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1718 the dkim code may be unix-mode line endings rather than smtp wire-format
1719 CRLF, so prepend a CR to any bare LF.
1721 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1722 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1724 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1725 the first were themselves being wrongly included in the feed into dkim
1726 processing; with most chunk sizes in use this resulted in an incorrect
1727 body hash calculated value.
1729 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1730 DKIM signature block, for verification. Although advised against by
1731 standards it is specifically not ruled illegal.
1733 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1735 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1736 missing a body hash (the bh= tag).
1738 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1739 It seems that HAProxy sends the Proxy Protocol information in clear and
1740 only then does a TLS startup, so do the same.
1742 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1743 TCP connections (such as for Spamd) unless the daemon successfully set
1744 Fast Open mode on its listening sockets. This fixes breakage seen on
1745 too-old kernels or those not configured for Fast Open, at the cost of
1746 requiring both directions being enabled for TFO, and TFO never being used
1747 by non-daemon-related Exim processes.
1749 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1750 endings, at least on the first header line. Try to canonify any that get
1751 past that check, despite the cost.
1753 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1754 now limited to an arbitrary five deep, while parsing addresses with the
1755 strip_excess_angle_brackets option enabled.
1757 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1758 instead leave the unprompted TLS handshake in socket buffer for the
1759 TLS library to consume.
1761 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1763 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1765 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1766 region in one operation, for speed, and this leaves assigned pointers
1767 dangling. Add checks run only under the testsuite which checks all
1768 variables at a store-reset and panics on a dangling pointer; add code
1769 explicitly nulling out all the variables discovered. Fixes one known
1770 bug: a transport crash, where a dangling pointer for $sending_ip_address
1771 originally assigned in a verify callout, is re-used.
1773 PP/06 Drop '.' from @INC in various Perl scripts.
1775 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1777 PP/08 Reduce a number of compilation warnings under clang; building with
1778 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1779 should be warning-free.
1781 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1783 HS/01 Fix portability problems introduced by PP/08 for platforms where
1784 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1786 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1787 chunk. This allows us to accept broken chunked messages. We need a more
1788 general solution here.
1790 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1791 already-broken messages in the queue.
1793 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1795 JH/19 Fix reference counting bug in routing-generated-address tracking.
1801 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1802 supports it and a size is available (ie. the sending peer gave us one).
1804 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1805 years of being deprecated). The replacements are the ACLs
1806 acl_smtp_mime and acl_not_smtp_mime.
1808 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1809 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1810 or even an in-clear connection were permitted. Now, if the host lookup
1811 was dnssec and dane was requested then the host is only used if the
1812 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1813 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1814 if one fails this test.
1815 This means that a poorly-configured remote DNS will make it incommunicado;
1816 but it protects against a DNS-interception attack on it.
1818 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1819 non-noisy when a race steals the message being considered.
1821 JH/05 If main configuration option tls_certificate is unset, generate a
1822 self-signed certificate for inbound TLS connections.
1824 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1825 in rewrites and routers.
1827 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1828 and logged a warning sing 4.83; now they are a configuration file error.
1830 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1831 (lacking @domain). Apply the same qualification processing as RCPT.
1833 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1835 JH/10 Support ${sha256:} applied to a string (as well as the previous
1838 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1839 a cutthrough deliver is pending, as we always want to make a connection.
1840 This also avoids re-routing the message when later placing the cutthrough
1841 connection after a verify cache hit.
1842 Do not update it with the verify result either.
1844 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1845 when routing results in more than one destination address.
1847 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1848 signing (which inhibits the cutthrough capability). Previously only
1849 the presence of an option was tested; now an expansion evaluating as
1850 empty is permissible (obviously it should depend only on data available
1851 when the cutthrough connection is made).
1853 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1854 the relevant preceding SMTP command did not note the pipelining mode.
1856 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1857 Previously they were not counted.
1859 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1860 as one having no matching records. Previously we deferred the message
1861 that needed the lookup.
1863 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1864 distinguished as "(=".
1866 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1867 for missing MX records. Previously it only worked for missing A records.
1869 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1871 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1872 after the data-go-ahead and data-ack. Patch from Jason Betts.
1874 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1875 even for a "none" policy. Patch from Tony Meyer.
1877 JH/22 Fix continued use of a connection for further deliveries. If a port was
1878 specified by a router, it must also match for the delivery to be
1881 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1882 When one of the recipients of a message was unsuitable for the connection
1883 (has no matching addresses), we lost track of needing to mark it
1884 deferred. As a result mail would be lost.
1886 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1888 JH/25 Decoding ACL controls is now done using a binary search; the source code
1889 takes up less space and should be simpler to maintain. Merge the ACL
1890 condition decode tables also, with similar effect.
1892 JH/26 Fix problem with one_time used on a redirect router which returned the
1893 parent address unchanged. A retry would see the parent address marked as
1894 delivered, so not attempt the (identical) child. As a result mail would
1897 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1898 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1899 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1902 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1903 Main config options check_{log,spool}_{inodes,space} are now
1904 100 inodes, 10MB unless set otherwise in the configuration.
1906 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1907 Previously it only applied to the main-section connection policy
1910 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1912 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1913 by me. Added RFC7919 DH primes as an alternative.
1915 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1916 are not in the system include path.
1918 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1919 GnuTLS, when a session startup failed (eg because the client disconnected)
1920 Exim did stdio operations after fclose. This was exposed by a recent
1921 change which nulled out the file handle after the fclose.
1923 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1924 signed directly by the cert-signing cert, rather than an intermediate
1925 OCSP-signing cert. This is the model used by LetsEncrypt.
1927 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1929 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1930 an incoming connection.
1932 HS/02 Bug 1802: Do not half-close the connection after sending a request
1935 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1936 fallback to "prime256v1".
1938 JH/34 SECURITY: Use proper copy of DATA command in error message.
1939 Could leak key material. Remotely exploitable. CVE-2016-9963.
1945 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1946 and 3.4.4 - once the server is enabled to respond to an OCSP request
1947 it does even when not requested, resulting in a stapling non-aware
1948 client dropping the TLS connection.
1950 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1951 support variable-length bit vectors. No functional change.
1953 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1954 The I= interface field on outgoing lines is now after the H= remote
1955 host field, same as incoming lines. There is a separate
1956 outgoing_interface log selector which allows you to disable the
1959 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1960 If not running log_selector +smtp_connection the mainlog would be held
1961 open indefinitely after a "too many connections" event, including to a
1962 deleted file after a log rotate. Leave the per net connection logging
1963 leaving it open for efficiency as that will be quickly detected by the
1964 check on the next write.
1966 HS/01 Bug 1671: Fix post transport crash.
1967 Processing the wait-<transport> messages could crash the delivery
1968 process if the message IDs didn't exist for some reason. When
1969 using 'split_spool_directory=yes' the construction of the spool
1970 file name failed already, exposing the same netto behaviour.
1972 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1973 mime_regex ACL conditions.
1975 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1976 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1977 to HELO, local diagnostic string.
1979 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1980 log line to debug. Even when configured with a tls authenticator many
1981 client connections are expected to not authenticate in this way, so
1982 an authenticate fail is not an error.
1984 HS/02 Add the Exim version string to the process info. This way exiwhat
1985 gives some more detail about the running daemon.
1987 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1988 matter for fast-change records such as DNSBLs.
1990 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1991 retry record, even if constant. There may be multiple transports with
1992 different interface settings and the retry behaviour needs to be kept
1995 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1997 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1999 JH/10 Bug 840: fix log_defer_output option of pipe transport
2001 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2002 hosts_randomize. This matters a lot when combined with mua_wrapper.
2004 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2005 ${quote_pgsql:<string>} operator.
2007 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2009 JH/14 Change header file name for freeradius-client. Relevant if compiling
2010 with Radius support; from the Gentoo tree and checked under Fedora.
2012 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2014 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2015 option transport_filter, meaning no filtering.
2017 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2019 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2020 defaults to "*" (all hosts). The variable is now available when not built
2021 with TLS, default unset, mainly to enable keeping the testsuite sane.
2022 If a server certificate is not supplied (via tls_certificate) an error is
2023 logged, and clients will find TLS connections fail on startup. Presumably
2024 they will retry in-clear.
2025 Packagers of Exim are strongly encouraged to create a server certificate
2026 at installation time.
2028 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2029 with the $config_file variable.
2031 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2032 in transport context, after the attempt, and per-recipient. The latter type
2033 is per host attempted. The event data is the error message, and the errno
2034 information encodes the lookup type (A vs. MX) used for the (first) host,
2035 and the trailing two digits of the smtp 4xx response.
2037 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2038 to write to mainlog (or rejectlog, paniclog) in the window between file
2039 creation and permissions/ownership being changed. Particularly affects
2040 installations where exicyclog is run as root, rather than exim user;
2041 result is that the running daemon panics and dies.
2043 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2045 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2046 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2047 "pri" and "weight". Note that the previous implicit priority given by the
2048 list order is no longer honoured.
2050 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2051 for DKIM processing.
2053 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2054 by defining SUPPORT_SOCKS.
2056 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2057 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2058 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2059 variables are renamed to proxy_{local,external}_{address,port}.
2061 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2062 for a build by defining SUPPORT_I18N
2064 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2065 of the query string, and make ${quote_redis:} do that quoting.
2067 JH/29 Move Events support from Experimental to mainline, enabled by default
2068 and removable for a build by defining DISABLE_EVENT.
2070 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2072 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2073 cached by the daemon.
2075 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2076 by defining LOOKUP_REDIS. The libhiredis library is required.
2078 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2079 keys are given for lookup.
2081 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2082 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2083 only supported when built with TLS support. The PolarSSL SHA routines
2084 are still used when the TLS library is too old for convenient support.
2086 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2087 openssl_options), for security. OpenSSL forces this from version 1.1.0
2088 server-side so match that on older versions.
2090 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2091 allocation for $value could be released as the expansion processing
2092 concluded, but leaving the global pointer active for it.
2094 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2095 and to use the domains and local_parts ACL conditions.
2097 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2098 incorrectly not doubled on cutthrough transmission, hence seen as a
2099 body-termination at the receiving system - resulting in truncated mails.
2100 Commonly the sender saw a TCP-level error, and retransmitted the message
2101 via the normal store-and-forward channel. This could result in duplicates
2102 received - but deduplicating mailstores were liable to retain only the
2103 initial truncated version.
2105 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2107 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2109 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2110 we're in there, support oversigning also; bug 1309.
2112 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2114 HS/04 Add support for keep_environment and add_environment options.
2116 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2117 either intentional arithmetic overflow during PRNG, or testing config-
2120 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2121 delivery resulted in actual delivery. Cancel cutthrough before DATA
2124 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2125 rejects a recipient: pass the reject to the originator.
2127 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2128 Many were false-positives and ignorable, but it's worth fixing the
2131 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2132 for the new environment-manipulation done at startup. Move the routines
2133 from being local to tls.c to being global via the os.c file.
2135 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2136 an extract embedded as result-arg for a map, the first arg for extract
2137 is unavailable so we cannot tell if this is a numbered or keyed
2138 extraction. Accept either.
2144 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2147 JH/02 The smtp transport option "multi_domain" is now expanded.
2149 JH/03 The smtp transport now requests PRDR by default, if the server offers
2152 JH/04 Certificate name checking on server certificates, when exim is a client,
2153 is now done by default. The transport option tls_verify_cert_hostnames
2154 can be used to disable this per-host. The build option
2155 EXPERIMENTAL_CERTNAMES is withdrawn.
2157 JH/05 The value of the tls_verify_certificates smtp transport and main options
2158 default to the word "system" to access the system default CA bundle.
2159 For GnuTLS, only version 3.0.20 or later.
2161 JH/06 Verification of the server certificate for a TLS connection is now tried
2162 (but not required) by default. The verification status is now logged by
2163 default, for both outbound TLS and client-certificate supplying inbound
2166 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2169 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2170 Status Notification (bounce) messages are now MIME format per RFC 3464.
2171 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2172 under the control of the dsn_advertise_hosts option, and routers may
2173 have a dsn_lasthop option.
2175 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2176 default, modifiable by a malware= option. The list separator for
2177 the options can now be changed in the usual way. Bug 68.
2179 JH/10 The smtp_receive_timeout main option is now expanded before use.
2181 JH/11 The incoming_interface log option now also enables logging of the
2182 local interface on delivery outgoing connections.
2184 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2185 if the interface and destination host and port all match.
2187 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2190 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2191 Patch from Andrew Lewis.
2193 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2194 now supports optional time-restrictions, weighting, and priority
2195 modifiers per server. Patch originally by <rommer@active.by>.
2197 JH/16 The spamd_address main option now supports a mixed list of local
2198 and remote servers. Remote servers can be IPv6 addresses, and
2199 specify a port-range.
2201 JH/17 Bug 68: The spamd_address main option now supports an optional
2202 timeout value per server.
2204 JH/18 Bug 1581: Router and transport options headers_add/remove can
2205 now have the list separator specified.
2207 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2210 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2213 JH/21 Support for the A6 type of dns record is withdrawn.
2215 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2216 rather than the verbs used.
2218 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2219 from 255 to 1024 chars.
2221 JH/24 Verification callouts now attempt to use TLS by default.
2223 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2224 are generic router options now. The defaults didn't change.
2226 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2227 Original patch from Alexander Shikoff, worked over by JH.
2229 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2230 files if ZCAT_COMMAND is not executable.
2232 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2234 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2236 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2237 Normally benign, it bites when the pair was led to by a CNAME;
2238 modern usage is to not canonicalize the domain to a CNAME target
2239 (and we were inconsistent anyway for A-only vs AAAA+A).
2241 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2243 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2244 when evaluating $sender_host_dnssec.
2246 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2247 $sender_helo_dnssec variable.
2249 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2251 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2253 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2255 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2256 documented as working, but never had. Support all but $spam_report.
2258 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2259 added for tls authenticator.
2261 HS/03 Add perl_taintmode main config option
2267 TL/01 When running the test suite, the README says that variables such as
2268 no_msglog_check are global and can be placed anywhere in a specific
2269 test's script, however it was observed that placement needed to be near
2270 the beginning for it to behave that way. Changed the runtest perl
2271 script to read through the entire script once to detect and set these
2272 variables, reset to the beginning of the script, and then run through
2273 the script parsing/test process like normal.
2275 TL/02 The BSD's have an arc4random API. One of the functions to induce
2276 adding randomness was arc4random_stir(), but it has been removed in
2277 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2278 function when detected.
2280 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2281 cause callback expansion.
2283 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2284 syntax errors in an expansion can be treated as a string instead of
2285 logging or causing an error, due to the internal use of bool_lax
2286 instead of bool when processing it.
2288 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2289 server certificates when making smtp deliveries.
2291 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2293 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2295 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2297 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2298 Merged patch from Sebastian Wiedenroth.
2300 JH/05 Fix results-pipe from transport process. Several recipients, combined
2301 with certificate use, exposed issues where response data items split
2302 over buffer boundaries were not parsed properly. This eventually
2303 resulted in duplicates being sent. This issue only became common enough
2304 to notice due to the introduction of connection certificate information,
2305 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2307 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2308 size buffer was used, resulting in syntax errors when an expansion
2311 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2312 version 3.3.6 or later.
2314 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2315 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2316 both become "event_action", the variables become $event_name, $event_data
2317 and $event_defer_errno. There is a new variable $verify_mode, usable in
2318 routers, transports and related events. The tls:cert event is now also
2319 raised for inbound connections, if the main configuration event_action
2322 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2323 early OCSP support, but no stapling (appears to be less than 1.0.0).
2325 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2326 server certificate names available under the smtp transport option
2327 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2330 JH/10 Time-related extraction expansions from certificates now use the main
2331 option "timezone" setting for output formatting, and are consistent
2332 between OpenSSL and GnuTLS compilations. Bug 1541.
2334 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2335 encoded parameter in the incoming message. Bug 1558.
2337 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2338 include certificate info, eximon was claiming there were spoolfile
2341 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2343 JH/14 Log delivery-related information more consistently, using the sequence
2344 "H=<name> [<ip>]" wherever possible.
2346 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2347 are problematic for Debian distribution, omit them from the release
2350 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2352 JH/16 Fix string representation of time values on 64bit time_t architectures.
2355 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2356 output list separator was used.
2361 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2362 checkers that were complaining about end of non-void function with no
2365 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2366 This was a regression introduced in 4.83 by another bugfix.
2368 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2370 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2371 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2377 TF/01 Correctly close the server side of TLS when forking for delivery.
2379 When a message was received over SMTP with TLS, Exim failed to clear up
2380 the incoming connection properly after forking off the child process to
2381 deliver the message. In some situations the subsequent outgoing
2382 delivery connection happened to have the same fd number as the incoming
2383 connection previously had. Exim would try to use TLS and fail, logging
2384 a "Bad file descriptor" error.
2386 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2387 utilities have not been installed.
2389 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2390 temporary space as the ACL may create new global variables.
2392 TL/01 LDAP support uses per connection or global context settings, depending
2393 upon the detected version of the libraries at build time.
2395 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2396 to extract and use the src ip:port in logging and expansions as if it
2397 were a direct connection from the outside internet. PPv2 support was
2398 updated based on HAProxy spec change in May 2014.
2400 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2402 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2403 Properly escape header and check for NULL return.
2405 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2408 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2410 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2411 characters in header names, implemented as a verify condition.
2412 Contributed by Michael Fischer v. Mollard.
2414 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2415 results permerror and temperror. Previous values are deprecated but
2416 still accepted. In a future release, err_perm and err_temp will be
2417 completely removed, which will be a backward incompatibility if the
2418 ACL tests for either of these two old results. Patch contributed by
2419 user bes-internal on the mailing list.
2421 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2423 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2424 selectors, in both main and reject logs.
2426 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2429 JH/07 Add malware type "sock" for talking to simple daemon.
2431 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2433 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2434 routers/transports under cutthrough routing.
2436 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2437 numbers. Touch up "bool" conditional to keep the same definition.
2439 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2441 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2443 JH/12 Expand items in router/transport headers_add or headers_remove lists
2444 individually rather than the list as a whole. Bug 1452.
2446 Required for reasonable handling of multiple headers_ options when
2447 they may be empty; requires that headers_remove items with embedded
2448 colons must have them doubled (or the list-separator changed).
2450 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2451 view the policy declared in the DMARC record. Currently, $dmarc_status
2452 is a combined value of both the record presence and the result of the
2455 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2457 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2458 dnslookup router and the smtp transport (applying to the forward
2461 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2462 of ldap servers used for a specific lookup. Patch provided by Heiko
2465 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2466 New variable $lookup_dnssec_authenticated for observability.
2468 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2469 Patch submitted by Lars Timman.
2471 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2473 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2474 Requires trusted mode and valid format message id, aborts otherwise.
2475 Patch contributed by Heiko Schlichting.
2477 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2478 certextract with support for various fields. Bug 1358.
2480 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2481 is requested by default, modifiable by smtp transport option
2484 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2485 operate on certificate variables to give certificate fingerprints
2486 Also new ${sha256:cert_variable}.
2488 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2490 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2491 Christian Aistleitner.
2493 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2495 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2496 file. Patch from Wolfgang Breyha.
2498 JH/25 Expand the coverage of the delivery $host and $host_address to
2499 client authenticators run in verify callout. Bug 1476.
2501 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2502 align with daemon_smtp_ports. Bug 72.
2504 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2505 support and error reporting did not work properly.
2507 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2508 and is readable. Patch from Andrew Colin Kissa.
2510 TL/14 Enhance documentation of ${run expansion and how it parses the
2511 commandline after expansion, particularly in the case when an
2512 unquoted variable expansion results in an empty value.
2514 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2516 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2517 by renaming away from it. Addresses GH issue 15, reported by
2520 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2522 TL/15 SECURITY: prevent double expansion in math comparison functions
2523 (can expand unsanitized data). Not remotely exploitable.
2530 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2532 PP/02 Make -n do something, by making it not do something.
2533 When combined with -bP, the name of an option is not output.
2535 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2538 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2539 $sender_host_name and config options to manage this, and basic check
2542 PP/05 DSCP support for outbound connections and control modifier for inbound.
2544 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2545 (Only plugin which currently uses this is kerberos4, which nobody should
2546 be using, but we should make it available and other future plugins might
2547 conceivably use it, even though it would break NAT; stuff *should* be
2548 using channel bindings instead).
2550 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2551 name; added for Sendmail compatibility; requires admin caller.
2552 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2553 just ignore it); requires trusted caller.
2554 Also parse but ignore: -Ac -Am -X<logfile>
2557 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2559 TL/02 Add +smtp_confirmation as a default logging option.
2561 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2562 Patch by Magnus Holmgren from 2007-02-20.
2564 TL/04 Bugzilla 1281 - Spec typo.
2565 Bugzilla 1283 - Spec typo.
2566 Bugzilla 1290 - Spec grammar fixes.
2568 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2570 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2572 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2573 mailing list by by Dmitry Isaikin.
2575 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2577 JH/02 Support "G" suffix to numbers in ${if comparisons.
2579 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2581 NM/01 Bugzilla 1197 - Spec typo
2582 Bugzilla 1196 - Spec examples corrections
2584 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2586 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2587 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2590 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2591 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2593 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2594 "acl {{name}{arg}...}", and optional args on acl condition
2597 JH/05 Permit multiple router/transport headers_add/remove lines.
2599 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2601 JH/07 Avoid using a waiting database for a single-message-only transport.
2602 Performance patch from Paul Fisher. Bugzilla 1262.
2604 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2607 JH/09 Add $headers_added variable, with content from use of ACL modifier
2608 add_header (but not yet added to the message). Bugzilla 199.
2610 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2611 Pulled from Bugzilla 817 by Wolfgang Breyha.
2613 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2615 (nb: this is the same fix as in Exim 4.80.1)
2617 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2620 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2623 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2625 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2626 advertise SMTP AUTH mechanism to us, instead of a generic
2627 protocol violation error. Also, make Exim more robust to bad
2628 data from the Dovecot auth socket.
2630 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2632 When a queue runner is handling a message, Exim first routes the
2633 recipient addresses, during which it prunes them based on the retry
2634 hints database. After that it attempts to deliver the message to
2635 any remaining recipients. It then updates the hints database using
2638 So if a recipient address works intermittently, it can get repeatedly
2639 deferred at routing time. The retry hints record remains fresh so the
2640 address never reaches the final cutoff time.
2642 This is a fairly common occurrence when a user is bumping up against
2643 their storage quota. Exim had some logic in its local delivery code
2644 to deal with this. However it did not apply to per-recipient defers
2645 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2647 This change adds a proper retry rule check during routing so that the
2648 final cutoff time is checked against the message's age. We only do
2649 this check if there is an address retry record and there is not a
2650 domain retry record; this implies that previous attempts to handle
2651 the address had the retry_use_local_parts option turned on. We use
2652 this as an approximation for the destination being like a local
2653 delivery, as in LMTP.
2655 I suspect this new check makes the old local delivery cutoff check
2656 redundant, but I have not verified this so I left the code in place.
2658 TF/02 Correct gecos expansion when From: is a prefix of the username.
2660 Test 0254 submits a message to Exim with the header
2664 When I ran the test suite under the user fanf2, Exim expanded
2665 the header to contain my full name, whereas it should have added
2666 a Resent-Sender: header. It erroneously treats any prefix of the
2667 username as equal to the username.
2669 This change corrects that bug.
2671 GF/01 DCC debug and logging tidyup
2672 Error conditions log to paniclog rather than rejectlog.
2673 Debug lines prefixed by "DCC: " to remove any ambiguity.
2675 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2677 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2678 Bug spotted by Jeremy Harris; was flawed since initial commit.
2679 Would have resulted in OCSP responses post-SNI triggering an Exim
2680 NULL dereference and crash.
2682 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2684 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2685 Bug detection, analysis and fix by Samuel Thibault.
2686 Bugzilla 1331, Debian bug #698092.
2688 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2690 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2691 Server implementation by Todd Lyons, client by JH.
2692 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2693 config variable "prdr_enable" controls whether the server
2694 advertises the facility. If the client requests PRDR a new
2695 acl_data_smtp_prdr ACL is called once for each recipient, after
2696 the body content is received and before the acl_smtp_data ACL.
2697 The client is controlled by both of: a hosts_try_prdr option
2698 on the smtp transport, and the server advertisement.
2699 Default client logging of deliveries and rejections involving
2700 PRDR are flagged with the string "PRDR".
2702 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2703 fclose(). Diagnosis by Todd Lyons.
2705 PP/17 Update configure.default to handle IPv6 localhost better.
2706 Patch by Alain Williams (plus minor tweaks).
2709 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2710 This is now consistent with GnuTLS, and is now documented: the
2711 previous undocumented portable approach to treating the option as
2712 unset was to force an expansion failure. That still works, and
2713 an empty string is now equivalent.
2715 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2716 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2717 not performing validation itself.
2719 PP/20 Added force_command boolean option to pipe transport.
2720 Patch from Nick Koston, of cPanel Inc.
2722 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2725 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2727 PP/21 Fix eximon continuous updating with timestamped log-files.
2728 Broken in a format-string cleanup in 4.80, missed when I repaired the
2729 other false fix of the same issue.
2730 Report and fix from Heiko Schlichting.
2733 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2734 Report from Prashanth Katuri.
2736 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2737 It's SecureTransport, so affects any MacOS clients which use the
2738 system-integrated TLS libraries, including email clients.
2740 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2741 using a MIME ACL for non-SMTP local injection.
2742 Report and assistance in diagnosis by Warren Baker.
2744 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2746 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2748 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2749 last id that failed so it may be referenced in subsequent ACL's.
2751 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2754 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2755 ldap library initialization, allowing self-signed CA's to be
2756 used. Also properly sets require_cert option later in code by
2757 using NULL (global ldap config) instead of ldap handle (per
2758 session). Bug diagnosis and testing by alxgomz.
2760 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2761 the src/util/ subdirectory.
2763 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2764 renamed to Transport Post Delivery Action by Jeremy Harris, as
2767 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2768 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2769 redis_servers = needs to be configured which will be used by the redis
2770 lookup. Patch from Warren Baker, of The Packet Hub.
2772 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2774 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2775 hostname or reverse DNS when processing a host list. Used suggestions
2776 from multiple comments on this bug.
2778 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2780 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2781 Missed a few lines, added it to make the runtest require no keyboard
2784 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2785 contains upper case chars. Make router use caseful_local_part.
2787 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2788 support when GnuTLS has been built with p11-kit.
2794 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2796 This, or similar/improved, will also be change PP/11 of 4.82.
2802 PP/01 Handle short writes when writing local log-files.
2803 In practice, only affects FreeBSD (8 onwards).
2804 Bugzilla 1053, with thanks to Dmitry Isaikin.
2806 NM/01 Bugzilla 949 - Documentation tweak
2808 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2811 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2813 PP/02 Implemented gsasl authenticator.
2815 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2817 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2818 `pkg-config foo` for cflags/libs.
2820 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2821 with rest of GSASL and with heimdal_gssapi.
2823 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2824 `pkg-config foo` for cflags/libs for the TLS implementation.
2826 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2827 properties get this fed in as external SSF. A number of robustness
2828 and debugging improvements to the cyrus_sasl authenticator.
2830 PP/08 cyrus_sasl server now expands the server_realm option.
2832 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2833 Patch by Jeremy Harris.
2835 PP/10 Added dbmjz lookup type.
2837 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2839 PP/12 MAIL args handles TAB as well as SP, for better interop with
2840 non-compliant senders.
2841 Analysis and variant patch by Todd Lyons.
2843 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2844 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2845 Patch from Dirk Mueller <dmueller@suse.com>
2847 PP/13 tls_peerdn now print-escaped for spool files.
2848 Observed some $tls_peerdn in wild which contained \n, which resulted
2849 in spool file corruption.
2851 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2852 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2853 or write after TLS renegotiation, which otherwise led to messages
2856 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2857 as a tracking header (ie: a signed header comes before the signature).
2858 Patch from Wolfgang Breyha.
2860 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2861 comma-sep list; embedded commas doubled.
2863 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2865 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2867 Report and patch from Dmitry Banschikov.
2869 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2870 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2871 failures. We appear to now support TLS1.1+ with Exim.
2873 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2874 lets Exim select keys and certificates based upon TLS SNI from client.
2875 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2876 before an outbound SMTP session. New log_selector, +tls_sni.
2878 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2879 NULL dereference. Report and patch from Alun Jones.
2881 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2882 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2883 Not seeing resolver debug output on NetBSD, but suspect this is a
2884 resolver implementation change.
2886 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2887 Left warnings. Added "eximon gdb" invocation mode.
2889 PP/21 Defaulting "accept_8bitmime" to true, not false.
2891 PP/22 Added -bw for inetd wait mode support.
2893 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2894 locate the relevant includes and libraries. Made this the default.
2896 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2897 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2899 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2900 This may cause build issues on older platforms.
2902 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2903 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2904 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2905 Added SNI support via GnuTLS too.
2906 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2908 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2910 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2911 Applied second patch from Janne, implementing suggestion to default
2912 multiple-strings-in-record handling to match SPF spec.
2914 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2916 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2917 read-only, out of scope).
2918 Patch from Wolfgang Breyha, report from Stuart Northfield.
2920 PP/29 Fix three issues highlighted by clang analyser static analysis.
2921 Only crash-plausible issue would require the Cambridge-specific
2922 iplookup router and a misconfiguration.
2923 Report from Marcin Mirosław.
2925 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2927 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2928 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2929 As part of this, removing so much warning spew let me fix some minor
2930 real issues in debug logging.
2932 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2933 assignment on my part. Fixed.
2935 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2936 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2937 Janne Snabb (who went above and beyond: thank you).
2939 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2940 string otherwise requires a connection and a bunch more work and it's
2941 relatively easy to get wrong. Should also expose TLS library linkage
2944 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2945 64-bit ${eval} (JH/03).
2947 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2948 GNU libc to support some of the 64-bit stuff, should not lead to
2949 conflicts. Defined before os.h is pulled in, so if a given platform
2950 needs to override this, it can.
2952 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2953 protection layer was required, which is not implemented.
2954 Bugzilla 1254, patch from Wolfgang Breyha.
2956 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2957 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2958 tls_dhparam take prime identifiers. Also unbreak combination of
2959 OpenSSL+DH_params+TLSSNI.
2961 PP/39 Disable SSLv2 by default in OpenSSL support.
2967 PP/01 Solaris build fix for Oracle's LDAP libraries.
2968 Bugzilla 1109, patch from Stephen Usher.
2970 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2972 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2975 TF/02 Fix a couple more cases where we did not log the error message
2976 when unlink() failed. See also change 4.74-TF/03.
2978 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2979 lock up or crash if it happened to be inside a call to libc when it
2980 got a SIGUSR1 from exiwhat.
2982 The SIGUSR1 handler appends the current process status to the process
2983 log which is later printed by exiwhat. It used to use the general
2984 purpose logging code to do this, but several functions it calls are
2985 not safe for signals.
2987 The new output code in the SIGUSR1 handler is specific to the process
2988 log, and simple enough that it's easy to inspect for signal safety.
2989 Removing some special cases also simplifies the general logging code.
2990 Removing the spurious timestamps from the process log simplifies
2993 TF/04 Improved ratelimit ACL condition.
2995 The /noupdate option has been deprecated in favour of /readonly which
2996 has clearer semantics. The /leaky, /strict, and /readonly update modes
2997 are mutually exclusive. The update mode is no longer included in the
2998 database key; it just determines when the database is updated. (This
2999 means that when you upgrade Exim will forget old rate measurements.)
3001 Exim now checks that the per_* options are used with an update mode that
3002 makes sense for the current ACL. For example, when Exim is processing a
3003 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3004 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3005 must specify per_mail/readonly. If you omit the update mode it defaults to
3006 /leaky where that makes sense (as before) or /readonly where required.
3008 The /noupdate option is now undocumented but still supported for
3009 backwards compatibility. It is equivalent to /readonly except that in
3010 ACLs where /readonly is required you may specify /leaky/noupdate or
3011 /strict/noupdate which are treated the same as /readonly.
3013 A useful new feature is the /count= option. This is a generalization
3014 of the per_byte option, so that you can measure the throughput of other
3015 aggregate values. For example, the per_byte option is now equivalent
3016 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3018 The per_rcpt option has been generalized using the /count= mechanism
3019 (though it's more complicated than the per_byte equivalence). When it is
3020 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3021 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3022 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3023 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3024 Note that using per_rcpt with a non-readonly update mode in more than
3025 one ACL will cause the recipients to be double-counted. (The per_mail
3026 and per_byte options don't have this problem.)
3028 The handling of very low rates has changed slightly. If the computed rate
3029 is less than the event's count (usually one) then this event is the first
3030 after a long gap. In this case the rate is set to the same as this event's
3031 count, so that the first message of a spam run is counted properly.
3033 The major new feature is a mechanism for counting the rate of unique
3034 events. The new per_addr option counts the number of different
3035 recipients that someone has sent messages to in the last time period. It
3036 behaves like per_rcpt if all the recipient addresses are different, but
3037 duplicate recipient addresses do not increase the measured rate. Like
3038 the /count= option this is a general mechanism, so the per_addr option
3039 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3040 example, measure the rate that a client uses different sender addresses
3041 with the options per_mail/unique=$sender_address. There are further
3042 details in the main documentation.
3044 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3046 TF/06 Removed a few PCRE remnants.
3048 TF/07 Automatically extract Exim's version number from tags in the git
3049 repository when doing development or release builds.
3051 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3052 Bugzilla 879. Patch from Paul Fisher.
3054 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3055 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3058 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3060 PP/05 Variable $av_failed, true if the AV scanner deferred.
3061 Bugzilla 1078. Patch from John Horne.
3063 PP/06 Stop make process more reliably on build failure.
3064 Bugzilla 1087. Patch from Heiko Schlittermann.
3066 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3067 Bugzilla 1089. Patch from Heiko Schlittermann.
3069 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3070 Bugzilla 1131. Patch from Holger Weiß.
3072 PP/09 Handle IPv6 addresses with SPF.
3073 Bugzilla 860. Patch from Wolfgang Breyha.
3075 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3077 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3080 PP/11 match_* no longer expand right-hand-side by default.
3081 New compile-time build option, EXPAND_LISTMATCH_RHS.
3082 New expansion conditions, "inlist", "inlisti".
3084 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3086 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3088 PP/14 fix log_write() format string regression from TF/03.
3089 Bugzilla 1152. Patch from Dmitry Isaikin.
3095 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3097 PP/02 Harmonised TLS library version reporting; only show if debugging.
3098 Layout now matches that introduced for other libraries in 4.74 PP/03.
3100 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3102 PP/04 New "dns_use_edns0" global option.
3104 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3107 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3108 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3110 TK/01 Updated PolarSSL code to 0.14.2.
3111 Bugzilla 1097. Patch from Andreas Metzler.
3113 PP/07 Catch divide-by-zero in ${eval:...}.
3114 Fixes bugzilla 1102.
3116 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3119 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3120 format-string attack -- SECURITY: remote arbitrary code execution.
3122 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3123 time unintentionally subject to list matching rules, letting the header
3124 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3125 arbitrary string expansion). This allowed for information disclosure.
3127 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3128 INT_MIN/-1 -- value coerced to INT_MAX.
3134 NM/01 Workaround for PCRE version dependency in version reporting
3137 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3138 This fixes portability to compilers other than gcc, notably
3139 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3141 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3142 makefiles for portability to HP-UX and POSIX correctness.
3144 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3145 Also via indented variable definition in the Makefile.
3146 (Debugging by Oliver Heesakkers).
3148 PP/02 Restore caching of spamd results with expanded spamd_address.
3149 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3151 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3152 Improves build reliability. Fix from: Frank Elsner
3154 NM/02 Fix wide character breakage in the rfc2047 coding
3155 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3157 NM/03 Allow underscore in dnslist lookups
3158 Fixes bug 1026. Patch from Graeme Fowler
3160 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3161 Code patches from Adam Ciarcinski of NetBSD.
3163 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3166 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3167 is logged, to avoid truncation. Patch from John Horne.
3169 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3170 Patch from Jakob Hirsch.
3172 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3173 SQL string expansion failure details.
3174 Patch from Andrey Oktyabrski.
3176 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3177 Patch from Simon Arlott.
3179 PP/09 New lookups functionality failed to compile on old gcc which rejects
3180 extern declarations in function scope.
3181 Patch from Oliver Fleischmann
3183 PP/10 Use sig_atomic_t for flags set from signal handlers.
3184 Check getgroups() return and improve debugging.
3185 Fixed developed for diagnosis in bug 927 (which turned out to be
3188 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3189 Patch from Mark Zealey.
3191 PP/12 Bugzilla 1056: Improved spamd server selection.
3192 Patch from Mark Zealey.
3194 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3195 Based on patch from Heiko Schlittermann.
3197 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3198 Patch from Uwe Doering, sign-off by Michael Haardt.
3200 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3201 variable declaration deep within a block. Bug and patch from
3204 PP/15 lookups-Makefile IRIX compatibility coercion.
3206 PP/16 Make DISABLE_DKIM build knob functional.
3208 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3209 Patch by Simon Arlott
3211 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3212 variable argument macros. Our copy now differs from upstream.
3218 TF/01 Failure to get a lock on a hints database can have serious
3219 consequences so log it to the panic log.
3221 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3222 controlled using the smtp_confirmation log selector.
3224 TF/03 Include the error message when we fail to unlink a spool file.
3226 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3227 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3228 for maintaining out-of-tree patches for some time.
3230 PP/01 Bugzilla 139: Documentation and portability issues.
3231 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3232 Handle per-OS dynamic-module compilation flags.
3234 PP/02 Let /dev/null have normal permissions.
3235 The 4.73 fixes were a little too stringent and complained about the
3236 permissions on /dev/null. Exempt it from some checks.
3237 Reported by Andreas M. Kirchwitz.
3239 PP/03 Report version information for many libraries, including
3240 Exim version information for dynamically loaded libraries. Created
3241 version.h, now support a version extension string for distributors
3242 who patch heavily. Dynamic module ABI change.
3244 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3245 privilege escalation vulnerability whereby the Exim run-time user
3246 can cause root to append content of the attacker's choosing to
3249 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3252 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3253 If dropping privileges for untrusted macros, we disabled normal logging
3254 on the basis that it would fail; for the Exim run-time user, this is not
3255 the case, and it resulted in successful deliveries going unlogged.
3256 Fixed. Reported by Andreas Metzler.
3262 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3263 only prepend for the Resent-* case. Fixes regression introduced in
3264 Exim 4.70 by NM/22 for Bugzilla 607.
3266 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3267 increasing numbers of administrators be bitten by this.
3269 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3271 PP/03 Bugzilla 994: added openssl_options main configuration option.
3273 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3275 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3277 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3279 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3280 without a peer certificate, leading to a segfault because of an
3281 assumption that peers always have certificates. Be a little more
3282 paranoid. Problem reported by Martin Tscholak.
3284 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3285 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3286 NB: ClamAV planning to remove STREAM in "middle of 2010".
3287 CL also introduces -bmalware, various -d+acl logging additions and
3288 more caution in buffer sizes.
3290 PP/09 Implemented reverse_ip expansion operator.
3292 PP/10 Bugzilla 937: provide a "debug" ACL control.
3294 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3296 PP/12 Bugzilla 973: Implement --version.
3298 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3300 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3302 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3304 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3305 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3306 ignore trailing whitespace.
3308 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3310 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3313 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3314 Notification from Dr Andrew Aitchison.
3316 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3317 ExtendedDetectionInfo response format.
3318 Notification from John Horne.
3320 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3323 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3324 XSL and documented dependency on system catalogs, with examples of how
3327 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3330 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3331 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3332 configuration file which is writeable by the Exim user or group.
3334 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3335 of configuration files to cover files specified with the -C option if
3336 they are going to be used with root privileges, not just the default
3339 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3340 option (effectively making it always true).
3342 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3343 files to be used while preserving root privileges.
3345 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3346 that rogue child processes cannot use them.
3348 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3349 run-time user, instead of root.
3351 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3352 Exim run-time user without dropping privileges.
3354 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3355 result string, instead of calling string_vformat() twice with the same
3358 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3359 for other users. Others should always drop root privileges if they use
3360 -C on the command line, even for a whitelisted configure file.
3362 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3364 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3370 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3371 $data_path, and $header_path variables; fixed documentation bugs and
3374 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3375 exipick to access non-standard spools, including the "frozen" queue
3378 NM/01 Bugzilla 965: Support mysql stored procedures.
3379 Patch from Alain Williams
3381 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3383 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3384 Patch from Andreas Metzler
3386 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3387 Patch from Kirill Miazine
3389 NM/05 Bugzilla 671: Added umask to procmail example.
3391 JJ/03 installed exipick 20100323.0, fixing doc bug
3393 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3394 directory. Notification and patch from Dan Rosenberg.
3396 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3398 TK/02 Improve log output when DKIM signing operation fails.
3400 MH/01 Treat the transport option dkim_domain as a colon separated
3401 list, not as a single string, and sign the message with each element,
3402 omitting multiple occurences of the same signer.
3404 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3405 Bugzilla 985, 986. Patch by Simon Arlott
3407 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3408 Patch by Simon Arlott
3410 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3411 MBX locking. Notification from Dan Rosenberg.
3417 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3419 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3421 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3423 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3425 NM/04 Bugzilla 915: Flag broken perl installation during build.
3431 TK/01 Added patch by Johannes Berg that expands the main option
3432 "spamd_address" if it starts with a dollar sign.
3434 TK/02 Write list of recipients to X-Envelope-Sender header when building
3435 the mbox-format spool file for content scanning (suggested by Jakob
3438 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3439 (http://www.dcc-servers.net/) support via dccifd. Activated by
3440 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3442 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3443 by Mark Daniel Reidel <mr@df.eu>.
3445 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3446 When building exim an external PCRE library is now needed -
3447 PCRE is a system library on the majority of modern systems.
3448 See entry on PCRE_LIBS in EDITME file.
3450 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3451 conversation. Added nologin parameter to request.
3452 Patch contributed by Kirill Miazine.
3454 TF/01 Do not log submission mode rewrites if they do not change the address.
3456 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3458 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3459 log files in place. Contributed by Roberto Lima.
3461 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3463 TF/03 Bugzilla 615: When checking the local_parts router precondition
3464 after a local_part_suffix or local_part_prefix option, Exim now
3465 does not use the address's named list lookup cache, since this
3466 contains cached lookups for the whole local part.
3468 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3469 Robert Millan. Documentation is in experimental-spec.txt.
3471 TF/04 Bugzilla 668: Fix parallel build (make -j).
3473 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3475 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3476 Patch provided by Jan Srzednicki.
3478 TF/05 Leading white space used to be stripped from $spam_report which
3479 wrecked the formatting. Now it is preserved.
3481 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3482 that they are available at delivery time.
3484 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3486 TF/08 TLS error reporting now respects the incoming_interface and
3487 incoming_port log selectors.
3489 TF/09 Produce a more useful error message if an SMTP transport's hosts
3490 setting expands to an empty string.
3492 NM/06 Bugzilla 744: EXPN did not work under TLS.
3493 Patch provided by Phil Pennock.
3495 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3496 Patch provided by Richard Godbee.
3498 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3499 acl_smtp_notquit, added index entry.
3501 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3502 Patch provided by Eugene Bujak.
3504 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3505 accept(). Patch provided by Maxim Dounin.
3507 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3508 Patch provided by Phil Pennock.
3510 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3512 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3513 Patch provided by Brad "anomie" Jorsch.
3515 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3516 Patch provided by Dean Brooks.
3518 TK/05 Add native DKIM support (does not depend on external libraries).
3520 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3521 Patch provided by Graeme Fowler.
3523 NM/16 Bugzilla 851: Documentation example syntax fix.
3525 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3527 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3530 NM/19 Bugzilla 745: TLS version reporting.
3531 Patch provided by Phil Pennock.
3533 NM/20 Bugzilla 167: bool: condition support.
3534 Patch provided by Phil Pennock.
3536 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3537 clients. Patch provided by Phil Pennock.
3539 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3540 Patch provided by Brad "anomie" Jorsch.
3542 NM/23 Bugzilla 687: Fix misparses in eximstats.
3543 Patch provided by Heiko Schlittermann.
3545 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3546 Patch provided by Heiko Schlittermann.
3548 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3549 plus update to original patch.
3551 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3553 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3554 Patch provided by David Brownlee.
3556 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3558 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3560 NM/30 Bugzilla 888: TLS documentation bugfixes.
3562 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3564 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3565 Unlike the original bugzilla I have changed all shell scripts in src tree.
3567 NM/33 Bugzilla 898: Transport filter timeout fix.
3568 Patch by Todd Rinaldo.
3570 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3571 Patch by Serge Demonchaux.
3573 NM/35 Bugzilla 39: Base64 decode bug fixes.
3574 Patch by Jakob Hirsch.
3576 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3578 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3580 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3582 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3588 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3589 ALT-N's libdkim that I have put here:
3590 http://duncanthrax.net/exim-experimental/
3592 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3593 were called 'true' and it seems that C99 defines that as a reserved
3594 keyword to be used with 'bool' variable types. That means you could
3595 not include C99-style headers which use bools without triggering
3596 build errors in sieve.c.
3598 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3599 as mailq or other aliases. Changed the --help handling significantly
3600 to do whats expected. exim_usage() emits usage/help information.
3602 SC/01 Added the -bylocaldomain option to eximstats.
3604 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3606 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3608 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3614 PH/01 Another patch from the Sieve maintainer.
3616 PH/02 When an IPv6 address is converted to a string for single-key lookup
3617 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3618 dots are used instead of colons so that keys in lsearch files need not
3619 contain colons. This was done some time before quoting was made available
3620 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3621 (notated using the quote facility) so as to distinguish them from IPv4
3622 keys. This meant that lookups for IP addresses in host lists did not work
3623 for iplsearch lookups.
3625 This has been fixed by arranging for IPv6 addresses to be expressed with
3626 colons if the lookup type is iplsearch. This is not incompatible, because
3627 previously such lookups could never work.
3629 The situation is now rather anomalous, since one *can* have colons in
3630 ordinary lsearch keys. However, making the change in all cases is
3631 incompatible and would probably break a number of configurations.
3633 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3636 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3637 conversion specification without a maximum field width, thereby enabling
3638 a rogue spamd server to cause a buffer overflow. While nobody in their
3639 right mind would setup Exim to query an untrusted spamd server, an
3640 attacker that gains access to a server running spamd could potentially
3641 exploit this vulnerability to run arbitrary code as the Exim user.
3643 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3644 $primary_hostname instead of what libspf2 thinks the hosts name is.
3646 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3647 a directory entry by the name of the lookup key. Previously, if a
3648 symlink pointed to a non-existing file or a file in a directory that
3649 Exim lacked permissions to read, a lookup for a key matching that
3650 symlink would fail. Now it is enough that a matching directory entry
3651 exists, symlink or not. (Bugzilla 503.)
3653 PH/03 The body_linecount and body_zerocount variables are now exported in the
3656 PH/04 Added the $dnslist_matched variable.
3658 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3659 This means they are set thereafter only if the connection becomes
3662 PH/06 Added the client_condition to authenticators so that some can be skipped
3663 by clients under certain conditions.
3665 PH/07 The error message for a badly-placed control=no_multiline_responses left
3666 "_responses" off the end of the name.
3668 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3670 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3671 (without spaces) instead of just copying the configuration text.
3673 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3675 PH/11 Added $max_received_linelength.
3677 PH/12 Added +ignore_defer and +include_defer to host lists.
3679 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3680 way in which PCRE > 7.0 is built.
3682 PH/14 Implemented queue_only_load_latch.
3684 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3685 MAIL command. The effect was to mangle the value on 64-bit systems.
3687 PH/16 Another patch from the Sieve maintainer.
3689 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3691 PH/18 If a system quota error occurred while trying to create the file for
3692 a maildir delivery, the message "Mailbox is full" was not appended to the
3693 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3694 only to a quota excession during the actual writing of the file.
3696 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3697 characters?) which causes problems in log lines. The DN values are now
3698 passed through string_printing() before being added to log lines.
3700 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3701 and InterBase are left for another time.)
3703 PH/21 Added message_body_newlines option.
3705 PH/22 Guard against possible overflow in moan_check_errorcopy().
3707 PH/23 POSIX allows open() to be a macro; guard against that.
3709 PH/24 If the recipient of an error message contained an @ in the local part
3710 (suitably quoted, of course), incorrect values were put in $domain and
3711 $local_part during the evaluation of errors_copy.
3717 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3718 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3721 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3722 issue a MAIL command.
3724 PH/02 In an ACL statement such as
3726 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3728 if a client was not listed at all, or was listed with a value other than
3729 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3730 the condition was not true (as it should be), so access was not denied.
3731 The bug was that the ! inversion was incorrectly passed on to the second
3732 item. This has been fixed.
3734 PH/03 Added additional dnslists conditions == and =& which are different from
3735 = and & when the dns lookup returns more than one IP address.
3737 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3738 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3740 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3741 FSYNC, which compiles an option called disable_fsync that allows for
3742 bypassing fsync(). The documentation is heavily laced with warnings.
3744 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3746 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3747 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3748 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3749 including adding "make clean"; (3) Added -fPIC when compiling the test
3750 dynamically loaded module, to get rid of a warning.
3752 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3753 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3754 = 0s. The bug is otherwise harmless.
3756 PH/07 There was a bug in the dovecot authenticator such that the value of
3757 $auth1 could be overwritten, and so not correctly preserved, after a
3758 successful authentication. This usually meant that the value preserved by
3759 the server_setid option was incorrect.
3761 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3763 PH/09 Installed PCRE release 7.0.
3765 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3766 run for batched SMTP input. It is now run at the start of every message
3767 in the batch. While fixing this I discovered that the process information
3768 (output by running exiwhat) was not always getting set for -bs and -bS
3769 input. This is fixed, and it now also says "batched" for BSMTP.
3771 PH/11 Added control=no_pipelining.
3773 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3774 patch, slightly modified), and move the expansion of helo_data till after
3775 the connection is made in the smtp transport (so it can use these
3778 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3780 PH/14 Added log_selector = +pid.
3782 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3784 PH/16 Add ${if forany and ${if forall.
3786 PH/17 Added dsn_from option to vary the From: line in DSNs.
3788 PH/18 Flush SMTP output before performing a callout, unless control =
3789 no_callout_flush is set.
3791 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3792 was true (the default) a successful delivery failed to delete the retry
3793 item, thus causing premature timeout of the address. The bug is now
3796 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3798 PH/21 Long custom messages for fakedefer and fakereject are now split up
3799 into multiline responses in the same way that messages for "deny" and
3800 other ACL rejections are.
3802 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3803 with slight modification.
3805 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3806 draft, changing the syntax and factoring some duplicate code".
3808 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3809 for deliveries of the second and subsequent messages over the same SMTP
3812 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3813 ${reduce, with only minor "tidies".
3815 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3817 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3818 expansion side effects.
3820 PH/27 When a message times out after an over-quota error from an Exim-imposed
3821 quota, the bounce message says "mailbox is full". This message was not
3822 being given when it was a system quota that was exceeded. It now should
3825 MH/03 Made $recipients available in local_scan(). local_scan() already has
3826 better access to the recipient list through recipients_list[], but
3827 $recipients can be useful in postmaster-provided expansion strings.
3829 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3830 in the case of a MAIL command with additional options following the
3831 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3832 were accidentally chopped off.
3834 PH/29 SMTP synchronization checks are implemented when a command is read -
3835 there is a check that no more input is waiting when there shouldn't be
3836 any. However, for some commands, a delay in an ACL can mean that it is
3837 some time before the response is written. In this time, more input might
3838 arrive, invalidly. So now there are extra checks after an ACL has run for
3839 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3840 pipelining has not been advertised.
3842 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3844 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3845 returned in the SMTP response when smtp_return_error_details was false.
3846 This has been fixed.
3848 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3849 instead of the C I/O that was originally supplied, because problems were
3850 reported on Solaris.
3852 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3853 Exim which did not show up earlier: it was assuming that a call to
3854 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3855 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3856 was a macro that became an assignment, so it seemed to work. This has
3857 changed to a proper function call with a void return, hence the compile
3858 error. Exim's code has been fixed.
3860 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3863 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3865 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3867 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3868 because it turns out to be too big or there is a timeout) there is no
3869 "Completed" line in the log. When some messages of this type were
3870 selected by exigrep, they were listed as "not completed". Others were
3871 picked up by some special patterns. I have improved the selection
3872 criteria to be more general.
3874 PH/38 The host_find_failed option in the manualroute router can now be set
3875 to "ignore", to completely ignore a host whose IP address cannot be
3876 found. If all hosts are ignored, the behaviour is controlled by the new
3877 host_all_ignored option.
3879 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3880 homing or because of multiple MX records with /mx) generated more than
3881 one IP address, and the following item turned out to be the local host,
3882 all the secondary addresses of the first item were incorrectly removed
3883 from the list, along with the local host and any following hosts (which
3884 is what is supposed to happen).
3886 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3887 whoever called Exim into the -H file. In the case of the daemon it was
3888 behaving confusingly. When first started, it used values for whoever
3889 started the daemon, but after a SIGHUP it used the Exim user (because it
3890 calls itself on a restart). I have changed the code so that it now always
3893 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3894 message are rejected with the same error (e.g. no authentication or bad
3895 sender address), and a DATA command is nevertheless sent (as can happen
3896 with PIPELINING or a stupid MUA), the error message that was given to the
3897 RCPT commands is included in the rejection of the DATA command. This is
3898 intended to be helpful for MUAs that show only the final error to their
3901 PH/42 Another patch from the Sieve maintainer.
3903 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3904 Eximstats - Fixed some broken HTML links and added missing column headers
3906 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3907 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3909 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3911 PH/43 Yet another patch from the Sieve maintainer.
3913 PH/44 I found a way to check for a TCP/IP connection going away before sending
3914 the response to the final '.' that terminates a message, but only in the
3915 case where the client has not sent further data following the '.'
3916 (unfortunately, this is allowed). However, in many cases there won't be
3917 any further data because there won't be any more messages to send. A call
3918 to select() can be used: if it shows that the input is "ready", there is
3919 either input waiting, or the socket has been closed. An attempt to read
3920 the next input character can distinguish the two cases. Previously, Exim
3921 would have sent an OK response which the client would never have see.
3922 This could lead to message repetition. This fix should cure that, at
3923 least in a lot of common cases.
3925 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3926 advertised in response to EHLO.
3932 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3933 fixed by 4.65/MH/01 (is this a record?) are fixed:
3935 (i) An empty string was always treated as zero by the numeric comparison
3936 operators. This behaviour has been restored.
3938 (ii) It is documented that the numeric comparison operators always treat
3939 their arguments as decimal numbers. This was broken in that numbers
3940 starting with 0 were being interpreted as octal.
3942 While fixing these problems I realized that there was another issue that
3943 hadn't been noticed. Values of message_size_limit (both the global option
3944 and the transport option) were treated as octal if they started with 0.
3945 The documentation was vague. These values are now always treated as
3946 decimal, and I will make that clear in the documentation.
3952 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3953 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3956 MH/01 Don't check that the operands of numeric comparison operators are
3957 integers when their expansion is in "skipping" mode (fixes bug
3958 introduced by 4.64-PH/07).
3960 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3961 child addresses, Exim now panics and dies. Previously, because the count
3962 is held in a short int, deliveries were likely to be lost. As such a
3963 large number of recipients for a single message is ridiculous
3964 (performance will be very, very poor), I have chosen to impose a limit
3965 rather than extend the field.
3971 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3972 leftover -K file (the existence of which was triggered by #402).
3973 While we were at it, introduced process PID as part of the -K
3974 filename. This should rule out race conditions when creating
3977 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3978 processing considerably. Previous code took too long for large mails,
3979 triggering a timeout which in turn triggers #401.
3981 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3982 in the DK code in transports.c. sendfile() is not really portable,
3983 hence the _LINUX specificness.
3985 TF/01 In the add_headers option to the mail command in an Exim filter,
3986 there was a bug that Exim would claim a syntax error in any
3987 header after the first one which had an odd number of characters
3990 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3991 callout verification, Exim cached a "reject" for the entire domain. This
3992 is correct for most verifications, but it is not correct for a recipient
3993 verification with use_sender or use_postmaster set, because in that case
3994 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3995 case of MAIL FROM:<> rejection from other early rejections (e.g.
3996 rejection of HELO). When verifying a recipient using a non-null MAIL
3997 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3998 Whatever the result of the callout, the value of the domain cache is
3999 left unchanged (for any other kind of callout, getting as far as trying
4000 RCPT means that the domain itself is ok).
4002 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4005 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4006 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4007 session, the daemon ignores ECONNECT errors and logs others; it now
4008 ignores EPIPE as well.
4010 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4011 (quoted-printable decoding).
4013 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4014 later the small subsequent patch to fix an introduced bug.
4016 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4018 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4020 PH/08 An error is now given if message_size_limit is specified negative.
4022 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4023 to be given (somewhat) arbitrary names.
4025 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4028 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4029 miscellaneous code fixes
4031 PH/10 Added the log_reject_target ACL modifier to specify where to log
4034 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4035 hostname. This is wrong, because it relates to the incoming message (and
4036 probably the interface on which it is arriving) and not to the outgoing
4037 callout (which could be using a different interface). This has been
4038 changed to use the value of the helo_data option from the smtp transport
4039 instead - this is what is used when a message is actually being sent. If
4040 there is no remote transport (possible with a router that sets up host
4041 addresses), $smtp_active_hostname is used.
4043 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4044 tweaks were necessary in order to get it to work (see also 21 below):
4045 (a) The code assumed that strncpy() returns a negative number on buffer
4046 overflow, which isn't the case. Replaced with Exim's string_format()
4048 (b) There were several signed/unsigned issues. I just did the minimum
4049 hacking in of casts. There is scope for a larger refactoring.
4050 (c) The code used strcasecmp() which is not a standard C function.
4051 Replaced with Exim's strcmpic() function.
4052 (d) The code set only $1; it now sets $auth1 as well.
4053 (e) A simple test gave the error "authentication client didn't specify
4054 service in request". It would seem that Dovecot has changed its
4055 interface. Fortunately there's a specification; I followed it and
4056 changed what the client sends and it appears to be working now.
4058 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4061 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4062 address A is aliased to B and C, where B exists and C does not. Without
4063 -v the output is "A verified" because verification stops after a
4064 successful redirection if more than one address is generated. However,
4065 with -v the child addresses are also verified. Exim was outputting "A
4066 failed to verify" and then showing the successful verification for C,
4067 with its parentage. It now outputs "B failed to verify", showing B's
4068 parentage before showing the successful verification of C.
4070 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4071 look up a TXT record in a specific list after matching in a combined
4074 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4075 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4076 they consult the DNS. I had assumed they would set it the way they
4077 wanted; and indeed my experiments on Linux seem to show that in some
4078 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4079 To be on the safe side, however, I have now made the interface to
4080 host_find_byname() similar to host_find_bydns(), with an argument
4081 containing the DNS resolver options. The host_find_byname() function now
4082 sets these options at its start, just as host_find_bydns() does. The smtp
4083 transport options dns_qualify_single and dns_search_parents are passed to
4084 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4085 of host_find_byname() use the default settings of RES_DEFNAMES
4086 (qualify_single) but not RES_DNSRCH (search_parents).
4088 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4089 spool_read_header() do less string testing, by means of a preliminary
4090 switch on the second character of optional "-foo" lines. (This is
4091 overdue, caused by the large number of possibilities that now exist.
4092 Originally there were few.) While I was there, I also converted the
4093 str(n)cmp tests so they don't re-test the leading "-" and the first
4094 character, in the hope this might squeeze out yet more improvement.
4096 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4097 flag allowing group syntax was set by the header_syntax check but not
4098 turned off, possible causing trouble later; (2) The flag was not being
4099 set at all for the header_verify test, causing "group"-style headers to
4100 be rejected. I have now set it in this case, and also caused header_
4101 verify to ignore an empty address taken from a group. While doing this, I
4102 came across some other cases where the code for allowing group syntax
4103 while scanning a header line wasn't quite right (mostly, not resetting
4104 the flag correctly in the right place). These bugs could have caused
4105 trouble for malformed header lines. I hope it is now all correct.
4107 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4108 with the "reply" argument non-NULL. The code, however (which originally
4109 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4110 but it didn't always do it. This confused somebody who was copying the
4111 code for some other use. I have removed all the tests.
4113 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4114 feature that was used to support insecure browsers during the U.S. crypto
4115 embargo. It requires special client support, and Exim is probably the
4116 only MTA that supported it -- and would never use it because real RSA is
4117 always available. This code has been removed, because it had the bad
4118 effect of slowing Exim down by computing (never used) parameters for the
4119 RSA_EXPORT functionality.
4121 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4122 authenticator to fail if there's a tab character in the incoming data
4123 (there should never be unless someone is messing about, as it's supposed
4124 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4125 if the connection is using TLS or if the remote IP is the same as the
4126 local IP, and the "valid-client-cert option" if a client certificate has
4129 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4130 authenticators. This can be used for authorization after authentication
4131 succeeds. (In the case of plaintext, it servers for both authentication
4134 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4135 if any retry times were supplied.
4137 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4138 connection, where there is no client IP address to check. In this
4139 situation, the verify now always succeeds.
4141 PH/25 Applied John Jetmore's -Mset patch.
4143 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4145 PH/27 In a string expansion for a processed (not raw) header when multiple
4146 headers of the same name were present, leading whitespace was being
4147 removed from all of them, but trailing whitespace was being removed only
4148 from the last one. Now trailing whitespace is removed from each header
4149 before concatenation. Completely empty headers in a concatenation (as
4150 before) are ignored.
4152 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4153 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4155 PH/29 [Removed. This was a change that I later backed out, and forgot to
4156 correct the ChangeLog entry (that I had efficiently created) before
4157 committing the later change.]
4159 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4160 address errors (4xx response to RCPT) over the same connection as other
4161 messages routed to the same hosts. Such deliveries are always "forced",
4162 so retry times are not inspected. This resulted in far too many retries
4163 for the affected addresses. The effect occurred only when there were more
4164 hosts than the hosts_max_try setting in the smtp transport when it had
4165 the 4xx errors. Those hosts that it had tried were not added to the list
4166 of hosts for which the message was waiting, so if all were tried, there
4167 was no problem. Two fixes have been applied:
4169 (i) If there are any address or message errors in an SMTP delivery, none
4170 of the hosts (tried or untried) are now added to the list of hosts
4171 for which the message is waiting, so the message should not be a
4172 candidate for sending over the same connection that was used for a
4173 successful delivery of some other message. This seems entirely
4174 reasonable: after all the message is NOT "waiting for some host".
4175 This is so "obvious" that I'm not sure why it wasn't done
4176 previously. Hope I haven't missed anything, but it can't do any
4177 harm, as the worst effect is to miss an optimization.
4179 (ii) If, despite (i), such a delivery is accidentally attempted, the
4180 routing retry time is respected, so at least it doesn't keep
4181 hammering the server.
4183 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4184 in ${readsocket because some servers need this prod.
4186 PH/32 Added some extra debug output when updating a wait-xxx database.
4188 PH/33 The hint "could be header name not terminated by colon", which has been
4189 given for certain expansion errors for a long time, was not being given
4190 for the ${if def:h_colon_omitted{... case.
4192 PH/34 The spec says: "With one important exception, whenever a domain list is
4193 being scanned, $domain contains the subject domain." There was at least
4194 one case where this was not true.
4196 PH/35 The error "getsockname() failed: connection reset by peer" was being
4197 written to the panic log as well as the main log, but it isn't really
4198 panic-worthy as it just means the connection died rather early on. I have
4199 removed the panic log writing for the ECONNRESET error when getsockname()
4202 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4203 runs only) independently of the message's sender address. This meant
4204 that, if the 4xx error was in fact related to the sender, a different
4205 message to the same recipient with a different sender could confuse
4206 things. In particular, this can happen when sending to a greylisting
4207 server, but other circumstances could also provoke similar problems.
4208 I have changed the default so that the retry time for these errors is now
4209 based a combination of the sender and recipient addresses. This change
4210 can be overridden by setting address_retry_include_sender=false in the
4213 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4214 remote server are returned as part of bounce messages. This was not
4215 happening for LMTP over a pipe (the lmtp transport), but now it is the
4216 same for both kinds of LMTP.
4218 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4219 in header lines that were in fact CNAMEs. This is no longer the case.
4221 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4222 and queue runs started by the daemon processed all messages. This has
4223 been fixed so that -R and -S can now usefully be given with -q<time>.
4225 PH/40 Import PCRE release 6.7 (fixes some bugs).
4227 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4229 PH/42 Give an error if -q is specified more than once.
4231 PH/43 Renamed the variables $interface_address and $interface_port as
4232 $received_ip_address and $received_port, to make it clear that these
4233 values apply to message reception, and not to the outgoing interface when
4234 a message is delivered. (The old names remain recognized, of course.)
4236 PH/44 There was no timeout on the connect() call when using a Unix domain
4237 socket in the ${readsocket expansion. There now is.
4239 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4240 be meaningful with "accept".
4242 SC/01 Eximstats V1.43
4243 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4245 SC/02 Eximstats V1.44
4246 Use a glob alias rather than an array ref in the generated
4247 parser. This improves both readability and performance.
4249 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4250 Collect SpamAssassin and rejection statistics.
4251 Don't display local sender or destination tables unless
4252 there is data to show.
4253 Added average volumes into the top table text output.
4255 SC/04 Eximstats V1.46
4256 Collect data on the number of addresses (recipients)
4257 as well as the number of messages.
4259 SC/05 Eximstats V1.47
4260 Added 'Message too big' to the list of mail rejection
4261 reasons (thanks to Marco Gaiarin).
4263 SC/06 Eximstats V1.48
4264 Mainlog lines which have GMT offsets and are too short to
4265 have a flag are now skipped.
4267 SC/07 Eximstats V1.49 (Alain Williams)
4268 Added the -emptyok flag.
4270 SC/08 Eximstats V1.50
4271 Fixes for obtaining the IP address from reject messages.
4273 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4274 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4275 whitespace changes from 4.64-PH/27
4277 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4280 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4281 are found, allow negative numbers in numeric criteria)
4283 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4285 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4286 to match changes made in 4.64-PH/43
4288 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4290 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4291 "domain" is actually an IP address, and force a failure. This locks out
4292 those revolvers/nameservers that support "A-for-A" lookups, in
4293 contravention of the specifications.
4295 PH/48 When a host name was looked up from an IP address, and the subsequent
4296 forward lookup of the name timed out, the host name was left in
4297 $sender_host_name, contrary to the specification.
4299 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4300 restricted to single-key lookups, Exim was not diagnosing an error if
4301 * or *@ was used with a query-style lookup.
4303 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4305 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4306 long ago, but noone interested enough thought of it. Let's just say that
4307 the "1.1" means that there are some new functions that weren't there at
4308 some point in the past.
4310 PH/51 Error processing for expansion failure of helo_data from an smtp
4311 transport during callout processing was broken.
4313 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4314 tested/used via the -bh/-bhc/-bs options.
4316 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4317 bug, fixed in subsequent PCRE releases).
4319 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4320 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4322 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4328 SC/01 Use a glob alias rather than an array ref in eximstats generated
4329 parser. This improves both readability and performance.
4331 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4332 Don't display local sender or destination tables in eximstats unless
4333 there is data to show.
4334 Added average volumes into the eximstats top table text output.
4336 SC/03 Collect data on the number of addresses (recipients) as well
4337 as the number of messages in eximstats.
4339 TF/01 Correct an error in the documentation for the redirect router. Exim
4340 does (usually) call initgroups() when daemonizing.
4342 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4343 with consistent privilege compared to when running as a daemon.
4345 TF/03 Note in the spec that $authenticated_id is not set for local
4346 submissions from trusted users.
4348 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4349 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4351 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4352 by adding some example configuration directives to the default
4353 configuration file. A little bit of work is required to uncomment the
4354 directives and define how usernames and passwords are checked, but
4355 there is now a framework to start from.
4357 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4358 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4359 without this. I don't know how relevant this is to other LDAP libraries.
4361 PH/02 Add the verb name to the "unknown ACL verb" error.
4363 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4365 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4367 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4368 directory not expanded when it should be if an expanded home directory
4369 was set for the address (which is overridden by the transport).
4371 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4374 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4375 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4376 because it is too late at that time, and has no effect.
4378 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4379 security issue with \' (bugzilla #107). I could not use the
4380 PQescapeStringConn() function, because it needs a PGconn value as one of
4383 PH/08 When testing addresses using -bt, indicate those final addresses that
4384 are duplicates that would not cause an additional delivery. At least one
4385 person was confused, thinking that -bt output corresponded to deliveries.
4386 (Suppressing duplicates isn't a good idea as you lose the information
4387 about possibly different redirections that led to the duplicates.)
4389 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4390 systems where poll() doesn't work, in particular OS X.
4392 PH/10 Added more information to debugging output for retry time not reached.
4394 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4395 operations in malware.c.
4397 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4400 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4401 syslog_duplication set false, log lines that would normally be written
4402 both the the main log and to the reject log were not written to syslog at
4405 PH/14 In the default configuration, change the use of "message" in ACL warn
4406 statements to "add_header".
4408 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4409 not followed by a command (e.g. "seen endif").
4411 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4412 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4415 PH/17 Added extra conditions to the default value of delay_warning_condition
4419 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4420 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4421 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4424 The Auto-Submitted: and various List- headers are standardised, whereas I
4425 don't think Precedence: ever was.
4427 PH/18 Refactored debugging code in route_finduser() to show more information,
4428 in particular, the error code if getpwnam() issues one.
4430 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4431 This is apparently needed in addition to the PH/07 change above to avoid
4432 any possible encoding problems.
4434 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4435 but not after initializing Perl.
4437 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4438 output them only if debugging. By default they are written stderr,
4439 apparently, which is not desirable.
4441 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4444 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4447 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4449 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4450 authenticated or an ident call has been made. Suppress the default
4451 values for $authenticated_id and $authenticated_sender (but permit -oMai
4452 and -oMas) when testing with -bh.
4454 PH/24 Re-jigged the order of the tests in the default configuration so that the
4455 tests for valid domains and recipients precede the DNS black list and CSA
4456 tests, on the grounds that those ones are more expensive.
4458 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4459 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4460 This bug exists in every version of Exim that I still have, right back to
4463 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4464 However, an attempt to turn on case-sensitivity in a regex key by
4465 including (?-i) didn't work because the subject string was already
4466 lowercased, and the effects were non-intuitive. It turns out that a
4467 one-line patch can be used to allow (?-i) to work as expected.
4473 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4474 other effects) broke the use of negated acl sub-conditions.
4476 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4479 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4480 "Deny" causes Exim to reject the incoming connection with a 554 error.
4481 Unfortunately, if there is a major crisis, such as a disk failure,
4482 tcp-wrappers gives "deny", whereas what one would like would be some
4483 kind of temporary error. A kludge has been added to help with this.
4484 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4485 554 error is used if errno is still zero or contains ENOENT (which occurs
4486 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4489 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4491 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4492 errors. Otherwise a message that provokes a temporary error (when other
4493 messages do not) can cause a whole host to time out.
4495 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4496 addresses were routed directly to files or pipes from a redirect router.
4497 File deliveries just didn't batch; pipe deliveries might have suffered
4500 PH/06 A failure to get a lock for a hints database would erroneously always say
4501 "Failed to get write lock", even when it was really a read lock.
4503 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4504 of 0600. This has been changed to use the value of the lockfile_mode
4505 option (which defaults to 0600).
4507 PH/08 Applied small patch from the Sieve maintainer.
4509 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4510 folder from quota calculations, a direct delivery into this folder messed
4511 up the contents of the maildirsize file. This was because the regex was
4512 used only to exclude .Trash (or whatever) when the size of the mailbox
4513 was calculated. There was no check that a delivery was happening into an
4514 excluded directory. This bug has been fixed by ignoring all quota
4515 processing for deliveries into excluded directories.
4517 PH/10 Added the maildirfolder_create_regex option to appendfile.
4523 PH/01 The code for finding all the local interface addresses on a FreeBSD
4524 system running IPv6 was broken. This may well have applied to all BSD
4525 systems, as well as to others that have similar system calls. The broken
4526 code found IPv4 interfaces correctly, but gave incorrect values for the
4527 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4528 that it would not match correctly against @[] and not recognize the IPv6
4531 PH/02 The ipliteral router was not recognizing addresses of the form user@
4532 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4534 PH/03 Added disable_ipv6.
4536 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4537 decoded form, because it is most often used to construct To: headers
4538 lines in autoreplies, and the decoded form may well be syntactically
4539 invalid. However, $reply_address has leading white space removed, and all
4540 newlines turned into spaces so that the autoreply transport does not
4543 PH/05 If group was specified without a user on a router, and no group or user
4544 was specified on a transport, the group from the router was ignored.
4546 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4547 for visible compile-time settings that can be used to change these
4548 numbers, for those that want even more. Backwards compatibility with old
4549 spool files has been maintained. However, going back to a previous Exim
4550 release will lost any variables that are in spool files.
4552 PH/07 Two small changes when running in the test harness: increase delay when
4553 passing a TCP/IP connection to a new process, in case the original
4554 process has to generate a bounce, and remove special handling of
4555 127.0.0.2 (sic), which is no longer necessary.
4557 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4558 be the same on different OS.
4560 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4563 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4564 whether --show-vars was specified or not
4566 JJ/02 exipick: Added support for new ACL variable spool format introduced
4569 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4570 syntactically invalid From: or Reply-to: line, and a filter used this to
4571 generate an autoreply, and therefore failed to obtain an address for the
4572 autoreply, Exim could try to deliver to a non-existent relative file
4573 name, causing unrelated and misleading errors. What now happens is that
4574 it logs this as a hard delivery error, but does not attempt to create a
4577 PH/11 The exinext utility has a -C option for testing purposes, but although
4578 the given file was scanned by exinext itself; it wasn't being passed on
4579 when Exim was called.
4581 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4582 an end-of-file indication when reading a command response.
4584 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4585 compiled. In many other places in Exim, IPv6 addresses are always
4586 recognized, so I have changed this. It also means that IPv4 domain
4587 literals of the form [IPV4:n.n.n.n] are now always recognized.
4589 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4590 used if the router is not running as root, for example, when verifying at
4591 ACL time, or when using -bh. The debugging output from this situation was
4592 non-existent - all you got was a failure to exec. I have made two
4595 (a) Failures to set uid/gid, the current directory, or a process leader
4596 in a subprocess such as that created by queryprogram now generate
4597 suitable debugging output when -d is set.
4599 (b) The queryprogram router detects when it is not running as root,
4600 outputs suitable debugging information if -d is set, and then runs
4601 the subprocess without attempting to change uid/gid.
4603 PH/15 Minor change to Makefile for building test_host (undocumented testing
4606 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4607 additional section of a DNS packet that returns MX or SRV records.
4608 Instead, it always explicitly searches for A/AAAA records. This avoids
4609 major problems that occur when a DNS server includes only records of one
4610 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4611 fixed another bug: if SRV records were looked up and the corresponding
4612 address records were *not* found in the additional section, the port
4613 values from the SRV records were lost.
4615 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4616 using the correct key (the original address) when searching the retry
4617 rules in order to find which one to use for generating the retry hint.
4619 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4620 adding the default one. Similarly, if it contains a Reply-To: header, the
4621 errors_reply_to option, if set, is not used.
4623 PH/19 When calculating a retry time, Exim used to measure the "time since
4624 failure" by looking at the "first failed" field in the retry record. Now
4625 it does not use this if it is later than than the arrival time of the
4626 message. Instead it uses the arrival time. This makes for better
4627 behaviour in cases where some deliveries succeed, thus re-setting the
4628 "first failed" field. An example is a quota failure for a huge message
4629 when small messages continue to be delivered. Without this change, the
4630 "time since failure" will always be short, possible causing more frequent
4631 delivery attempts for the huge message than are intended.
4632 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4634 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4635 $1, $2, $3) because the numerical variables can be reset during some
4636 expansion items (e.g. "match"), thereby losing the authentication data.
4638 PH/21 Make -bV show the size of off_t variables so that the test suite can
4639 decide whether to run tests for quotas > 2G.
4641 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4642 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4643 filecount value is greater than 2G or if a quota value is greater than 2G
4644 on a system where the size of off_t is not greater than 4, a panic error
4647 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4648 never match. The debug and -bh output now contains an explicit error
4649 message indicating a malformed IPv4 address or mask.
4651 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4652 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4653 PH/23 above applies.
4655 PH/25 Do not write to syslog when running in the test harness. The only
4656 occasion when this arises is a failure to open the main or panic logs
4657 (for which there is an explicit test).
4659 PH/26 Added the /no_tell option to "control=freeze".
4661 PH/27 If a host name lookup failed very early in a connection, for example, if
4662 the IP address matched host_lookup and the reverse lookup yielded a name
4663 that did not have a forward lookup, an error message of the form "no IP
4664 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4665 could be logged. Now it outputs the IP address instead of "NULL".
4667 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4668 allows the sender and the authenticated sender to be set when
4669 submitting a message from within Exim. Since child_open_exim() is
4670 documented for local_scan(), the new function should be too.
4672 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4673 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4674 results in an empty string is now treated as unset.
4676 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4678 PH/31 Added sender_verify_fail logging option.
4680 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4681 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4682 tidied the source and removed it altogether.
4684 PH/33 When a queue run was abandoned because the load average was too high, a
4685 log line was always written; now it is written only if the queue_run log
4686 selector is set. In addition, the log line for abandonment now contains
4687 information about the queue run such as the pid. This is always present
4688 in "start" and "stop" lines but was omitted from the "abandon" line.
4690 PH/34 Omit spaces between a header name and the colon in the error message that
4691 is given when verify = headers_syntax fails (if there are lots of them,
4692 the message gets confusing).
4694 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4695 names, as there are now some PTR records that contain slashes. This check
4696 is only to protect against broken name servers that fall over on strange
4697 characters, so the fact that it applies to all lookups doesn't matter.
4699 PH/36 Now that the new test suite is complete, we can remove some of the
4700 special code in Exim that was needed for the old test suite. For example,
4701 sorting DNS records because real resolvers return them in an arbitrary
4702 order. The new test suite's fake resolver always returns records in the
4705 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4706 bounces) except when queue_only is set, to avoid logging races between
4707 the different processes.
4709 PH/38 Panic-die if .include specifies a non-absolute path.
4711 PH/39 A tweak to the "H" retry rule from its user.
4713 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4714 a label. They prevented compilation on older perls.
4716 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4717 a warning to be raised on newish perls.
4719 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4720 on queue. Changes to match documented behaviour of showing count of
4721 messages matching specified criteria.
4723 PH/40 Changed the default ident timeout from 30s to 5s.
4725 PH/41 Added support for the use of login_cap features, on those BSD systems
4726 that have them, for controlling the resources used by pipe deliveries.
4728 PH/42 The content-scanning code uses fopen() to create files in which to put
4729 message data. Previously it was not paying any attention to the mode of
4730 the files. Exim runs with umask(0) because the rest of the code creates
4731 files with open(), and sets the required mode explicitly. Thus, these
4732 files were ending up world-writeable. This was not a big issue, because,
4733 being within the spool directory, they were not world-accessible. I have
4734 created a function called modefopen, which takes an additional mode
4735 argument. It sets umask(777), creates the file, chmods it to the required
4736 mode, then resets the umask. All the relevant calls to fopen() in the
4737 content scanning code have been changed to use this function.
4739 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4740 to 24 hours. This avoids potential overflow problems when processing G
4741 and H retry rules. I suspect nobody ever tinkers with this value.
4743 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4745 PH/45 When the plaintext authenticator is running as a client, the server's
4746 challenges are checked to ensure they are valid base64 strings. By
4747 default, the authentication attempt is cancelled if an invalid string is
4748 received. Setting client_ignore_invalid_base64 true ignores these errors.
4749 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4750 they are received. Thus, the responses can be made to depend on the
4751 challenges. If an invalid string is ignored, an empty string is placed in
4754 PH/46 Messages that are created by the autoreply transport now contains a
4755 References: header, in accordance with RFCs 2822 and 3834.
4757 PH/47 Added authenticated_sender_force to the smtp transport.
4759 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4761 PH/49 Installed latest patch from the Sieve maintainer.
4763 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4764 was also set, the appendfile transport was unnecessarily scanning a
4765 directory of message files (e.g. for maildir delivery) to find the count
4766 of files (along with the size), even though it did not need this
4767 information. It now does the scan only if it needs to find either the
4768 size of the count of files.
4770 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4772 PH/52 Two bugs concerned with error handling when the smtp transport is
4775 (i) Exim was not creating retry information for temporary errors given
4776 for individual recipients after the DATA command when the smtp transport
4777 was used in LMTP mode. This meant that they could be retried too
4778 frequently, and not timed out correctly.
4780 (ii) Exim was setting the flag that allows error details to be returned
4781 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4782 recipients that were returned after the DATA command.
4784 PH/53 This is related to PH/52, but is more general: for any failing address,
4785 when detailed error information was permitted to be returned to the
4786 sender, but the error was temporary, then after the final timeout, only
4787 "retry timeout exceeded" was returned. Now it returns the full error as
4788 well as "retry timeout exceeded".
4790 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4791 do this, and (what is worse) MTAs that accept it.
4793 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4794 will now be deprecated.
4796 PH/56 New os.c-cygwin from the Cygwin maintainer.
4798 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4799 formats (previously only available in exim formats via -bpr, -bpru,
4800 and -bpra. Now also available in native and exiqgrep formats)
4802 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4803 with very large, slow to parse queues
4805 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4807 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4809 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4810 responses to authentication challenges, though it was showing the
4811 challenges; (ii) I've removed the CR characters from the debug output for
4814 PH/58 Allow for the insertion of a newline as well as a space when a string
4815 is turned into more than one encoded-word during RFC 2047 encoding. The
4816 Sieve code now uses this.
4818 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4819 data_4xx, lost_connection, tls_required.
4821 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4822 message was being sent as an SMTP response.
4824 PH/61 Add -l and -k options to exicyclog.
4826 PH/62 When verifying, if an address was redirected to one new address, so that
4827 verification continued, and the new address failed or deferred after
4828 having set something in $address_data, the value of $address_data was not
4829 passed back to the ACL. This was different to the case when no
4830 redirection occurred. The value is now passed back in both cases.
4832 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4833 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4834 use login_cap.h, so on its own it isn't the distinguishing feature. The
4835 new name refers directly to the setclassresources() function.
4837 PH/65 Added configuration files for NetBSD3.
4839 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4841 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4842 is preferred over IPv4.
4844 PH/68 The bounce_return_message and bounce_return_body options were not being
4845 honoured for bounces generated during the reception of non-SMTP messages.
4846 In particular, this applied to messages rejected by the ACL. This bug has
4847 been fixed. However, if bounce_return_message is true and bounce_return_
4848 body is false, the headers that are returned for a non-SMTP message
4849 include only those that have been read before the error was detected.
4850 (In the case of an ACL rejection, they have all been read.)
4852 PH/69 The HTML version of the specification is now built in a directory called
4853 spec_html instead of spec.html, because the latter looks like a path with
4854 a MIME-type, and this confuses some software.
4856 PH/70 Catch two compiler warnings in sieve.c.
4858 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4859 function verify_get_ident() calls ip_connect() to connect a socket, but
4860 if the "connect()" function timed out, ip_connect() used to close the
4861 socket. However, verify_get_ident() also closes the socket later, and in
4862 between Exim writes to the log, which may get opened at this point. When
4863 the socket was closed in ip_connect(), the log could get the same file
4864 descriptor number as the socket. This naturally causes chaos. The fix is
4865 not to close the socket in ip_connect(); the socket should be closed by
4866 the function that creates it. There was only one place in the code where
4867 this was missing, in the iplookup router, which I don't think anybody now
4868 uses, but I've fixed it anyway.
4870 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4871 well as to direct DNS lookups. Otherwise the handling of names in host
4872 lists is inconsistent and therefore confusing.
4878 PH/01 Two changes to the default runtime configuration:
4880 (1) Move the checks for relay_from_hosts and authenticated clients from
4881 after to before the (commented out) DNS black list checks.
4883 (2) Add control=submission to the relay_from_hosts and authenticated
4884 clients checks, on the grounds that messages accepted by these
4885 statements are most likely to be submissions.
4887 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4889 (1) Generate an error if the third argument for the ${prvs expansion is
4892 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4895 (3) Reset the variables that are obtained from the first argument of
4896 ${prvscheck and used in the second argument before leaving the code,
4897 because their memory is reclaimed, so using them afterwards may do
4900 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4901 one (it's much easier than Tom thought :-).
4903 (5) Because of (4), we can now allow for the use of $prvscheck_result
4904 inside the third argument.
4906 PH/03 For some reason, the default setting of PATH when running a command from
4907 a pipe transport was just "/usr/bin". I have changed it to
4910 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4911 anything to be listed in the output from -bV.
4913 PH/05 When a filter generated an autoreply, the entire To: header line was
4914 quoted in the delivery log line, like this:
4916 => >A.N.Other <ano@some.domain> <original@ddress> ...
4918 This has been changed so that it extracts the operative address. There
4919 may be more than one such address. If so, they are comma-separated, like
4922 => >ano@some.domain,ona@other.domain <original@ddress> ...
4924 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4925 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4926 not being looked up in the rDNS to get a host name, Exim was showing the
4927 IP address twice in Received: lines, even though the IP addresses were
4928 identical. For example:
4930 Received: from [1.2.3.4] (helo=[1.2.3.4])
4932 However, if the real host name was known, it was omitting the HELO data
4933 if it matched the actual IP address. This has been tidied up so that it
4934 doesn't show the same IP address twice.
4936 PH/07 When both +timestamp and +memory debugging was on, the value given by
4937 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4938 called by the debug printing, thereby overwriting the timestamp buffer.
4939 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4941 PH/08 When the original message was included in an autoreply transport, it
4942 always said "this is a copy of the message, including all the headers",
4943 even if body_only or headers_only was set. It now gives an appropriate
4946 PH/09 Applied a patch from the Sieve maintainer which:
4948 o fixes some comments
4949 o adds the (disabled) notify extension core
4950 o adds some debug output for the result of if/elsif tests
4951 o points to the current vacation draft in the documentation
4952 and documents the missing references header update
4956 o fixes a bug in processing the envelope test (when testing
4957 multiple envelope elements, the last element determined the
4960 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4961 Electronic Mail") by including:
4963 Auto-submitted: auto-generated
4965 in the messages that it generates (bounce messages and others, such as
4966 warnings). In the case of bounce messages for non-SMTP messages, there was
4967 also a typo: it was using "Auto_submitted" (underscore instead of
4968 hyphen). Since every message generated by Exim is necessarily in response
4969 to another message, thes have all been changed to:
4971 Auto-Submitted: auto-replied
4973 in accordance with these statements in the RFC:
4975 The auto-replied keyword:
4977 - SHOULD be used on messages sent in direct response to another
4978 message by an automatic process,
4980 - MUST NOT be used on manually-generated messages,
4982 - MAY be used on Delivery Status Notifications (DSNs) and Message
4983 Disposition Notifications (MDNs),
4985 - MUST NOT be used on messages generated by automatic or periodic
4986 processes, except for messages which are automatic responses to
4989 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4990 to the default Received: header definition.
4992 PH/12 Added log selector acl_warn_skipped (default on).
4994 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4995 variables because (a) they are in the wrong storage pool and (b) even if
4996 they were copied, it wouldn't work properly because of the caching.
4998 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4999 checking when decoding. Apparently there are clients that generate
5000 overlong encoded strings. Why am I not surprised?
5002 PH/15 If the first argument of "${if match_address" was not empty, but did not
5003 contain an "@" character, Exim crashed. Now it writes a panic log message
5004 and treats the condition as false.
5006 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5008 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5009 extension "envelope-auth". The code is finished and in agreement with
5010 other implementations, but there is no documentation so far and in fact,
5011 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5012 not changing the active code.
5014 Print executed "if" and "elsif" statements when debugging is used. This
5015 helps a great deal to understand what a filter does.
5017 Document more things not specified clearly in RFC3028. I had all this
5018 sorted out, when out of a sudden new issues came to my mind. Oops."
5020 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5023 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5024 canonical form (as documented). However, after a host name lookup from
5025 the IP address, check_host() was doing a simple string comparison with
5026 addresses acquired from the DNS when checking that the found name did
5027 have the original IP as one of its addresses. Since any found IPv6
5028 addresses are likely to be in abbreviated form, the comparison could
5029 fail. Luckily, there already exists a function for doing the comparison
5030 by converting both addresses to binary, so now that is used instead of
5031 the text comparison.
5033 PH/20 There was another similar case to PH/19, when a complete host name was
5034 given in a host list; looking up its IP address could give an abbreviated
5035 form, whereas the current host's name might or might not be abbreviated.
5036 The same fix has been applied.
5042 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5043 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5046 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5047 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5049 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5051 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5052 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5053 Also, for "Auto-Submitted", treat anything other than "no" as
5054 non-personal, in accordance with RFC 3834. (Previously it treated
5055 anything starting "auto-" as non-personal.)
5057 TF/01 The control=submission/name=... option had a problem with syntax
5058 errors if the name included a slash character. The /name= option
5059 now slurps the rest of the string, so it can include any characters
5060 but it must come last in the list of options (after /sender_retain
5063 PH/05 Some modifications to the interface to the fake nameserver for the new
5071 TK/01 Added the "success_on_redirect" address verification option. See
5072 NewStuff for rationale and an example.
5074 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5076 PH/02 Patch to exigrep to allow it to work on syslog lines.
5078 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5079 fread() to skip over the body file's header line, because in Cygwin the
5080 header line is locked and is inaccessible.
5082 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5083 co-exist for some time) to make it clear that it is the Exim ID that is
5084 referenced, not the Message-ID: header line.
5086 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5087 string_format() function, because snprintf() does not exist on all
5090 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5091 ${stat: expansion item.
5093 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5094 protocol synchronization error", to keep the pedants happy.
5096 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5097 well as for IRIX systems, when gcc is being used. See the host.c source
5100 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5102 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5105 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5106 transport and to the smtp transport in LMTP mode.
5108 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5110 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5111 run a filter in a subprocess. This could lead to confusion in subsequent
5112 lookups in the parent process. There should also be a search_tidyup() at
5113 the end of the subprocess.
5115 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5116 only if the host matched helo_try_verify_hosts, which caused the
5117 verification to occur when the EHLO/HELO command was issued. The ACL just
5118 tested the remembered result. Now, if a previous verification attempt has
5119 not happened, "verify = helo" does it there and then.
5121 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5123 TK/03 Fix log output including CR from clamd.
5125 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5126 exist provoked a memory error which could cause a segfault.
5128 PH/15 Installed PCRE 6.2
5130 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5132 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5133 of the problem. Specifically, suggested +O2 rather than +O1 for the
5136 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5138 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5139 record, the port was not correctly logged when the outgoing_port log
5140 selector was set (it logged the transort's default port).
5142 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5143 fallback_hosts, and "hosts" in the smtp transport.
5145 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5146 host errors such as "Connection refused".
5148 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5149 authentication with radiusclient 0.4.9:
5151 - Error returned from rc_read_config was caught wrongly
5152 - Username/password not passed on to radius server due to wrong length.
5154 The presumption is that some radiusclient API changes for 4.51/PH/17
5155 were not taken care of correctly. The code is still untested by me (my
5156 Linux distribution still has 0.3.2 of radiusclient), but it was
5157 contributed by a Radius user.
5159 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5160 expanding the "port" option of the smtp transport.
5162 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5163 while reading a MIME header. Thanks to Tom Hughes for a patch.
5165 PH/24 Include config.h inside local_scan.h so that configuration settings are
5168 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5169 that in an ACL for RCPT (for example), you can examine exactly what was
5172 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5173 commands, but it was not correctly comparing the address with the actual
5174 client host address. Thus, it would show the EHLO address in Received:
5175 header lines when this was not necessary.
5177 PH/27 Added the % operator to ${eval:}.
5179 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5180 it should be ignoring failures (because with -C, for example, it has lost
5181 privilege). It wasn't ignoring creation failures other than "already
5184 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5187 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5188 because an input line was too long, either on its own, or by virtue of
5189 too many continuations, the temporary file was not being removed, and the
5190 return code was incorrect.
5192 PH/31 Missing "BOOL" in function definition in filtertest.c.
5194 PH/32 Applied Sieve patches from the maintainer.
5196 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5198 PH/33 Added "verify = not_blind".
5200 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5201 Local/Makefile (with some defaults set). These are used in built scripts
5202 such as exicyclog, but they have never been used in the exim_install
5203 script (though there are many overriding facilities there). I have
5204 arranged that the exim_install script now takes note of these two
5207 PH/35 Installed configuration files for Dragonfly.
5209 PH/36 When a locally submitted message by a trusted user did not contain a
5210 From: header, and the sender address was obtained from -f or from an SMTP
5211 MAIL command, and the trusted user did not use -F to supply a sender
5212 name, $originator_name was incorrectly used when constructing a From:
5213 header. Furthermore, $originator_name was used for submission mode
5214 messages from external hosts without From: headers in a similar way,
5215 which is clearly wrong.
5217 PH/37 Added control=suppress_local_fixups.
5219 PH/38 When log_selector = +received_sender was set, and the addition of the
5220 sender made the log line's construction buffer exactly full, or one byte
5221 less than full, an overflow happened when the terminating "\n" was
5224 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5225 when the result of a list match is failure because a DNS lookup failed.
5227 PH/40 RM_COMMAND is now used in the building process.
5229 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5230 the "build-* directories that it finds.
5232 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5233 address in a domain literal was a prefix of an interface address.
5235 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5236 when verifying a sender address, unless rewrite_headers is false.
5238 PH/44 Wrote a long comment about why errors_to addresses are verified as
5239 recipients, not senders.
5241 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5242 the ratelimit ACL was added.
5244 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5246 PH/46 Added extra information about PostgreSQL errors to the error string.
5248 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5249 avoiding the need to install special zones in a real server. This is
5250 backwards compatible; if it can't find the fake resolver, it drops back.
5251 Thus, both old and new test suites can be run.
5253 TF/02 Added util/ratelimit.pl
5255 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5256 clock is set back in time.
5258 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5259 Candler <B.Candler@pobox.com>.
5261 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5262 OK for addresses that are the result of redirections.
5264 PH/48 A number of further additions for the benefit of the new test suite,
5265 including a fake gethostbyname() that interfaces to the fake DNS resolver
5268 TF/06 The fix for widen_domains has also been applied to qualify_single and
5269 search_parents which are the other dnslookup options that can cause
5272 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5275 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5277 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5278 service" if it encountered an MX record with an empty target hostname.
5279 The message is now "an MX or SRV record indicated no SMTP service".
5281 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5282 if the DNS of the sending site is misconfigured. This is quite a
5283 common situation. This change restores the behaviour of treating a
5284 helo verification defer as a failure.
5286 PH/51 If self=fail was set on a router, the bounce message did not include the
5287 actual error message.
5293 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5295 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5296 command itself ended in error, the underlying message about the transport
5297 filter timeout was being overwritten with the pipe command error. Now the
5298 underlying error message should be appended to the second error message.
5300 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5302 PH/02 Reduce the amount of output that "make" produces by default. Full output
5303 can still be requested.
5305 PH/03 The warning log line about a condition test deferring for a "warn" verb
5306 was being output only once per connection, rather than after each
5307 occurrence (because it was using the same function as for successful
5308 "warn" verbs). This seems wrong, so I have changed it.
5310 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5311 it should not have, which might have caused a crash in the right
5312 circumstances, but probably never did.
5314 PH/04 Installed a modified version of Tony Finch's patch to make submission
5315 mode fix the return path as well as the Sender: header line, and to
5316 add a /name= option so that you can make the user's friendly name appear
5319 TF/03 Added the control = fakedefer ACL modifier.
5321 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5322 Mark Lowes for thorough testing.
5324 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5326 TK/03 Merged latest SRS patch from Miles Wilton.
5328 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5329 with the definition in sysexits.h (which is #included earlier).
5330 Fortunately, Exim does not actually use EX_OK. The code used to try to
5331 preserve the sysexits.h value, by assuming that macro definitions were
5332 scanned for macro replacements. I have been disabused of this notion,
5333 so now the code just undefines EX_OK before #including unistd.h.
5335 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5336 in the smtp transport. When a block could not be written in a single
5337 write() function, the timeout was being re-applied to each part-write.
5338 This seems wrong - if the receiver was accepting one byte at a time it
5339 would take for ever. The timeout is now adjusted when this happens. It
5340 doesn't have to be particularly precise.
5342 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5343 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5345 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5346 without a domain if the check to <postmaster@domain> fails.
5348 SC/01 Eximstats: added -xls and the ability to specify output files
5349 (patch written by Frank Heydlauf).
5351 SC/02 Eximstats: use FileHandles for outputting results.
5353 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5355 SC/04 Eximstats: fixed display of large numbers with -nvr option
5357 SC/05 Eximstats: fixed merging of reports with empty tables.
5359 SC/06 Eximstats: added the -include_original_destination flag
5361 SC/07 Eximstats: removed tabs and trailing whitespace.
5363 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5365 TK/06 MBOX spool code: Add real "From " MBOX separator line
5366 so the .eml file is really in mbox format (even though
5367 most programs do not really care). Patch from Alex Miller.
5369 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5370 The latter is generated from $received_to and is only set if the
5371 message has one envelope recipient. SA can use these headers,
5372 obviously out-of-the-box. Patch from Alex Miller.
5374 PH/08 The ${def test on a variable was returning false if the variable's
5375 value was "0", contrary to what the specification has always said!
5376 The result should be true unless the variable is empty.
5378 PH/09 The syntax error of a character other than { following "${if
5379 def:variable_name" (after optional whitespace) was not being diagnosed.
5380 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5381 accidental colon was present, for example, could give incorrect results.
5383 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5384 result is used (not including appendfile, where other changes are about
5387 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5388 This involved changing a lot of size variables from int to off_t. It
5389 should work with maildirs and everything.
5391 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5392 spamd dying while we are connected to it.
5394 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5397 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5398 function for BDB 4.3.
5400 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5402 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5403 the sources for PCRE have been re-arranged and more files are now
5406 PH/15 The code I had for printing potentially long long variables in PH/11
5407 above was not the best (it lost precision). The length of off_t variables
5408 is now inspected at build time, and an appropriate printing format (%ld
5409 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5410 to be "long long int" or "long int". This is needed for the internal
5411 formatting function string_vformat().
5413 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5414 the configuration file to be ":syslog", then the script "guesses" where
5415 the logs files are, rather than using the compiled in default. In our
5416 case the guess is not the same as the compiled default, so the script
5417 suddenly stopped working when I started to use syslog. The patch checks
5418 to see if log_file_path is "". If so, it attempts to read it from exim
5419 with no configuration file to get the compiled in version, before it
5420 falls back to the previous guessing code."
5422 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5423 implementing BATV in an Exim configuration. See NewStuff for the gory
5426 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5427 Makefile that are specific to HP-UX.
5429 PH/18 If the "use_postmaster" option was set for a recipient callout together
5430 with the "random" option, the postmaster address was used as the MAIL
5431 FROM address for the random test, but not for the subsequent recipient
5432 test. It is now used for both.
5434 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5435 patch removes a few documentation additions to RFC 3028, because the
5436 latest draft now contains them. It adds the new en;ascii-case comparator
5437 and a new error check for 8bit text in MIME parts. Comparator and
5438 require names are now matched exactly. I enabled the subaddress
5439 extension, but it is not well tested yet (read: it works for me)."
5441 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5442 rework some of the code of TK/09 above to avoid the hardwired use of
5443 "%lld" and "long long". Replaced the call to snprintf() with a call to
5446 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5447 records point to non-existent hosts", "retry timeout exceeded", and
5448 "retry time not reached for any host after a long failure period".
5450 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5451 experimental DomainKeys support:
5453 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5454 (2) On an error such as an illegally used "control", the wrong name for
5455 the control was given.
5457 These problems did NOT occur unless DomainKeys support was compiled.
5459 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5461 PH/24 Added ${if match_ip condition.
5463 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5464 databases so that it will be absolutely obvious if a crash occurs in the
5465 DB library. This is a regular occurrence (often caused by mis-matched
5468 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5469 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5470 picked up on a user's system that detects such things. There doesn't seem
5471 to be a gcc warning option for this - only an attribute that has to be
5472 put on the function's prototype. It seems that in Fedora Core 4 they have
5473 set this on a number of new functions. No doubt there will be more in due
5476 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5477 specify a transport. However, if an address that was verified by such a
5478 router was the subject of a callout, Exim crashed because it tried to
5479 read the rcpt_include_affixes from the non-existent transport. Now it
5480 just assumes that the setting of that option is false. This bug was
5481 introduced by 4.51/PH/31.
5483 PH/28 Changed -d+all to exclude +memory, because that information is very
5484 rarely of interest, but it makes the output a lot bigger. People tend to
5485 do -d+all out of habit.
5487 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5488 code in os-type was giving problems when libc.so lives in lib64, like on
5491 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5492 aren't the modern standard, and it seems that some systems' include files
5493 don't always have them. Exim was already checking for some of the newer
5494 ones like T_AAAA, and defining it itself. I've added checks for all the
5495 record types that Exim uses.
5497 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5498 not automatically generating a new one, as it is supposed to. This
5499 prevented TLS from working. If the file did exist, but contained invalid
5500 data, a new version was generated, as expected. It was only the case of a
5501 non-existent file that was broken.
5503 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5504 with a change in libdomainkeys > 0.64.
5506 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5507 from DNS. If the selector record carries the flag, it now has
5508 precedence over the domain-wide flag.
5510 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5512 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5513 the use of a port name that isn't defined in /etc/services) occurred, the
5514 message was deferred as in a normal delivery, and thus remained on the
5515 spool, instead of being failed because of the mua_wrapper setting. This
5516 is now fixed, and I tidied up some of the mua_wrapper messages at the
5519 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5520 the messages in a hash of arrays rather than using individual hashes.
5521 This is a bit cleaner and results in dramatic memory savings, albeit
5522 at a slight CPU cost.
5524 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5525 as requested by Marc Sherman.
5527 SC/10 Eximstats: added histograms for user specified patterns as requested
5530 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5532 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5533 fopen() in the content-scanning modules that did not already have it.
5539 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5540 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5542 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5544 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5546 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5547 to test Sieve filters that use "vacation".
5549 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5550 that changes the way the GnuTLS parameters are stored in the cache file.
5551 The new format can be generated externally. For backward compatibility,
5552 if the data in the cache doesn't make sense, Exim assumes it has read an
5553 old-format file, and it generates new data and writes a new file. This
5554 means that you can't go back to an older release without removing the
5557 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5558 work if there are any delivery delays because "one_time" forces the
5559 parent to be marked "delivered", so its unseen clone is never tried
5560 again. For this reason, Exim now forbids the simultaneous setting of
5563 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5564 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5565 entry if you want to know the details. The fix, however, neglected to
5566 consider the case when local delivery batching is involved. The test for
5567 "previously delivered" was not happening when checking to see if an
5568 address could be batched with a previous (undelivered) one; under
5569 certain circumstances this could lead to multiple deliveries to the same
5572 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5573 in its include files, and this causes problems building Exim.
5575 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5576 header_syntax) but Exim was just ignoring anything given after a slash.
5577 In particular, this caused confusion with an attempt to use "verify =
5578 reverse_host_lookup/defer_ok". An error is now given when options are
5579 supplied for verify items that do not have them. (Maybe reverse_host_
5580 lookup should have a defer_ok option, but that's a different point.)
5582 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5583 defined by RFC 821) to 2048, because there were problems with some AUTH
5584 commands, and RFC 1869 says the size should be increased for extended
5585 SMTP commands that take arguments.
5587 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5590 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5591 "unknown" error; now it says that the functionality isn't in the binary.
5593 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5594 an address' error message when a string expansion fails (syntax or
5595 whatever). Otherwise the password may appear in the log. Following change
5596 PH/42 below, there is no longer a chance of it appearing in a bounce
5599 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5601 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5602 of its addresses was ever tried. (Bugzilla bug #2.)
5604 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5605 the result incorrectly in the debug output. (It correctly added a newline
5606 to what was transported.)
5608 TF/01 Added $received_time.
5610 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5611 commented out examples of how to interface to a virus scanner and to
5612 SpamAssassin. Also added commented examples of av_scanner and
5613 spamd_address settings.
5615 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5616 and controls are allowed in which ACLs. There were a couple of minor
5617 errors. Some of the entries in the conditions table (which is a table of
5618 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5619 negation of where the condition IS allowed.
5621 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5623 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5624 header file does not have a version number, so I've had to invent a new
5625 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5626 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5627 radiusclient), but it was contributed by a Radius user.
5629 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5630 files or directories, for OpenSSL.
5632 PH/19 When an Exim process that is running as root has to create an Exim log
5633 file, it does so in a subprocess that runs as exim:exim so as to get the
5634 ownership right at creation (otherwise, other Exim processes might see
5635 the file with the wrong ownership). There was no test for failure of this
5636 fork() call, which would lead to the process getting stuck as it waited
5637 for a non-existent subprocess. Forks do occasionally fail when resources
5638 run out. I reviewed all the other calls to fork(); they all seem to check
5641 PH/20 When checking for unexpected SMTP input at connect time (before writing
5642 the banner), Exim was not dealing correctly with a non-positive return
5643 from the read() function. If the client had disconnected by this time,
5644 the result was a log entry for a synchronization error with an empty
5645 string after "input=" when read() returned zero. If read() returned -1
5646 (an event I could not check), uninitialized data bytes were printed.
5647 There were reports of junk text (parts of files, etc) appearing after
5650 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5652 PH/22 Added support for macro redefinition, and (re)definition in between
5653 driver and ACL definitions.
5655 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5656 forgetting to use the resulting value; it was using the unexpanded value.
5658 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5659 hadn't been configured. The fix is from Juergen Kreileder, who
5660 understands it better than I do:
5662 "Here's what I see happening with three configured cyrus_sasl
5663 authenticators configured (plain, login, cram-md5):
5665 On startup auth_cyrus_sasl_init() gets called for each of these.
5666 This means three calls to sasl_listmech() without a specified mech_list.
5667 => SASL tests which mechs of all available mechs actually work
5668 => three warnings about OTP not working
5669 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5671 With the patch, sasl_listmech() also gets called three times. But now
5672 SASL's mech_list option is set to the server_mech specified in the the
5673 authenticator. Or in other words, the answer from sasl_listmech()
5674 gets limited to just the mech you're testing for (which is different
5676 => the return list contains just 'plain' or 'login', 'cram-md5' or
5677 nothing depending on the value of ob->server_mech.
5679 I've just tested the patch: Authentication still works fine,
5680 unavailable mechs specified in the exim configuration are still
5681 caught, and the auth.log warnings about OTP are gone."
5683 PH/25 When debugging is enabled, the contents of the command line are added
5684 to the debugging output, even when log_selector=+arguments is not
5687 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5688 answer is "GNU", and only if the return is "GNU/something" is the answer
5691 PH/27 $acl_verify_message is now set immediately after the failure of a
5692 verification in an ACL, and so is available in subsequent modifiers. In
5693 particular, the message can be preserved by coding like this:
5695 warn !verify = sender
5696 set acl_m0 = $acl_verify_message
5698 Previously, $acl_verify_message was set only while expanding "message"
5699 and "log_message" when a very denied access.
5701 PH/28 Modified OS/os.c-Linux with
5703 -#ifndef OS_LOAD_AVERAGE
5704 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5706 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5707 nomenclature these days.)
5709 PH/29 Installed patch from the Sieve maintainer that adds the options
5710 sieve_useraddress and sieve_subaddress to the redirect router.
5712 PH/30 In these circumstances:
5713 . Two addresses routed to the same list of hosts;
5714 . First host does not offer TLS;
5715 . First host accepts first address;
5716 . First host gives temporary error to second address;
5717 . Second host offers TLS and a TLS session is established;
5718 . Second host accepts second address.
5719 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5720 and peerdn, if requested) that were in fact used only for the second
5723 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5724 attention to any local part prefix or suffix that was matched by the
5725 router that accepted the address. It now behaves in the same way as it
5726 does for delivery: the affixes are removed from the local part unless
5727 rcpt_include_affixes is set on the transport.
5729 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5730 timeout during the DATA phase of an incoming message.
5732 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5733 applied a patch sanctioned by the Sieve maintainer.
5735 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5736 the uid or gid is negative. A case of a negative gid caused this to be
5737 noticed. The fix allows for either to be negative.
5739 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5740 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5743 PH/36 The change PH/12 above was broken. Fixed it.
5745 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5746 the grounds that routing the same address twice would always produce the
5747 same answer. This might have been true once, but it is certainly no
5748 longer true now. Routing a child address may depend on the previous
5749 routing that produced that child. Some complicated redirection strategies
5750 went wrong when messages had multiple recipients, and made Exim's
5751 behaviour dependent on the order in which the addresses were given.
5753 I have moved the duplicate checking until after the routing is complete.
5754 Exim scans the addresses that are assigned to local and remote
5755 transports, and removes any duplicates. This means that more work will be
5756 done, as duplicates will always all be routed, but duplicates are
5757 presumably rare, so I don't expect this is of any significance.
5759 For deliveries to pipes, files, and autoreplies, the duplicate checking
5760 still happens during the routing process, since they are not going to be
5763 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5764 It corrects a timeout issue with spamd. This is Ian's comment: "The
5765 background is that sometimes spamd either never reads data from a
5766 connection it has accepted, or it never writes response data. The exiscan
5767 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5768 blindly assumes that writes won't block so it may never time out."
5770 PH/39 Allow G after quota size as well as K and M.
5772 PH/40 The value set for $authenticated_id in an authenticator may not contain
5773 binary zeroes or newlines because the value is written to log lines and
5774 to spool files. There was no check on this. Now the value is run through
5775 the string_printing() function so that such characters are converted to
5776 printable escape sequences.
5778 PH/41 $message_linecount is a new variable that contains the total number of
5779 lines in the message. Compare $body_linecount, which is the count for the
5782 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5783 bounce and delay warning messages, except in certain special cases, which
5786 (a) An SMTP error message from a remote host;
5787 (b) A message specified in a :fail: redirection;
5788 (c) A message specified in a "fail" command in a system filter;
5789 (d) A message specified in a FAIL return from the queryprogram router;
5790 (e) A message specified by the cannot_route_message router option.
5792 In these cases only, Exim does include the error details in bounce and
5793 warning messages. There are also a few cases where bland messages such
5794 as "unrouteable address" or "local delivery error" are given.
5796 PH/43 $value is now also set for the "else" part of a ${run expansion.
5798 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5799 being worked on, but at least Exim now implements the latest version to
5802 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5803 process to complete was treated as a delivery failure, a timeout while
5804 writing the message to the pipe was logged, but erroneously treated as a
5805 successful delivery. Such timeouts include transport filter timeouts. For
5806 consistency with the overall process timeout, these timeouts are now
5807 treated as errors, giving rise to delivery failures by default. However,
5808 there is now a new Boolean option for the pipe transport called
5809 timeout_defer, which, if set TRUE, converts the failures into defers for
5810 both kinds of timeout. A transport filter timeout is now identified in
5813 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5814 systems where "make" and "gmake" are different, calling "gmake" at top
5815 level broke things. I've arranged for the value of $(MAKE) to be passed
5816 from the Makefile to this script so that it can call the same version of
5820 A note about Exim versions 4.44 and 4.50
5821 ----------------------------------------
5823 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5824 changes of various kinds. As a consequence, a big documentation update was
5825 needed. This delayed the release for rather longer than seemed good, especially
5826 in the light of a couple of (minor) security issues. Therefore, the changes
5827 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5828 release. So 4.44 and 4.50 are in effect two different branches that both start
5831 I have left the 4.50 change log unchanged; it contains all the changes since
5832 4.43. The change log for 4.44 is below; many of its items are identical to
5833 those for 4.50. This seems to be the most sensible way to preserve the
5834 historical information.
5840 1. Minor wording change to the doc/README.SIEVE file.
5842 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5843 computation of the current number of files was incorrect.
5845 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5846 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5849 4. Give more explanation in the error message when the command for a transport
5850 filter fails to execute.
5852 5. There are several places where Exim runs a non-Exim command in a
5853 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5854 was being done only for the command run by the queryprogram router. It is
5855 now done for all such subprocesses. The other cases are: ${run, transport
5856 filters, and the commands run by the lmtp and pipe transports.
5858 6. Added CONFIGURE_GROUP build-time option.
5860 7. Some older OS have a limit of 256 on the maximum number of file
5861 descriptors. Exim was using setrlimit() to set 1000 as a large value
5862 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5863 systems. I've change it so that if it can't get 1000, it tries for 256.
5865 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5866 was an oversight, and furthermore, ever since the addition of extra
5867 controls (e.g. 4.43/32), the checks on when to allow different forms of
5868 "control" were broken. There should now be diagnostics for all cases when a
5869 control that does not make sense is encountered.
5871 9. Added the /retain_sender option to "control=submission".
5873 10. $recipients is now available in the predata ACL (oversight).
5875 11. Tidy the search cache before the fork to do a delivery from a message
5876 received from the command line. Otherwise the child will trigger a lookup
5877 failure and thereby defer the delivery if it tries to use (for example) a
5878 cached ldap connection that the parent has called unbind on.
5880 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5881 of $address_data from the recipient verification was clobbered by the
5882 sender verification.
5884 13. The value of address_data from a sender verification is now available in
5885 $sender_address_data in subsequent conditions in the ACL statement.
5887 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5889 15. Added a new option "connect=<time>" to callout options, to set a different
5892 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5893 was its contents. (It was OK if the option was not defined at all.)
5895 17. A "Completed" log line is now written for messages that are removed from
5896 the spool by the -Mrm option.
5898 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5899 information about exactly what failed.
5901 19. Added -dd to debug only the daemon process.
5903 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5904 handles timeouts, both on the server side and network timeouts. Renamed the
5905 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5907 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5908 It is now set to "smtps".
5910 22. $host_address is now set to the target address during the checking of
5911 ignore_target_hosts.
5913 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5914 being passed; this would have caused $sender_host_name to have been used if
5915 matching the list had actually called for a host name (not very likely,
5916 since this list is usually IP addresses). A host name is now passed as
5919 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5920 code that specifies a non-restarting handler (typically sigaction() in
5921 modern systems) in an attempt to fix a rare and obscure crash bug.
5923 25. Narrowed the window for a race in the daemon that could cause it to ignore
5924 SIGCHLD signals. This is not a major problem, because they are used only to
5925 wake it up if nothing else does.
5927 26. A malformed maildirsize file could cause Exim to calculate negative values
5928 for the mailbox size or file count. Odd effects could occur as a result.
5929 The maildirsize information is now recalculated if the size or filecount
5932 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5933 support for a long time. Removed HAVE_SYS_VFS_H.
5935 28. Installed the latest version of exipick from John Jetmore.
5937 29. In an address list, if the pattern was not a regular expression, an empty
5938 subject address (from a bounce message) matched only if the pattern was an
5939 empty string. Non-empty patterns were not even tested. This was the wrong
5940 because it is perfectly reasonable to use an empty address as part of a
5941 database query. An empty address is now tested by patterns that are
5942 lookups. However, all the other forms of pattern expect the subject to
5943 contain a local part and a domain, and therefore, for them, an empty
5944 address still always fails if the pattern is not itself empty.
5946 30. Exim went into a mad DNS loop when attempting to do a callout where the
5947 host was specified on an smtp transport, and looking it up yielded more
5948 than one IP address.
5950 31. Re-factored the code for checking spool and log partition space into a
5951 function that finds that data and another that does the check. The former
5952 is then used to implement four new variables: $spool_space, $log_space,
5953 $spool_inodes, and $log_inodes.
5955 32. The RFC2047 encoding function was originally intended for short strings
5956 such as real names; it was not keeping to the 75-character limit for
5957 encoded words that the RFC imposes. It now respects the limit, and
5958 generates multiple encoded words if necessary. To be on the safe side, I
5959 have increased the buffer size for the ${rfc2047: expansion operator from
5962 33. It is now permitted to omit both strings after an "if" condition; if the
5963 condition is true, the result is "true". As before, when the second string
5964 is omitted, a false condition yields an empty string. This makes it less
5965 cumbersome to write custom ACL and router conditions.
5967 34. Failure to deliver a bounce message always caused it to be frozen, even if
5968 there was an errors_to setting on the router. The errors_to setting is now
5971 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5972 canonical form (fully expanded) before being placed in
5973 $sender_host_address.
5975 36. The table in the code that translates DNS record types into text (T_A to
5976 "A" for instance) was missing entries for NS and CNAME. It is just possible
5977 that this could have caused confusion if both these types were looked up
5978 for the same domain, because the text type is used as part of Exim's
5979 per-process caching. But the chance of anyone hitting this buglet seems
5982 37. The dnsdb lookup has been extended in a number of ways.
5984 (1) There is a new type, "zns", which walks up the domain tree until it
5985 finds some nameserver records. It should be used with care.
5987 (2) There is a new type, "mxh", which is like "mx" except that it returns
5988 just the host names, not the priorities.
5990 (3) It is now possible to give a list of domains (or IP addresses) to be
5991 looked up. The behaviour when one of the lookups defers can be
5992 controlled by a keyword.
5994 (4) It is now possible to specify the separator character for use when
5995 multiple records are returned.
5997 38. The dnslists ACL condition has been extended: it is now possible to supply
5998 a list of IP addresses and/or domains to be looked up in a particular DNS
6001 39. Added log_selector=+queue_time_overall.
6003 40. When running the queue in the test harness, wait just a tad after forking a
6004 delivery process, to get repeatability of debugging output.
6006 41. Include certificate and key file names in error message when GnuTLS fails
6007 to set them up, because the GnuTLS error message doesn't include the name
6008 of the failing file when there is a problem reading it.
6010 42. Allow both -bf and -bF in the same test run.
6012 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6014 44. The "Exiscan patch" is now merged into the mainline Exim source.
6016 45. Sometimes the final signoff response after QUIT could fail to get
6017 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6018 < 0 before doing a fflush(). This bug looks as though it goes back to the
6019 introduction of TLS in release 3.20, but "sometimes" must have been rare
6020 because the tests only now provoked it.
6022 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6023 (this can affect the format of dates).
6025 47. exim_tidydb, when checking for the continued existence of a message for
6026 which it has found a message-specific retry record, was not finding
6027 messages that were in split spool directories. Consequently, it was
6028 deleting retry records that should have stayed in existence.
6030 48. Steve fixed some bugs in eximstats.
6032 49. The SPA authentication driver was not abandoning authentication and moving
6033 on to the next authenticator when an expansion was forced to fail,
6034 contradicting the general specification for all authenticators. Instead it
6035 was generating a temporary error. It now behaves as specified.
6037 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6038 (the order specifies the preference for clients). The order is now AES256,
6039 AES128, 3DES, ARCFOUR128.
6041 51. Small patch to Sieve code - explicitly set From: when generating an
6044 52. Exim crashed if a remote delivery caused a very long error message to be
6045 recorded - for instance if somebody sent an entire SpamAssassin report back
6046 as a large number of 550 error lines. This bug was coincidentally fixed by
6047 increasing the size of one of Exim's internal buffers (big_buffer) that
6048 happened as part of the Exiscan merge. However, to be on the safe side, I
6049 have made the code more robust (and fixed the comments that describe what
6052 53. Now that there can be additional text after "Completed" in log lines (if
6053 the queue_time_overall log selector is set), a one-byte patch to exigrep
6054 was needed to allow it to recognize "Completed" as not the last thing in
6057 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6058 patch that reportedly fixes this has been added. I am not expert enough to
6059 create a test for it. This is what the patch creator wrote:
6061 "I found a little strange behaviour of ldap code when working with
6062 Windows 2003 AD Domain, where users was placed in more than one
6063 Organization Units. When I tried to give exim partial DN, the exit code
6064 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6065 But simultaneously result of request was absolutely normal ldap result,
6066 so I produce this patch..."
6068 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6069 REFERENCE, so I have modified the code to exclude the patch when that macro
6072 55. Some experimental protocols are using DNS PTR records for new purposes. The
6073 keys for these records are domain names, not reversed IP addresses. The
6074 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6075 leaves it alone. Component reversal etc. now happens only for IP addresses.
6078 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6080 57. Double the size of the debug message buffer (to 2048) so that more of very
6081 long debug lines gets shown.
6083 58. The exicyclog utility now does better if the number of log files to keep
6084 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6086 59. Two changes related to the smtp_active_hostname option:
6088 (1) $smtp_active_hostname is now available as a variable.
6089 (2) The default for smtp_banner uses $smtp_active_hostname instead
6090 of $primary_hostname.
6092 60. The host_aton() function is supposed to be passed a string that is known
6093 to be a valid IP address. However, in the case of IPv6 addresses, it was
6094 not checking this. This is a hostage to fortune. Exim now panics and dies
6095 if the condition is not met. A case was found where this could be provoked
6096 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6097 components; fortuitously, this particular loophole had already been fixed
6098 by change 4.50/55 above.
6100 If there are any other similar loopholes, the new check in host_aton()
6101 itself should stop them being exploited. The report I received stated that
6102 data on the command line could provoke the exploit when Exim was running as
6103 exim, but did not say which command line option was involved. All I could
6104 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6105 running as the user.
6108 61. There was a buffer overflow vulnerability in the SPA authentication code
6109 (which came originally from the Samba project). I have added a test to the
6110 spa_base64_to_bits() function which I hope fixes it.
6113 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6114 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6116 63. The daemon start-up calls getloadavg() while still root for those OS that
6117 need the first call to be done as root, but it missed one case: when
6118 deliver_queue_load_max is set with deliver_drop_privilege. This is
6119 necessary for the benefit of the queue runner, because there is no re-exec
6120 when deliver_drop_privilege is set.
6122 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6123 This has been fixed.
6125 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6126 was in use, was not putting the data itself into the right store pool;
6127 consequently, it could be overwritten for a subsequent message in the same
6128 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6131 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6133 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6134 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6135 the function were treating the return as a boolean value, which happened to
6136 work because 0=false and not-0=true, but is not correct code.
6138 68. The host_aton() function was not handling scoped IPv6 addresses (those
6139 with, for example, "%eth0" on the end) correctly.
6141 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6142 negated items (that is, ~something) in unsigned ints. Some compilers
6143 apparently mutter when there is no cast.
6145 70. If an address verification called from an ACL failed, and did not produce a
6146 user-specific message (i.e. there was only a "system" message), nothing was
6147 put in $acl_verify_message. In this situation, it now puts the system
6150 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6151 session; change 4.31/43 added the unwanted input to the log line - except
6152 that it did not do this in the start of session case. It now does.
6154 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6155 This is wrong and can cause the other end to generate a synchronization
6156 error if it is another Exim or anything else that does the synchronization
6157 check. A QUIT command is no longer sent after a timeout.
6159 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6160 during host lookups.
6162 74. The defer_ok option of callout verification was not working if it was used
6163 when verifying addresses in header lines, that is, for this case:
6165 verify = header_sender/callout=defer_ok
6167 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6168 those file descriptors could be used for SMTP connections. If anything
6169 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6170 could be sent to the SMTP client, causing chaos. The daemon now opens
6171 stdin, stdout, and stderr to /dev/null when it puts itself into the
6174 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6175 log by default. The user can override this with suitable Perl magic.
6177 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6178 the log message when discard triggers, was not working for the DATA ACL or
6179 for the non-SMTP ACL.
6181 78. Error message wording change in sieve.c.
6183 79. If smtp_accept_max_per_host was set, the number of connections could be
6184 restricted to fewer than expected, because the daemon was trying to set up
6185 a new connection before checking whether the processes handling previous
6186 connections had finished. The check for completed processes is now done
6187 earlier. On busy systems, this bug wouldn't be noticed because something
6188 else would have woken the daemon, and it would have reaped the completed
6191 80. If a message was submitted locally by a user whose login name contained one
6192 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6193 It caused a spool format error. I have fixed the spool reading code. A
6194 related problem was that the "from" clause in the Received: line became
6195 illegal because of the space(s). It is now covered by ${quote_local_part.
6197 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6200 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6202 83. Patch from Sieve maintainer to fix unterminated string problem in
6203 "vacation" handling.
6205 84. Some minor changes to the Linux configuration files to help with other
6206 OS variants using glibc.
6208 85. One more patch for Sieve to update vacation handling to latest spec.
6211 ----------------------------------------------------
6212 See the note above about the 4.44 and 4.50 releases.
6213 ----------------------------------------------------
6219 1. Change 4.43/35 introduced a bug that caused file counts to be
6220 incorrectly computed when quota_filecount was set in an appendfile
6223 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6224 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6227 3. Give more explanation in the error message when the command for a transport
6228 filter fails to execute.
6230 4. There are several places where Exim runs a non-Exim command in a
6231 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6232 was being done only for the command run by the queryprogram router. It is
6233 now done for all such subprocesses. The other cases are: ${run, transport
6234 filters, and the commands run by the lmtp and pipe transports.
6236 5. Some older OS have a limit of 256 on the maximum number of file
6237 descriptors. Exim was using setrlimit() to set 1000 as a large value
6238 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6239 systems. I've change it so that if it can't get 1000, it tries for 256.
6241 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6242 was an oversight, and furthermore, ever since the addition of extra
6243 controls (e.g. 4.43/32), the checks on when to allow different forms of
6244 "control" were broken. There should now be diagnostics for all cases when a
6245 control that does not make sense is encountered.
6247 7. $recipients is now available in the predata ACL (oversight).
6249 8. Tidy the search cache before the fork to do a delivery from a message
6250 received from the command line. Otherwise the child will trigger a lookup
6251 failure and thereby defer the delivery if it tries to use (for example) a
6252 cached ldap connection that the parent has called unbind on.
6254 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6255 of $address_data from the recipient verification was clobbered by the
6256 sender verification.
6258 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6259 was its contents. (It was OK if the option was not defined at all.)
6261 11. A "Completed" log line is now written for messages that are removed from
6262 the spool by the -Mrm option.
6264 12. $host_address is now set to the target address during the checking of
6265 ignore_target_hosts.
6267 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6268 being passed; this would have caused $sender_host_name to have been used if
6269 matching the list had actually called for a host name (not very likely,
6270 since this list is usually IP addresses). A host name is now passed as
6273 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6274 code that specifies a non-restarting handler (typically sigaction() in
6275 modern systems) in an attempt to fix a rare and obscure crash bug.
6277 15. Narrowed the window for a race in the daemon that could cause it to ignore
6278 SIGCHLD signals. This is not a major problem, because they are used only to
6279 wake it up if nothing else does.
6281 16. A malformed maildirsize file could cause Exim to calculate negative values
6282 for the mailbox size or file count. Odd effects could occur as a result.
6283 The maildirsize information is now recalculated if the size or filecount
6286 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6287 support for a long time. Removed HAVE_SYS_VFS_H.
6289 18. Updated exipick to current release from John Jetmore.
6291 19. Allow an empty sender to be matched against a lookup in an address list.
6292 Previously the only cases considered were a regular expression, or an
6295 20. Exim went into a mad DNS lookup loop when doing a callout where the
6296 host was specified on the transport, if the DNS lookup yielded more than
6299 21. The RFC2047 encoding function was originally intended for short strings
6300 such as real names; it was not keeping to the 75-character limit for
6301 encoded words that the RFC imposes. It now respects the limit, and
6302 generates multiple encoded words if necessary. To be on the safe side, I
6303 have increased the buffer size for the ${rfc2047: expansion operator from
6306 22. Failure to deliver a bounce message always caused it to be frozen, even if
6307 there was an errors_to setting on the router. The errors_to setting is now
6310 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6311 canonical form (fully expanded) before being placed in
6312 $sender_host_address.
6314 24. Updated eximstats to version 1.33
6316 25. Include certificate and key file names in error message when GnuTLS fails
6317 to set them up, because the GnuTLS error message doesn't include the name
6318 of the failing file when there is a problem reading it.
6320 26. Expand error message when OpenSSL has problems setting up cert/key files.
6323 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6324 (this can affect the format of dates).
6326 28. exim_tidydb, when checking for the continued existence of a message for
6327 which it has found a message-specific retry record, was not finding
6328 messages that were in split spool directories. Consequently, it was
6329 deleting retry records that should have stayed in existence.
6331 29. eximstats updated to version 1.35
6332 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6333 1.35 - bugfix such that pie charts by volume are generated correctly
6335 30. The SPA authentication driver was not abandoning authentication and moving
6336 on to the next authenticator when an expansion was forced to fail,
6337 contradicting the general specification for all authenticators. Instead it
6338 was generating a temporary error. It now behaves as specified.
6340 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6341 (the order specifies the preference for clients). The order is now AES256,
6342 AES128, 3DES, ARCFOUR128.
6344 31. Small patch to Sieve code - explicitly set From: when generating an
6347 32. Exim crashed if a remote delivery caused a very long error message to be
6348 recorded - for instance if somebody sent an entire SpamAssassin report back
6349 as a large number of 550 error lines. This bug was coincidentally fixed by
6350 increasing the size of one of Exim's internal buffers (big_buffer) that
6351 happened as part of the Exiscan merge. However, to be on the safe side, I
6352 have made the code more robust (and fixed the comments that describe what
6355 33. Some experimental protocols are using DNS PTR records for new purposes. The
6356 keys for these records are domain names, not reversed IP addresses. The
6357 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6358 leaves it alone. Component reversal etc. now happens only for IP addresses.
6361 34. The host_aton() function is supposed to be passed a string that is known
6362 to be a valid IP address. However, in the case of IPv6 addresses, it was
6363 not checking this. This is a hostage to fortune. Exim now panics and dies
6364 if the condition is not met. A case was found where this could be provoked
6365 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6366 components; fortuitously, this particular loophole had already been fixed
6367 by change 4.50/55 or 4.44/33 above.
6369 If there are any other similar loopholes, the new check in host_aton()
6370 itself should stop them being exploited. The report I received stated that
6371 data on the command line could provoke the exploit when Exim was running as
6372 exim, but did not say which command line option was involved. All I could
6373 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6374 running as the user.
6377 35. There was a buffer overflow vulnerability in the SPA authentication code
6378 (which came originally from the Samba project). I have added a test to the
6379 spa_base64_to_bits() function which I hope fixes it.
6382 36. The daemon start-up calls getloadavg() while still root for those OS that
6383 need the first call to be done as root, but it missed one case: when
6384 deliver_queue_load_max is set with deliver_drop_privilege. This is
6385 necessary for the benefit of the queue runner, because there is no re-exec
6386 when deliver_drop_privilege is set.
6388 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6389 was in use, was not putting the data itself into the right store pool;
6390 consequently, it could be overwritten for a subsequent message in the same
6391 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6394 38. Sometimes the final signoff response after QUIT could fail to get
6395 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6396 < 0 before doing a fflush(). This bug looks as though it goes back to the
6397 introduction of TLS in release 3.20, but "sometimes" must have been rare
6398 because the tests only now provoked it.
6404 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6405 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6406 Now it returns an integer. A number of calls were still expecting a T/F
6407 return. Fortuitously, in all cases, the tests worked in OK situations,
6408 which is the norm. However, things would have gone wrong on any write
6409 failures on the smtp file descriptor. This function is used when sending
6410 messages over SMTP and also when doing verify callouts.
6412 2. When Exim is called to do synchronous delivery of a locally submitted
6413 message (the -odf or -odi options), it no longer closes stderr before doing
6416 3. Implemented the mua_wrapper option.
6418 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6420 5. Implemented the functions header_remove(), header_testname(),
6421 header_add_at_position(), and receive_remove_recipient(), and exported them
6424 6. If an ACL "warn" statement specified the addition of headers, Exim already
6425 inserted X-ACL-Warn: at the start if there was no header name. However, it
6426 was not making this test for the second and subsequent header lines if
6427 there were newlines in the string. This meant that an invalid header could
6428 be inserted if Exim was badly configured.
6430 7. Allow an ACL "warn" statement to add header lines at the start or after all
6431 the Received: headers, as well as at the end.
6433 8. Added the rcpt_4xx retry error code.
6435 9. Added postmaster_mailfrom=xxx to callout verification option.
6437 10. Added mailfrom=xxxx to the callout verification option, for verify=
6440 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6441 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6443 12. Inserted some casts to stop certain compilers warning when using pointer
6444 differences as field lengths or precisions in printf-type calls (mostly
6445 affecting debugging statements).
6447 13. Added optional readline() support for -be (dynamically loaded).
6449 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6450 same clock tick as a message's arrival, so that its received time was the
6451 same as the "first fail" time on the retry record, and that message
6452 remained on the queue past the ultimate address timeout, every queue runner
6453 would try a delivery (because it was past the ultimate address timeout) but
6454 after another failure, the ultimate address timeout, which should have then
6455 bounced the address, did not kick in. This was a "< instead of <=" error;
6456 in most cases the first failure would have been in the next clock tick
6457 after the received time, and all would be well.
6459 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6460 being recognized when the domain list was tested by the match_domain
6461 condition in an expansion string.
6463 16. Added the ${str2b64: operator.
6465 17. Exim was always calling setrlimit() to set a large limit for the number of
6466 processes, without checking whether the existing limit was already
6467 adequate. (It did check for the limit on file descriptors.) Furthermore,
6468 errors from getrlimit() and setrlimit() were being ignored. Now they are
6469 logged to the main and panic logs, but Exim does carry on, to try to do its
6470 job under whatever limits there are.
6472 18. Imported PCRE 5.0.
6474 19. Trivial typo in log message " temporarily refused connection" (the leading
6477 20. If the log selector return_path_on_delivery was set and an address was
6478 redirected to /dev/null, the delivery process crashed because it assumed
6479 that a return path would always be set for a "successful" delivery. In this
6480 case, the whole delivery is bypassed as an optimization, and therefore no
6483 21. Internal re-arrangement: the function for sending a challenge and reading
6484 a response while authentication was assuming a zero-terminated challenge
6485 string. It's now changed to take a pointer and a length, to allow for
6486 binary data in such strings.
6488 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6490 23. Exim was not respecting finduser_retries when seeking the login of the
6491 uid under which it was called; it was always trying 10 times. (The default
6492 setting of finduser_retries is zero.) Also, it was sleeping after the final
6493 failure, which is pointless.
6495 24. Implemented tls_on_connect_ports.
6497 25. Implemented acl_smtp_predata.
6499 26. If the domain in control=submission is set empty, Exim assumes that the
6500 authenticated id is a complete email address when it generates From: or
6501 Sender: header lines.
6503 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6504 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6505 chown and chgrp in /bin and hostname in /usr/bin.
6507 28. Exim was keeping the "process log" file open after each use, just as it
6508 does for the main log. This opens the possibility of it remaining open for
6509 long periods when the USR1 signal hits a daemon. Occasional processlog
6510 errors were reported, that could have been caused by this. Anyway, it seems
6511 much more sensible not to leave this file open at all, so that is what now
6514 29. The long-running daemon process does not normally write to the log once it
6515 has entered its main loop, and it closes the log before doing so. This is
6516 so that log files can straightforwardly be renamed and moved. However,
6517 there are a couple of unusual error situations where the daemon does write
6518 log entries, and I had neglected to close the log afterwards.
6520 30. The text of an SMTP error response that was received during a remote
6521 delivery was being truncated at 512 bytes. This is too short for some of
6522 the long messages that one sometimes sees. I've increased the limit to
6525 31. It is now possible to make retry rules that apply only when a message has a
6526 specific sender, in particular, an empty sender.
6528 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6529 it possible to be selective about when SMTP synchronization is enforced.
6531 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6533 32. Implemented hosts_connection_nolog.
6535 33. Added an ACL for QUIT.
6537 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6540 35. Added mailbox_size and mailbox_filecount to appendfile.
6542 36. Added control = no_multiline_responses to ACLs.
6544 37. There was a bug in the logic of the code that waits for the clock to tick
6545 in the case where the clock went backwards by a substantial amount such
6546 that the microsecond fraction of "now" was more than the microsecond
6547 fraction of "then" (but the whole seconds number was less).
6549 38. Added support for the libradius Radius client library this is found on
6550 FreeBSD (previously only the radiusclient library was supported).
6556 1. When certain lookups returned multiple values in the form name=value, the
6557 quoting of the values was not always being done properly. Specifically:
6558 (a) If the value started with a double quote, but contained no whitespace,
6560 (b) If the value contained whitespace other than a space character (i.e.
6561 tabs or newlines or carriage returns) it was not quoted.
6562 This fix has been applied to the mysql and pgsql lookups by writing a
6563 separate quoting function and calling it from the lookup code. The fix
6564 should probably also be applied to nisplus, ibase and oracle lookups, but
6565 since I cannot test any of those, I have not disturbed their existing code.
6567 2. A hit in the callout cache for a specific address caused a log line with no
6568 reason for rejecting RCPT. Now it says "Previous (cached) callout
6569 verification failure".
6571 3. There was an off-by-one bug in the queryprogram router. An over-long
6572 return line was truncated at 256 instead of 255 characters, thereby
6573 overflowing its buffer with the terminating zero. As well as fixing this, I
6574 have increased the buffer size to 1024 (and made a note to document this).
6576 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6577 when Exim is waiting for an SMTP response from a remote server, Exim
6578 restarts its select() call on the socket, thereby resetting its timeout.
6579 This is not a problem when such interrupts are rare. Somebody set up a cron
6580 job to run exiwhat every 2 minutes, which is less than the normal select()
6581 timeout (5 or 10 minutes). This meant that the select() timeout never
6582 kicked in because it was always reset. I have fixed this by comparing the
6583 time when an interrupt arrives with the time at the start of the first call
6584 to select(). If more time than the timeout has elapsed, the interrupt is
6585 treated as a timeout.
6587 5. Some internal re-factoring in preparation for the addition of Sieve
6588 extensions (by MH). In particular, the "personal" test is moved to a
6589 separate function, and given an option for scanning Cc: and Bcc: (which is
6590 not set for Exim filters).
6592 6. When Exim created an email address using the login of the caller as the
6593 local part (e.g. when creating a From: or Sender: header line), it was not
6594 quoting the local part when it contained special characters such as @.
6596 7. Installed new OpenBSD configuration files.
6598 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6599 try to make them clearer.
6601 9. Callout options, other than the timeout value, were being ignored when
6602 verifying sender addresses in header lines. For example, when using
6604 verify = header_sender/callout=no_cache
6606 the cache was (incorrectly) being used.
6608 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6609 only the Cygwin environment.
6611 11. When return_path_on_delivery was set as a log selector, if different remote
6612 addresses in the same message used different return paths and parallel
6613 remote delivery occurred, the wrong values would sometimes be logged.
6614 (Whenever a remote delivery process finished, the return path value from
6615 the most recently started remote delivery process was logged.)
6617 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6618 lines when AUTH and/or TLS are in use. This is the "received protocol"
6619 field. Exim used to use "asmtp" for authenticated SMTP, without any
6620 indication (in the protocol name) for TLS use. Now it follows the RFC and
6621 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6622 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6623 names appear in log lines as well as in Received: header lines.
6625 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6626 extensions, and comparison tests, and to fix some bugs.
6628 14. Changes to the "personal" filter test:
6630 (1) The test was buggy in that it was just doing the equivalent of
6631 "contains" tests on header lines. For example, if a user's address was
6632 anne@some.where, the "personal" test would incorrectly be true for
6634 To: susanne@some.where
6636 This test is now done by extracting each address from the header in turn,
6637 and checking the entire address. Other tests that are part of "personal"
6638 are now done using regular expressions (for example, to check local parts
6639 of addresses in From: header lines).
6641 (2) The list of non-personal local parts in From: addresses has been
6642 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6643 taken from the Sieve specification recommendations.
6645 (3) If the message contains any header line starting with "List-" it is
6646 treated as non-personal.
6648 (4) The test for "circular" in the Subject: header line has been removed
6649 because it now seems ill-conceived.
6651 15. Minor typos in src/EDITME comments corrected.
6653 16. Installed latest exipick from John Jetmore.
6655 17. If headers_add on a router specified a text string that was too long for
6656 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6657 of string_sprintf() is now avoided.
6659 18. $message_body_size was not set (it was always zero) when running the DATA
6660 ACL and the local_scan() function.
6662 19. For the "mail" command in an Exim filter, no default was being set for
6663 the once_repeat time, causing a random time value to be used if "once" was
6664 specified. (If the value happened to be <= 0, no repeat happened.) The
6665 default is now 0s, meaning "never repeat". The "vacation" command was OK
6666 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6667 (I found it when inspecting the code).
6669 20. There is now an overall timeout for performing a callout verification. It
6670 defaults to 4 times the callout timeout, which applies to individual SMTP
6671 commands during the callout. The overall timeout applies when there is more
6672 than one host that can be tried. The timeout is checked before trying the
6673 next host. This prevents very long delays if there are a large number of
6674 hosts and all are timing out (e.g. when the network connections are timing
6675 out). The value of the overall timeout can be changed by specifying an
6676 additional sub-option for "callout", called "maxwait". For example:
6678 verify = sender/callout=5s,maxwait=20s
6680 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6681 to the end before writing, but this should make it even safer).
6683 22. Exim was forgetting that it had advertised PIPELINING for the second and
6684 subsequent messages on an SMTP connection. It was also not resetting its
6685 memory on STARTTLS and an internal HELO.
6687 23. When Exim logs an SMTP synchronization error within a session, it now
6688 records whether PIPELINING has been advertised or not.
6690 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6691 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6692 rather than long int.
6694 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6696 26. Added the never_mail option to autoreply.
6702 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6703 crash if the getsockname() call failed; this can happen if a connection is
6704 closed very soon after it is established. The problem was simply in the
6705 order in which certain operations were done, causing Exim to try to write
6706 to the SMTP stream before it had set up the file descriptor. The bug has
6707 been fixed by making things happen in the correct order.
6713 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6714 before the connection was closed, thus losing the rejection response.
6716 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6717 some early Solaris releases, but causes trouble in current releases where
6718 socklen_t is defined.
6720 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6723 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6726 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6727 that releasing the top of it at the end releases what was used for sub-
6728 expansions (unless the block got too big). However, discard this block if
6729 the first thing is a variable or header, so that we can use its block when
6730 it is dynamic (useful for very large $message_headers, for example).
6732 6. Lookups now cache *every* query, not just the most recent. A new, separate
6733 store pool is used for this. It can be recovered when all lookup caches are
6734 flushed. Lookups now release memory at the end of their result strings.
6735 This has involved some general refactoring of the lookup sources.
6737 7. Some code has been added to the store_xxx() functions to reduce the amount
6738 of flapping under certain conditions.
6740 8. log_incoming_interface used to affect only the <= reception log lines. Now
6741 it causes the local interface and port to be added to several more SMTP log
6742 lines, for example "SMTP connection from", and rejection lines.
6744 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6746 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6748 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6749 could be overwritten at the end of the current message (or the start of a
6750 new message if it was set in a HELO ACL). The value is now preserved for
6751 the duration of the SMTP connection.
6753 12. If a transport had a headers_rewrite setting, and a matching header line
6754 contained an unqualified address, that address was qualified, even if it
6755 did not match any rewriting rules. The underlying bug was that the values
6756 of the flags that permit the existence of unqualified sender and recipient
6757 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6758 non-local messages, and by -bnq for local messages) were not being
6759 preserved with the message after it was received.
6761 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6762 "next input=" as part of the text comprising the host identity instead of
6763 the correct text. The code was using the same buffer for two different
6764 strings. However, depending on which order the printing function evaluated
6765 its arguments, the bug did not always show up. Under Linux, for example, my
6766 test suite worked just fine.
6768 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6769 doesn't work with some older versions of Perl. It has been changed to "my",
6770 which in any case is probably the better facility to use.
6772 15. A really picky compiler found some instances of statements for creating
6773 error messages that either had too many or two few arguments for the format
6776 16. The size of the buffer for calls to the DNS resolver has been increased
6777 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6778 for addresses that have a lot of PTR records. This alleviates a problem; it
6779 does not fully solve it.
6781 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6782 buffer now truncates the list and logs the incident, which is the same
6783 action as happens when Exim is looking up a host name and its aliases.
6784 Previously in this situation something unpredictable would happen;
6785 sometimes it was "internal error: store_reset failed".
6787 18. If a server dropped the connection unexpectedly when an Exim client was
6788 using GnuTLS and trying to read a response, the client delivery process
6789 crashed while trying to generate an error log message.
6791 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6792 string, for example:
6794 warn message = H1: something\nH2: something
6796 the text was added as a single header line from Exim's point of view
6797 though it ended up OK in the delivered message. However, searching for the
6798 second and subsequent header lines using $h_h2: did not work. This has been
6799 fixed. Similarly, if a system filter added multiple headers in this way,
6800 the routers could not see them.
6802 20. Expanded the error message when iplsearch is called with an invalid key to
6803 suggest using net-iplsearch in a host list.
6805 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6806 ACLs are no longer actually imposed (and a message to that effect is
6809 22. If a "gecos" field in a passwd entry contained escaped characters, in
6810 particular, if it contained a \" sequence, Exim got it wrong when building
6811 a From: or a Sender: header from that name. A second bug also caused
6812 incorrect handling when an unquoted " was present following a character
6813 that needed quoting.
6815 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6816 was not being matched caselessly.
6818 24. Arranged for all hyphens in the exim.8 source to be escaped with
6821 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6822 properly. Recipient callout cache records were still being keyed to include
6823 the sender, even when use_sender was set false. This led to far more
6824 callouts that were necessary. The sender is no longer included in the key
6825 when use_sender is false.
6827 26. Added "control = submission" modifier to ACLs.
6829 27. Added the ${base62d: operator to decode base 62 numbers.
6831 28. dnsdb lookups can now access SRV records.
6833 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6834 the configuration file.
6836 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6837 (-v) mode. This makes the output for a verbose queue run more intelligible.
6839 31. Added a use_postmaster feature to recipient callouts.
6841 32. Added the $body_zerocount variable, containing the number of binary zero
6842 bytes in the message body.
6844 33. The time of last modification of the "new" subdirectory is now used as the
6845 "mailbox time last read" when there is a quota error for a maildir
6848 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6850 35. Added +ignore_unknown as a special item in host lists.
6852 36. Code for decoding IPv6 addresses in host lists is now included, even if
6853 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6854 address was recognized as an IP address, but was then not correctly decoded
6855 into binary, causing unexpected and incorrect effects when compared with
6862 1. Very minor rewording of debugging text in manualroute to say "list of
6863 hosts" instead of "hostlist".
6865 2. If verify=header_syntax was set, and a header line with an unqualified
6866 address (no domain) and a large number of spaces between the end of the
6867 name and the colon was received, the reception process suffered a buffer
6868 overflow, and (when I tested it) crashed. This was caused by some obsolete
6869 code that should have been removed. The fix is to remove it!
6871 3. When running in the test harness, delay a bit after writing a bounce
6872 message to get a bit more predictability in the log output.
6874 4. Added a call to search_tidyup() just before forking a reception process. In
6875 theory, someone could use a lookup in the expansion of smtp_accept_max_
6876 per_host which, without the tidyup, could leave open a database connection.
6878 5. Added the variables $recipient_data and $sender_data which get set from a
6879 lookup success in an ACL "recipients" or "senders" condition, or a router
6880 "senders" option, similar to $domain_data and $local_part_data.
6882 6. Moved the writing of debug_print from before to after the "senders" test
6885 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6886 problems for message scanning, either using a data ACL, or using
6887 local_scan() because the Received: header was not generated till after they
6888 were called (in order to set the time as the time of reception completion).
6889 I have revised the way this works. The header is now generated after the
6890 body is received, but before the ACL or local_scan() are called. After they
6891 are run, the timestamp in the header is updated.
6897 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6898 before starting a queue runner without re-exec. This happened only when
6899 deliver_drop_privilege was set or when the Exim user was set to root. The
6900 effect of the bug was that timeouts during subsequent deliveries caused
6901 crashes instead of being properly handled. The handler is now left at its
6902 default (and expected) setting.
6904 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6905 message, again when deliver_drop_privilege is set or Exim is run as root.
6906 The bug described in (1) was not present in this case, but the tidying up
6907 of the other signals was missing. I have made the two cases consistent.
6909 3. The ignore_target_hosts setting on a manualroute router was being ignored
6910 for hosts that were looked up using the /MX notation.
6912 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6915 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6916 operated on the sender address. After changing the $sender_address to <>
6917 for the sender address verify, Exim was re-instated it as the original
6918 (before rewriting) address, but remembering that it had rewritten it, so it
6919 wasn't rewriting it again. This bug also had the effect of breaking the
6920 sender address verification caching when the sender address was rewritten.
6922 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6923 This has been changed so that if the ip literal address matches
6924 ignore_target_hosts, the router declines.
6926 7. Added expansion conditions match_domain, match_address, and match_local_
6927 part (NOT match_host).
6929 8. The placeholder for the Received: header didn't have a length field set.
6931 9. Added code to Exim itself and to exim_lock to test for a specific race
6932 condition that could lead to file corruption when using MBX delivery. The
6933 issue is with the lockfile that is created in /tmp. If this file is removed
6934 after a process has opened it but before that process has acquired a lock,
6935 there is the potential for a second process to recreate the file and also
6936 acquire a lock. This could lead to two Exim processes writing to the file
6937 at the same time. The added code performs the same test as UW imapd; it
6938 checks after acquiring the lock that its file descriptor still refers to
6939 the same named file.
6941 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6942 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6945 11. Added the smtp_active_hostname option. If used, this will typically be made
6946 to depend on the incoming interface address. Because $interface_address is
6947 not set up until the daemon has forked a reception process, error responses
6948 that can happen earlier (such as "too many connections") no longer contain
6951 12. If an expansion in a condition on a "warn" statement fails because a lookup
6952 defers, the "warn" statement is abandoned, and the next ACL statement is
6953 processed. Previously this caused the whole ACL to be aborted.
6955 13. Added the iplsearch lookup type.
6957 14. Added ident_timeout as a log selector.
6959 15. Added tls_certificate_verified as a log selector.
6961 16. Added a global option tls_require_ciphers (compare the smtp transport
6962 option of the same name). This controls incoming TLS connections.
6964 17. I finally figured out how to make tls_require_ciphers do a similar thing
6965 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6966 before starting the TLS session.
6968 18. Tabs are now shown as \t in -bP output.
6970 19. If the log selector return_path_on_delivery was set, Exim crashed when
6971 bouncing a message because it had too many Received: header lines.
6973 20. If two routers both had headers_remove settings, and the first one included
6974 a superfluous trailing colon, the final name in the first list and the
6975 first name in the second list were incorrectly joined into one item (with a
6976 colon in the middle).
6982 1. Added -C and -D options to the exinext utility, mainly to make it easier
6983 to include in the automated testing, but these could be helpful when
6984 multiple configurations are in use.
6986 2. The exinext utility was not formatting the output nicely when there was
6987 an alternate port involved in the retry record key, nor when there was a
6988 message id as well (for retries that were specific to a specific message
6989 and a specific host). It was also confused by IPv6 addresses, because of
6990 the additional colons they contain. I have fixed the IPv4 problem, and
6991 patched it up to do a reasonable job for IPv6.
6993 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6994 delivery, the log line now contains "pipelined" if PIPELINING was used.
6996 4. An SMTP transport process used to panic and die if the bind() call to set
6997 an explicit outgoing interface failed. This has been changed; it is now
6998 treated in the same way as a connect() failure.
7000 5. A reference to $sender_host_name in the part of a conditional expansion
7001 that was being skipped was still causing a DNS lookup. This no longer
7004 6. The def: expansion condition was not recognizing references to header lines
7005 that used bh_ and bheader_.
7007 7. Added the _cache feature to named lists.
7009 8. The code for checking quota_filecount in the appendfile transport was
7010 allowing one more file than it should have been.
7012 9. For compatibility with Sendmail, the command line option
7020 and sets the incoming protocol and host name (for trusted callers). The
7021 host name and its colon can be omitted when only the protocol is to be set.
7022 Note the Exim already has two private options, -pd and -ps, that refer to
7023 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7024 "s", but I don't think that's a major issue.
7026 10. A number of refactoring changes to the code, none of which should affect
7029 (a) The number of logging options was getting close to filling up the
7030 32-bit word that was used as a bit map. I have split them into two classes:
7031 those that are passed in the argument to log_write(), and those that are
7032 only ever tested independently outside of that function. These are now in
7033 separate 32-bit words, so there is plenty of room for expansion again.
7034 There is no change in the user interface or the logging behaviour.
7036 (b) When building, for example, log lines, the code previously used a
7037 macro that called string_cat() twice, in order to add two strings. This is
7038 not really sufficiently general. Furthermore, there was one instance where
7039 it was actually wrong because one of the argument was used twice, and in
7040 one call a function was used. (As it happened, calling the function twice
7041 did not affect the overall behaviour.) The macro has been replaced by a
7042 function that can join an arbitrary number of extra strings onto a growing
7045 (c) The code for expansion conditions now uses a table and a binary chop
7046 instead of a serial search (which was left over from when there were very
7047 few conditions). Also, it now recognizes conditions like "pam" even when
7048 the relevant support is not compiled in: a suitably worded error message is
7049 given if an attempt is made to use such a condition.
7051 11. Added ${time_interval:xxxxx}.
7053 12. A bug was causing one of the ddress fields not to be passed back correctly
7054 from remote delivery subprocesses. The field in question was not being
7055 subsequently used, so this caused to problems in practice.
7057 13. Added new log selectors queue_time and deliver_time.
7059 14. Might have fixed a bug in maildirsizefile handling that threw up
7060 "unexpected character" debug warnings, and recalculated the data
7061 unnecessarily. In any case, I expanded the warning message to give more
7064 15. Added the message "Restricted characters in address" to the statements in
7065 the default ACL that block characters like @ and % in local parts.
7067 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7068 Three changes have been made:
7070 (a) There was a serious bug; a negative response to MAIL caused the whole
7071 recipient domain to be cached as invalid, thereby blocking all messages
7072 to all local parts at the same domain, from all senders. This bug has
7073 been fixed. The domain is no longer cached after a negative response to
7074 MAIL if the sender used is not empty.
7076 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7079 (c) A new callout option, "use_sender" has been added for people who want
7080 the modified behaviour.
7086 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7089 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7090 indeed breaks things for older releases.
7092 3. Added additional logging to the case where there is a problem reading data
7093 from a filter that is running in a subprocess using a pipe, in order to
7094 try to track down a specific problem.
7096 4. Testing facility fudge: when running in the test harness and attempting
7097 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7098 getting "No route to host". Convert this to a timeout.
7100 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7103 6. Some OS don't have socklen_t but use size_t instead. This affects the
7104 fifth argument of getsockopt() amongst other things. This is now
7105 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7106 can be set for individual OS. I have set it for SunOS5, OSF1, and
7107 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7108 some earlier ones do not.
7110 7. Change 4.30/15 was not doing the test caselessly.
7112 8. The standard form for an IPv6 address literal was being rejected by address
7113 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7114 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7115 this, as well as the form without the "IPv6" on the front (but only when
7116 address literals are enabled, of course).
7118 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7120 10. Exim crashed if a message with an empty sender address specified by -f
7121 encountered a router with an errors_to setting. This could be provoked only
7122 by a command such as
7126 where an empty string was supplied; "<>" did not hit this bug.
7128 11. Installed PCRE release 4.5.
7130 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7131 remained set. It is now erased.
7133 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7134 times from message ids (which are base 36 rather than the normal 62).
7136 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7137 were being counted as actual protocol errors, and logged if the log
7138 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7139 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7140 and DATA following a set of rejected RCPTs do not count as protocol errors.
7141 In other words, Exim assumes they were pipelined, though this may not
7142 actually be the case. Of course, in all cases the client gets an
7143 appropriate error code.
7145 15. If a lookup fails in an ACL condition, a message about the failure may
7146 be available; it is used if testing the ACL cannot continue, because most
7147 such messages specify what the cause of the deferral is. However, some
7148 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7149 that caused an old message to be retained and used if a later statement
7150 caused a defer, replacing the real cause of the deferral.
7152 16. If an IP address had so many PTR records that the DNS lookup buffer
7153 was not large enough to hold them, Exim could crash while trying to process
7154 the truncated data. It now detects and logs this case.
7156 17. Further to 4.21/58, another change has been made: if (and only if) the
7157 first line of a message (the first header line) ends with CRLF, a bare LF
7158 in a subsequent header line has a space inserted after it, so as not to
7159 terminate the header.
7161 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7162 unnecessarily, used atoi() instead of strtol(), and didn't check the
7163 termination when getting file sizes from file names by regex.
7165 19. Completely re-implemented the support for maildirsize files, in the light
7166 of a number of problems with the previous contributed implementation
7167 (4.30/29). In particular:
7169 . If the quota is zero, the maildirsize file is maintained, but no quota is
7172 . If the maildir directory does not exist, it is created before any attempt
7173 to write a maildirsize file.
7175 . The quota value in the file is just a cache; if the quota is changed in
7176 the transport, the new value overrides.
7178 . A regular expression is available for excluding directories from the
7181 20. The autoreply transport checks the characters in options that define the
7182 message's headers; it allows continued headers, but it was checking with
7183 isspace() after an embedded newline instead of explicitly looking for a
7186 21. If all the "regular" hosts to which an address was routed had passed their
7187 expiry times, and had not reached their retry times, the address was
7188 bounced, even if fallback hosts were defined. Now Exim should go on to try
7191 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7192 equivalent code in the SMTP transport. Some hosts send humungous responses
7193 to HELO/EHLO, more than 1024 it seems.
7195 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7196 gives compiler warnings in some environments. I've now done it "properly",
7199 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7200 (because of problems with the built-in one) was declared to return uschar *
7201 instead of char *, causing compiler failure.
7203 25. Fixed a file descriptor leak when processing alias/forward files.
7205 26. Fixed a minor format string issue in dbfn.c.
7207 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7209 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7210 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7213 29. When a sender address is verified, it is cached, to save repeating the test
7214 when there is more than one recipient in a message. However, when the
7215 verification involves a callout, it is possible for different callout
7216 options to be set for different recipients. It is too complicated to keep
7217 track of this in the cache, so now Exim always runs a verification when a
7218 callout is required, relying on the callout cache for the optimization.
7219 The overhead is duplication of the address routing, but this should not be
7222 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7223 to be verified with callout=postmaster, and the main callout worked but the
7224 postmaster check failed, the verification correctly failed. However, if a
7225 subsequent RCPT command asked for sender verification *without* the
7226 postmaster check, incorrect caching caused this verification also to fail,
7229 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7230 it was not caching the DNS options (qualify_single, search_parents) that
7231 were used when the lookup failed. A subsequent lookup with different
7232 options therefore always gave the same answer, though there were cases
7233 where it should not have. (Example: a "domains = !$mx_any" option on a
7234 dnslookup router: the "domains" option is always processed without any
7235 widening, but the router might have qualify_single set.) Now Exim uses the
7236 cached value only when the same options are set.
7238 32. Added John Jetmore's "exipick" utility to the distribution.
7240 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7241 than a timeout (e.g. a certificate is required, and is not provided), an
7242 Exim server now closes the connection immediately. Previously it waited for
7243 the client to close - but if the client is SSL, it seems that they each
7244 wait for each other, leading to a delay before one of them times out.
7246 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7247 maintained 0.8.x compatibility because I don't think many are using it, and
7248 it is clearly obsolete.
7250 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7253 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7254 client certificate was expired. A simple patch fixes this, though I don't
7255 understand the full logic of why the verify callback is called multiple
7258 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7259 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7260 which causes problems with some clients (such as the Certicom SSL Plus
7261 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7262 disables the coutermeasure allowing Eudora to connect."
7264 38. Exim was not checking that a write() to a log file succeeded. This could
7265 lead to Bad Things if a log got too big, in particular if it hit a file
7266 size limit. Exim now panics and dies if it cannot write to a log file, just
7267 as it does if it cannot open a log file.
7269 39. Modified OS/Makefile-Linux so that it now contains
7271 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7273 The two -D definitions ensure that Exim is compiled with large file
7274 support, which makes it possible to handle log files that are bigger than
7277 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7278 instance) a domain was checked against a named list that involved a lookup,
7279 causing $domain_data to be set, then another domain was checked against the
7280 same list, then the first domain was re-checked, the value of $domain_data
7281 after the final check could be wrong. In particular, if the second check
7282 failed, it could be set empty. This bug probably also applied to
7285 41. The strip_trailing_dot option was not being applied to the address given
7286 with the -f command-line option.
7288 42. The code for reading a message's header from the spool was incrementing
7289 $received_count, but never initializing it. This meant that the value was
7290 incorrect (doubled) while delivering a message in the same process in which
7291 it was received. In the most common configuration of Exim, this never
7292 happens - a fresh exec is done - but it can happen when
7293 deliver_drop_privilege is set.
7295 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7296 it now includes up to 150 characters of the unexpected data in the log
7299 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7300 and building data strings. The size of both of these buffers was 10 000
7301 bytes - far larger than anybody would *ever* want, thought I. Needless to
7302 say, somebody hit the limit. I have increased the maximum line length to
7303 20 000 and the maximum data length of concatenated lines to 100 000. I have
7304 also fixed two bugs, because there was no checking on these buffers. Tsk,
7305 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7306 buffer is too small.
7308 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7309 lsearch lookups. Now it does.
7311 46. When parsing a route_list item in a manualroute router, a fixed-length
7312 buffer was used for the list of hosts. I made this 1024 bytes long,
7313 thinking that nobody would ever have a list of hosts that long. Wrong.
7314 Somebody had a whole pile of complicated expansion conditions, and the
7315 string was silently truncated, leading to an expansion error. It turns out
7316 that it is easier to change to an unlimited length (owing to other changes
7317 that have happened since this code was originally written) than to build
7318 structure for giving a limitation error. The length of the item that
7319 expands into the list of hosts is now unlimited.
7321 47. The lsearch lookup could not handle data where the length of text line was
7322 more than 4095 characters. Such lines were truncated, leading to shortened
7323 data being returned. It should now handle lines of any length.
7325 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7326 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7329 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7330 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7331 no longer happens in comment lines. A list of replacements is now placed
7332 at the head of all of the source files, except those whose only change is
7333 to replace PERL_COMMAND in the very first #! line.
7335 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7336 messages on the queue, with a bottom-up merge sort, using code contributed
7337 by Michael Haardt. This should make operations like -bp somewhat faster on
7338 large queues. It won't affect queue runners, except when queue_run_in_order
7341 51. Installed eximstats 1.31 in the distribution.
7343 52. Added support for SRV lookups to the dnslookup router.
7345 53. If an ACL referred to $message_body or $message_body_end, the value was not
7346 reset for any messages that followed in the same SMTP session.
7348 54. The store-handling optimization for building very long strings was not
7349 differentiating between the different store pools. I don't think this
7350 actually made any difference in practice, but I've tidied it.
7352 55. While running the routers to verify a sender address, $sender_address
7353 was still set to the sender address. This is wrong, because when routing to
7354 send a bounce to the sender, it would be empty. Therefore, I have changed
7355 it so that, while verifying a sender address, $sender_address is set to <>.
7356 (There is no change to what happens when verifying a recipient address.)
7358 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7359 target A or AAAA records (if not already returned) without resetting the
7360 qualify_single or search_parents options of the DNS resolver. These are
7361 inappropriate in this case because the targets of MX and SRV records must
7362 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7363 target that, when qualified, matched something in the local domain. These
7364 two options are now turned off when doing these lookups.
7366 57. It seems that at least some releases of Reiserfs (which does not have the
7367 concept of a fixed number of inodes) returns zero and not -1 for the
7368 number of available inodes. This interacted badly with check_spool_inodes,
7369 which assumed that -1 was the "no such thing" setting. What I have done is
7370 to check that the total number of inodes is greater than zero before doing
7371 the test of how many are available.
7373 58. When a "warn" ACL statement has a log_message modifier, the message is
7374 remembered, and not repeated. This is to avoid a lot of repetition when a
7375 message has many recipients that cause the same warning to be written.
7376 However, Exim was preserving the list of already written lines for an
7377 entire SMTP session, which doesn't seem right. The memory is now reset if a
7378 new message is started.
7380 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7381 debugging output unless log_rewrite was also set.
7383 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7384 of (int)(handle) when we know that handle contains (void *)(-1).
7386 61. The Exim daemon panic-logs an error return when it closes the incoming
7387 connection. However "connection reset by peer" seems to be common, and
7388 isn't really an error worthy of noting specially, so that particular error
7391 62. When Exim is trying to find all the local interfaces, it used to panic and
7392 die if the ioctl to get the interface flags failed. However, it seems that
7393 on at least one OS (Solaris 9) it is possible to have an interface that is
7394 included in the list of interfaces, but for which you get a failure error
7395 for this call. This happens when the interface is not "plumbed" into a
7396 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7397 failure of the "get flags" call assumes that the interface is down.
7399 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7400 makes life easier for people who are doing arithmetic on fields extracted
7401 from dates, where you often get leading zeros that should not be
7402 interpreted as octal.
7404 64. Added qualify_domain to the redirect router, to override the global
7407 65. If a pathologically long header line contained very many addresses (the
7408 report of this problem mentioned 10 000) and each of them was rewritten,
7409 Exim could use up a very large amount of memory. (It kept on making new
7410 copies of the header line as it rewrote, and never released the old ones.)
7411 At the expense of a bit more processing, the header rewriting function has
7412 been changed so that it no longer eats memory in this way.
7414 66. The generation of the Received: header has been moved from the time that a
7415 message starts to be received, to the time that it finishes. The timestamp
7416 in the Received: header should now be very close to that of the <= log
7417 line. There are two side-effects of this change:
7419 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7420 logged header lines no longer include the local Received: line, because
7421 it has not yet been created. The same applies to a copy of the message
7422 that is returned to a non-SMTP sender when a message is rejected.
7424 (b) When a filter file is tested using -bf, no additional Received: header
7425 is added to the test message. After some thought, I decided that this
7428 This change does not affect the value of $received_for. It is still set
7429 after address rewriting, but before local_scan() is called.
7431 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7433 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7434 gave an unhelpful panic error message, and a defer error. I have managed to
7435 change this behaviour so that it now rejects any supplied certificate,
7436 which seems right, as the list of acceptable certificates is empty.
7438 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7439 gave an unhelpful defer error. I have not managed to make this reject any
7440 supplied certificates, but the error message it gives is "no certificate
7441 supplied", which is not helpful.
7443 70. exigrep's output now also includes lines that are not associated with any
7444 message, but which match the given pattern. Implemented by a patch from
7445 Martin Sluka, which also tidied up the Perl a bit.
7447 71. Recipient callout verification, like sender verification, was using <> in
7448 the MAIL FROM command. This isn't really the right thing, since the actual
7449 sender may affect whether the remote host accepts the recipient or not. I
7450 have changed it to use the actual sender in the callout; this means that
7451 the cache record is now keyed on a recipient/sender pair, not just the
7452 recipient address. There doesn't seem to be a real danger of callout loops,
7453 since a callout by the remote host to check the sender would use <>.
7454 [SEE ABOVE: changed after hitting problems.]
7456 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7457 temporary errors. However, in the case of such a code being given after
7458 the end of a data transmission (i.e. after ".") Exim was failing to write
7459 a retry record for the message. (Yes, there was some broken host that was
7460 actually sending 8xx at this point.)
7462 73. An unknown lookup type in a host list could cause Exim to panic-die when
7463 the list was checked. (An example that provoked this was putting <; in the
7464 middle of a list instead of at the start.) If this happened during a DATA
7465 ACL check, a -D file could be left lying around. This kind of configuration
7466 error no longer causes Exim to die; instead it causes a defer error. The
7467 incident is still logged to the main and panic logs.
7469 74. Buglet left over from Exim 3 conversion. The message "too many messages
7470 in one connection" was written to the rejectlog but not the mainlog, except
7471 when address rewriting (yes!) was being logged.
7473 75. Added write_rejectlog option.
7475 76. When a system filter was run not as root (that is, when system_filter_user
7476 was set), the values of the $n variables were not being returned to the
7477 main process; thus, they were not subsequently available in the $sn
7480 77. Added +return_path_on_delivery log selector.
7482 78. A connection timeout was being treated differently from recipients deferred
7483 when testing hosts_max_try with a message that was older than the host's
7484 retry timeout. (The host should not be counted, thus allowing all hosts to
7485 be tried at least once before bouncing.) This may have been the cause of an
7486 occasionally reported bug whereby a message would remain on the queue
7487 longer than the retry timeout, but would be bounced if a delivery was
7488 forced. I say "may" because I never totally pinned down the problem;
7489 setting up timeout/retry tests is difficult. See also the next item.
7491 79. The ultimate address timeout was not being applied to errors that involved
7492 a combination of host plus message (for example, a timeout on a MAIL
7493 command). When an address resolved to a number of possible hosts, and they
7494 were not all tried for each delivery (e.g. because of hosts_max_try), a
7495 message could remain on the queue longer than the retry timeout.
7497 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7500 81. Fixed an obscure SMTP outgoing bug which required at least the following
7501 conditions: (a) there was another message waiting for the same server;
7502 (b) the server returned 5xx to all RCPT commands in the first message so
7503 that the message was not completed; (c) the server dropped the connection
7504 or gave a negative response to the RSET that Exim sends to abort the
7505 transaction. The observed case was a dropped connection after DATA that had
7506 been sent in pipelining mode. That is, the server had advertised PIPELINING
7507 but was not implementing it correctly. The effect of the bug was incorrect
7508 behaviour, such as trying another host, and this could lead to a crash.
7514 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7515 and daemon.c were passed as pointers to ints; they should have been
7516 pointers to socklen_t variables (which are typically unsigned ints).
7518 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7521 3. Fixed a really odd bug that affected only the testing scheme; patching a
7522 certain fixed string in the binary changed the value of another string that
7523 happened to be identical to the end of the original first string.
7525 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7526 name", it returns that address as the IP address. On some operating
7527 systems (e.g. Solaris), it also passes back the IP address string as the
7528 "host name". However, on others (e.g. Linux), it passes back an empty
7529 string. Exim wasn't checking for this, and was changing the host name to an
7530 empty string, assuming it had been canonicalized.
7532 5. Although rare, it is permitted to have more than one PTR record for a given
7533 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7534 all the names associated with an address, because they do in Solaris.
7535 However, it seems that they do not in Linux for data that comes from the
7536 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7537 I found this out when I moved to a new Linux workstation and tried to run
7538 the Exim test suite.
7540 To get round this problem I have changed the code so that it now does its
7541 own call to the DNS to look up PTR records when searching for a host name.
7542 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7543 addresses that are only in /etc/hosts are still found.
7545 This behaviour is, however, controlled by an option called host_lookup_
7546 order, which defaults to "bydns:byaddr". If people want to use the other
7547 order, or indeed, just use one or the other means of lookup, they can
7548 specify it in this variable.
7550 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7551 some operating systems, this comes back from gethostbyaddr() as an empty
7552 string, and this is what Exim used to test for. However, it seems that in
7553 other systems, "." is yielded. Exim now tests for this case too.
7555 7. The values of check_spool_space and check_log_space are now held internally
7556 as a number of kilobytes instead of an absolute number of bytes. If a
7557 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7558 kilobyte. This means that much larger values can be stored.
7560 8. Exim monitor: an attempt to get the action menu when not actually pointing
7561 at a message produces an empty menu entitled "No message selected". This
7562 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7563 no entries in it ("Shell widget menu has zero width and/or height"). So I
7564 have added a single, blank menu entry in this case.
7566 9. Added ${quote_local_part.
7568 10. MIME decoding is now applied to the contents of Subject: header lines when
7571 11. Now that a reference to $sender_host_address automatically causes a reverse
7572 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7573 host lookup before query-style lookups in lists that might use this
7574 variable. This has therefore been abolished, and the "net-" prefix is no
7575 longer necessary for query-style lookups.
7577 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7578 have been a typo for LFLAGS, so it has been changed.
7580 13. The install script calls Exim with "-C /dev/null" in order to find the
7581 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7582 to be output. However, since Exim outputs its version number before the
7583 error, it didn't break the script. It just looked ugly. I fixed this by
7584 always allowing "-C /dev/null" if the caller is root.
7586 14. Ignore overlarge ACL variable number when reading spool file - insurance
7587 against a later release with more variables having written the file.
7589 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7590 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7591 this, as well as the form without the "IPv6" on the front.
7593 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7594 OS/Makefile-Darwin file.
7596 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7597 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7599 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7600 variable. However, some OS use ints for inodes. I've added cast to long int
7601 to get rid of the compiler warning.
7603 19. I had forgotten to lock out "/../" in configuration file names when
7604 ALT_CONFIG_PREFIX was set.
7606 20. Routers used for verification do not need to specify transports. However,
7607 if such a router generated a host list, and callout was configured, Exim
7608 crashed, because it could not find a port number from the (non-existent)
7609 transport. It now assumes port 25 in this circumstance.
7611 21. Added the -t option to exigrep.
7613 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7614 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7615 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7616 them is set, LOOKUP_LSEARCH is forced.
7618 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7620 24. Added sender and host information to the "rejected by local_scan()" log
7621 line; previously there was no indication of these.
7623 25. Added .include_if_exists.
7625 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7626 Linux. It turns out that not all file systems support this. Apparently some
7627 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7628 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7629 not supported on the file descriptor, is now ignored when Exim is trying to
7630 sync a directory. This applies only to Linux.
7632 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7634 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7635 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7638 29. Added support for maildirsize files from supplied patch (modified a bit).
7640 30. The use of :fail: followed by an empty string could lead Exim to respond to
7641 sender verification failures with (e.g.):
7643 550 Verification failed for <xxx>
7644 550 Sender verify failed
7646 where the first response line was missing the '-' that indicates it is not
7647 the final line of the response.
7649 31. The loop for finding the name of the user that called Exim had a hardwired
7650 limit of 10; it now uses the value of finduser_retries, which is used for
7651 all other user lookups.
7653 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7656 33. Exim was neglecting to zero errno before one call of strtol() when
7657 expanding a string and expecting an integer value. On some systems this
7658 resulted in spurious "integer overflow" errors. Also, it was casting the
7659 result into an int without checking.
7661 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7662 did not work. The code looks as if it has *never* worked, though it appears
7663 to have been documented since at least release 1.62. I have made it work.
7665 35. The "timeout_DNS" error in retry rules, also documented since at least
7666 1.62, also never worked. As it isn't clear exactly what this means, and
7667 clearly it isn't a major issue, I have abolished the feature by treating it
7668 as "timeout", and writing a warning to the main and panic logs.
7670 36. The display of retry rules for -brt wasn't always showing the error code
7673 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7674 timeout_connect_A, timeout_connect_MX.
7676 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7677 to the empty sender.
7679 39. The daemon was not analysing the content of -oX till after it had closed
7680 stderr and disconnected from the controlling terminal. This meant that any
7681 syntax errors were only noted on the panic log, and the return code from
7682 the command was 0. By re-arranging the code a little, I've made the
7683 decoding happen first, so such errors now appear on stderr, and the return
7684 code is 1. However, the actual setting up of the sockets still happens in
7685 the disconnected process, so errors there are still only recorded on the
7688 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7689 connections (as happens on some IP stacks) was logged at start up time as
7690 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7691 it from "IPv6 and IPv4", which means that two separate sockets are being
7694 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7695 says whether AF_INET or AF_INET6 was passed as an argument.
7697 42. Exiwhat output was messed up when time zones were included in log
7700 43. Exiwhat now gives more information about the daemon's listening ports,
7701 and whether -tls-on-connect was used.
7703 44. The "port" option of the smtp transport is now expanded.
7705 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7706 silently ignored. Now an error message is written to the main and panic
7709 46. There's a new ACL modifier called "logwrite" which writes to a log file
7710 as soon as it is encountered.
7712 47. Added $local_user_uid and $local_user_gid at routing time.
7714 48. Exim crashed when trying to verify a sender address that was being
7717 49. Exim was recognizing only a space character after ".include". It now also
7718 recognizes a tab character.
7720 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7721 extracting the relevant information from the specification. The man page no
7722 longer contains scrambled data for the -d option, and I've added a section
7723 at the front about calling Exim under different names.
7725 51. Added "extra_headers" argument to the "mail" command in filter files.
7727 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7730 53. Installed eximstats 1.29.
7732 54. Added transport_filter_timeout as a generic transport option.
7734 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7735 Cc: header lines. This was required by RFC 822, but it not required by RFC
7738 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7739 incoming messages that did not have them. Now it does so only if the
7740 message originates locally, that is, if there is no associated remote host
7741 address. When Resent- header lines are present, this applies to the Resent-
7742 lines rather than the non-Resent- lines.
7744 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7745 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7747 58. Messages for configuration errors now include the name of the main
7748 configuration file - useful now that there may be more than one file in a
7749 list (.included file names were always shown).
7751 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7752 for those rare installations that do not start the daemon as root or run it
7753 setuid root. I've cut out the call to initgroups() if the daemon is not
7756 60. The Exim user and group can now be bound into the binary as text strings
7757 that are looked up at the start of Exim's processing.
7759 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7761 62. Added $mailstore_basename variable.
7763 63. Installed patch to sieve.c from Michael Haardt.
7765 64. When Exim failed to open the panic log after failing to open the main log,
7766 the original message it was trying to log was written to stderr and debug
7767 output, but if they were not available (the usual case in production), it
7768 was lost. Now it is written to syslog before the two lines that record the
7769 failures to open the logs.
7771 65. Users' Exim filters run in subprocesses under the user's uid. It is
7772 possible for a "deliver" command or an alias in a "personal" command to
7773 provoke an address rewrite. If logging of address rewriting is configured,
7774 this fails because the process is not running as root or exim. There may be
7775 a better way of dealing with this, but for the moment (because 4.30 needs
7776 to be released), I have disabled address rewrite logging when running a
7777 filter in a non-root, non-exim process.
7783 1. The buildconfig auxiliary program wasn't quoting the value set for
7784 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7785 not defined. This bug was masked in 4.22 by the effect that was fixed in
7788 2. Some messages that were rejected after a message id was allocated were
7789 shown as "incomplete" by exigrep. It no longer does this for messages that
7790 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7792 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7793 have allow_domain_literals set, the ID did not get logged in the <= line.
7794 Domain literals are now always recognized in Message-ID: header lines.
7796 4. The first argument for a ${extract expansion item is the key name or field
7797 number. Leading and trailing spaces in this item were not being ignored,
7798 causing some misleading effects.
7800 5. When deliver_drop_privilege was set, single queue runner processes started
7801 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7802 same command in the process it spins off) were not dropping privilege.
7804 6. When the daemon running as "exim" started a queue runner, it always
7805 re-executed Exim in the spun-off process. This is a waste of effort when
7806 deliver_drop_privilege is set. The new process now just calls the
7807 queue-runner function directly.
7813 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7816 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7817 ignored. Though the use of -oP was forcing the writing of a pid file, it
7818 was always written to the default place.
7820 3. If the message "no IP address found for host xxxx" is generated during
7821 incoming verification, it is now followed by identification of the incoming
7822 connection (so you can more easily find what provoked it).
7824 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7826 5. Added some features to "harden" Exim a bit more against certain attacks:
7828 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7829 be put in Local/Makefile. This is like the never_users runtime option,
7830 but it cannot be overridden. The default setting is "root".
7832 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7833 prefix string with which any file named in a -C command line option
7836 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7837 is retained for -C and -D only if the caller of Exim is root. Without
7838 it, the exim user may also use -C and -D and retain privilege.
7840 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7841 command line option is disabled.
7843 6. Macro names set by the -D option must start with an upper case letter, just
7844 like macro names defined in the configuration file.
7846 7. Added "dereference=" facility to LDAP.
7848 8. Two instances of the typo "uknown" in the source files are fixed.
7850 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7851 the Configure-Makefile script screwed up while processing it.
7853 10. Incorporated PCRE 4.4.
7855 11. The SMTP synchronization check was not operating right at the start of an
7856 SMTP session. For example, it could not catch a HELO sent before the client
7857 waited for the greeting. There is now a check for outstanding input at the
7858 point when the greeting is written. Because of the duplex, asynchronous
7859 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7860 way, but not yet received, when the check is performed.
7862 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7863 on TCP/IP sockets, because this apparently causes some broken clients to
7866 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7867 unchanged) from the Cygwin maintainer.
7869 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7870 was supported instead of testing for mbx. Effectively a typo.
7872 15. The spa authenticator server code was not checking that the input it
7873 received was valid base64.
7875 16. The debug output line for the "set" modifier in ACLs was not showing the
7876 name of the variable that was being set.
7878 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7880 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7881 lookup on its own. Something else was needed to trigger the lookup. For
7882 example, a match in host_lookup or the need for a host name in a host list.
7883 Now, if $sender_host_name is referenced and the host name has not yet been
7884 looked up, a lookup is performed. If the lookup fails, the variable remains
7885 empty, and $host_lookup_failed is set to "1".
7887 19. Added "eqi" as a case-independent comparison operator.
7889 20. The saslauthd authentication condition could segfault if neither service
7890 nor realm was specified.
7892 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7893 error message that was logged was misleading, and incoming SMTP
7894 connections were dropped. The message is now more accurate, and temporary
7895 errors are given to SMTP connections.
7897 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7898 (except RSET) with a 421 error, until QUIT is received. However, it was
7899 failing to send a response to QUIT.
7901 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7902 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7903 ACL is now run after the helo_try_verify_hosts code.
7905 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7906 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7907 case-independent, so other case variants are also recognized). Apparently
7908 some systems use these upper case variants.
7910 25. If more than two messages were waiting for the same host, and a transport
7911 filter was specified for the transport, Exim sent two messages over the
7912 same TCP/IP connection, and then failed with "socket operation on non-
7913 socket" when it tried to send the third.
7915 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7917 27. The extern definition of crypt16() in expand.c was not being excluded when
7918 the OS had its own crypt16() function.
7920 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7921 as a preferred synonym for return_size_limit, both as an option and as an
7924 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7926 30. Changed the default configuration ACL to relax the local part checking rule
7927 for addresses that are not in any local domains. For these addresses,
7928 slashes and pipe symbols are allowed within local parts, but the sequence
7929 /../ is explicitly forbidden.
7931 31. SPA server authentication was not clearing the challenge buffer before
7934 32. log_message in a "warn" ACL statement was writing to the reject log as
7935 well as to the main log, which contradicts the documentation and doesn't
7936 seem right (because no rejection is happening). So I have stopped it.
7938 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7939 However, I am unable to do any testing of this.
7941 34. Fixed an infelicity in the appendfile transport. When checking directories
7942 for a mailbox, to see if any needed to be created, it was accidentally
7943 using path names with one or more superfluous leading slashes; tracing
7944 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7946 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7947 contents are added to the log line that is written for every discarded
7948 recipient. (Previously a log_message setting was ignored.)
7950 36. The ${quote: operator now quotes the string if it is empty.
7952 37. The install script runs exim in order to find its version number. If for
7953 some reason other than non-existence or emptiness, which it checks, it
7954 could not run './exim', it was installing it with an empty version number,
7955 i.e. as "exim-". This error state is now caught, and the installation is
7958 38. An argument was missing from the function that creates an error message
7959 when Exim fails to connect to the socket for saslauthd authentication.
7960 This could cause Exim to crash, or give a corrupted message.
7962 39. Added isip, isip4, and isip6 to ${if conditions.
7964 40. The ACL variables $acl_xx are now saved with the message, and can be
7965 accessed later in routers, transports, and filters.
7967 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7968 strings in the file are not string-expanded.
7970 42. If a MAIL command specified a SIZE value that was too large to fit into an
7971 int variable, the check against message_size_limit failed. Such values are
7972 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7973 day this will have to be increased, but I don't think I want to be around
7974 when emails are that large.
7981 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7982 iconv() is not standard in FreeBSD.
7984 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7985 IPv6 enabled. The observed symptom was a segmentation fault on return from
7986 the function os_common_find_running_interfaces() in src/os.c.
7988 3. In the check_special_case() function in daemon.c I had used "errno" as an
7989 argument name, which causes warnings on some systems. This was basically a
7990 typo, since it was named "eno" in the comments!
7992 4. The code that waits for the clock to tick (at a resolution of some fraction
7993 of a second) so as to ensure message-id uniqueness was always waiting for
7994 at least one whole tick, when it could have waited for less. [This is
7995 almost certainly not relevant at current processor speeds, where it is
7996 unlikely to ever wait at all. But we try to future-proof.]
7998 5. The function that sleeps for a time interval that includes fractions of a
7999 second contained a race. It did not block SIGALRM between setting the
8000 timer, and suspending (a couple of lines later). If the interval was short
8001 and the sigsuspend() was delayed until after it had expired, the suspension
8002 never ended. On busy systems this could lead to processes getting stuck for
8005 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8006 process, before it forks any delivery processes. The open lookup caching
8007 mechanism meant that the open file or database connection was passed into
8008 the delivery process. The problem was that delivery processes always tidy
8009 up cached lookup data. This could cause a problem for the next delivery
8010 process started by the queue runner, because the external queue runner
8011 process does not know about the closure. So the next delivery process
8012 still has data in the lookup cache. In the case of a file lookup, there was
8013 no problem because closing a file descriptor in a subprocess doesn't affect
8014 the parent. However, if the lookup was caching a connection to a database,
8015 the connection was closed, and the second delivery process was likely to
8016 see errors such as "PGSQL: query failed: server closed the connection
8017 unexpectedly". The problem has been fixed by closing all cached lookups
8018 in a queue runner before running a delivery process.
8020 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8021 seem to have the "const" qualifier which it has on other OS. I've
8024 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8025 *of the same type* (client or server) with the same public name that an
8026 error should be diagnosed.
8028 9. When Exim looked up a host name for an IP address, but failed to find the
8029 original IP address when looking up the host name (a safety check), it
8030 output the message "<ip address> does not match any IP for NULL", which was
8031 confusing, to say the least. The bug was that the host name should have
8032 appeared instead of "NULL".
8034 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8035 user that is built into the binary, and the -C or -D options is used, root
8036 privilege is dropped before the configuration file is read. In addition,
8037 logging is switched to stderr instead of the normal log files. If the
8038 configuration then re-defines the Exim user, the unprivileged environment
8039 is probably not what is expected, so Exim logs a panic warning message (but
8042 However, if deliver_drop_privilege is set, the unprivileged state may well
8043 be exactly what is intended, so the warning has been cut out in that case,
8044 and Exim is allowed to try to write to its normal log files.
8050 1. smtp_return_error_details was not giving details for temporary sender
8051 or receiver verification errors.
8053 2. Diagnose a configuration error if two authenticators have the same public
8056 3. Exim used not to create the message log file for a message until the first
8057 delivery attempt. This could be confusing when incoming messages were held
8058 for policy or load reasons. The message log file is now created at the time
8059 the message is received, and an initial "Received" line is written to it.
8061 4. The automatically generated man page for command line options had a minor
8062 bug that caused no ill effects; however, a more serious problem was that
8063 the procedure for building the man page automatically didn't always
8064 operate. Consequently, release 4.20 contains an out-of-date version. This
8065 shouldn't happen again.
8067 5. When building Exim with embedded Perl support, the script that builds the
8068 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8069 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8071 6. The freeze_tell option was not being used for messages that were frozen on
8072 arrival, either by an ACL or by local_scan().
8074 7. Added the smtp_incomplete_transaction log selector.
8076 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8077 was accepting AUTH without a new EHLO.
8079 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8080 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8083 10. Logging of TCP/IP connections (when configured) now happens in the main
8084 daemon process instead of the child process, so that the TCP/IP connection
8085 count is more accurate (but it can never be perfect).
8087 11. The use of "drop" in a nested ACL was not being handled correctly in the
8088 outer ACL. Now, if condition failure induced by the nested "drop" causes
8089 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8090 or "require"), the connection is dropped.
8092 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8093 that yield "discard" can now be used with an "accept" or a "discard" verb,
8094 but an error is generated for any others (because I can't see a useful way
8095 to define what should happen).
8097 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8098 are now processed in the same way as lines in the Exim configuration file.
8099 In particular, continuation lines are supported.
8101 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8103 15. Added -ti meaning -t -i.
8105 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8106 domains, and warn by logging if others are found.
8108 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8109 returned from GIFCONF when Exim is trying to find the list of interfaces on
8110 a host. The code in os.c has been modified to copy each ifreq to an aligned
8111 structure in all cases.
8113 Also, in some cases, the returned ifreq's were being copied to a 'struct
8114 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8115 means the last couple of bytes of an IPv6 address could be chopped if the
8116 ifreq contained only a normal sockaddr (14 bytes storage).
8118 18. Named domain lists were not supported in the hosts_treat_as_local option.
8119 An entry such as +xxxx was not recognized, and was treated as a literal
8122 19. Ensure that header lines added by a DATA ACL are included in the reject log
8123 if the ACL subsequently rejects the message.
8125 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8126 MD5 (which is deprecated).
8128 21. When testing a filter file using -bf, Exim was writing a message when it
8129 took the sender from a "From " line in the message, but it was not doing so
8130 when it took $return_path from a Return-Path: header line. It now does.
8132 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8133 with a valid header line field name (a series of printing characters
8134 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8136 23. Changed "disc" in the source to "disk" to conform to the documentation and
8137 the book and for uniformity.
8139 24. Ignore Sendmail's -Ooption=value command line item.
8141 25. When execve() failed while trying to run a command in a pipe transport,
8142 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8143 could be confused with a return value of 69 from the command itself. This
8144 has been changed to 127, the value the shell returns if it is asked to run
8145 a non-existent command. The wording for the related log line suggests a
8146 non-existent command as the problem.
8148 26. If received_header_text expands to an empty string, do not add a Received:
8149 header line to the message. (Well, it adds a token one on the spool, but
8150 marks it "old" so that it doesn't get used or transmitted.)
8152 27. Installed eximstats 1.28 (addition of -nt option).
8154 28. There was no check for failure on the call to getsockname() in the daemon
8155 code. This can fail if there is a shortage of resources on the system, with
8156 ENOMEM, for example. A temporary error is now given on failure.
8158 29. Contrary to the C standard, it seems that in some environments, the
8159 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8160 program. Exim now does this explicitly; it affects the formatting of
8161 timestamps using strftime().
8163 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8164 complaints. I've now initialized all the variables, to avoid this.
8166 32. Header lines added by a system filter were not being "seen" during
8167 transport-time rewrites.
8169 33. The info_callback() function passed to OpenSSL is set up with type void
8170 (*)(SSL *, int, int), as described somewhere. However, when calling the
8171 function (actually a macro) that sets it up, the type void(*)() is
8172 expected. I've put in a cast to prevent warnings from picky compilers.
8174 34. If a DNS black list lookup found a CNAME record, but there were no A
8175 records associated with the domain it pointed at, Exim crashed.
8177 35. If a DNS black list lookup returned more than one A record, Exim ignored
8178 all but the first. It now scans all returned addresses if a particular IP
8179 value is being sought. In this situation, the contents of the
8180 $dnslist_value variable are a list of all the addresses, separated by a
8183 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8184 to accept a host name and all its aliases if the forward lookup for any of
8185 them yielded the IP address of the incoming connection. Now it accepts only
8186 those names whose forward lookup yields the correct IP address. Any other
8187 names are discarded. This closes a loophole whereby a rogue DNS
8188 administrator could create reverse DNS records to break through a
8189 wildcarded host restriction in an ACL.
8191 37. If a user filter or a system filter that ran in a subprocess used any of
8192 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8193 the wrong values were passed to the pipe command ($thisaddress had the
8194 value of $0, $0 had the value of $1, etc). This bug was introduced by
8195 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8197 38. Improved the line breaking for long SMTP error messages from ACLs.
8198 Previously, if there was no break point between 40 and 75 characters, Exim
8199 left the rest of the message alone. Two changes have been made: (a) I've
8200 reduced the minimum length to 35 characters; (b) if it can't find a break
8201 point between 35 and 75 characters, it looks ahead and uses the first one
8202 that it finds. This may give the occasional overlong line, but at least the
8203 remaining text gets split now.
8205 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8206 file descriptors might be low, and that setting 1000 would always raise it.
8207 It turns out that in some environments, the limit is already over 1000 and
8208 that lowering it causes trouble. So now Exim takes care not to decrease it.
8210 40. When delivering a message, the value of $return_path is set to $sender_
8211 address at the start of routing (routers may change the value). By an
8212 oversight, this default was not being set up when an address was tested by
8213 -bt or -bv, which affected the outcome if any router or filter referred to
8216 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8217 log only when "message" or "log_message" are set. However, if one of the
8218 conditions was an address verification, or a call to a nested ACL, the
8219 messages generated by the underlying test were being passed through. This
8220 no longer happens. The underlying message is available in $acl_verify_
8221 message for both "message" and "log_message" expansions, so it can be
8222 passed through if needed.
8224 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8225 new expansion $bh_ to give the encoded byte string without charset
8226 translation. Translation happens only if iconv() is available; HAVE_ICONV
8227 indicates this at build time. HEADERS_CHARSET gives the charset to
8228 translate to; headers_charset can change it in the configuration, and
8229 "headers charset" can change it in an individual filter file.
8231 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8232 that creates RFC 2047 encoded "words" labels them as that charset instead
8233 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8234 expansion operator; (ii) when Exim creates a From: line for a local
8235 message; (iii) when a header line is rewritten to include a "phrase" part.
8237 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8238 buggy, causing it to skip the first lines of messages whose message ID
8239 ended in 'D'. This would not have bitten before Exim release 4.14, because
8240 message IDs were unlikely to end in 'D' before then. The effect was to have
8241 incorrect size information for certain domains.
8243 45. #include "config.h" was missing at the start of the crypt16.c module. This
8244 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8247 46. If there was a timeout during a "random" callout check, Exim treated it as
8248 a failure of the random address, and carried on sending RSET and the real
8249 address. If the delay was just some slowness somewhere, the response to the
8250 original RCPT would be taken as a response to RSET and so on, causing
8251 mayhem of various kinds.
8253 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8254 when I implemented it. It didn't allow for the fact that some option values
8255 may legitimately be negative (e.g. size_addition), and it didn't even do
8256 the right test for positive values.
8258 48. Domain names in DNS records are case-independent. Exim always looks them up
8259 in lower case. Some resolvers return domain names in exactly the case they
8260 appear in the zone file, that is, they may contain uppercase letters. Not
8261 all resolvers do this - some return always lower case. Exim was treating a
8262 change of case by a resolver as a change of domain, similar to a widening
8263 of a domain abbreviation. This triggered its re-routing code and so it was
8264 trying to route what was effectively the same domain again. This normally
8265 caused routing to fail (because the router wouldn't handle the domain
8266 twice). Now Exim checks for this case specially, and just changes the
8267 casing of the domain that it ultimately uses when it transmits the message
8270 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8273 50. If a filter generated a file delivery with a non-absolute name (possible if
8274 no home directory exists for the router), the forbid_file option was not
8277 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8278 the existing equality matching.
8280 52. Exim was using ints instead of ino_t variables in some places where it was
8281 dealing with inode numbers.
8283 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8284 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8285 called TMPDIR, and if it finds it is different, it changes its value.
8287 54. The smtp_printf() function is now made available to local_scan() so
8288 additional output lines can be written before returning. There is also an
8289 smtp_fflush() function to enable the detection of a dropped connection.
8290 The variables smtp_input and smtp_batched_input are exported to
8293 55. Changed the default runtime configuration: the message "Unknown user"
8294 has been removed from the ACL, and instead placed on the localuser router,
8295 using the cannot_route_message feature. This means that any verification
8296 failures that generate their own messages won't get overridden. Similarly,
8297 the "Unrouteable address" message that was in the ACL for unverifiable
8298 relay addresses has also been removed.
8300 56. Added hosts_avoid_esmtp to the smtp transport.
8302 57. The exicyclog script was not checking for the esoteric option
8303 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8304 will work only if exicyclog is run under the appropriate euid.
8306 58. Following a discussion on the list, the rules by which Exim recognises line
8307 endings on incoming messages have been changed. The -dropcr and drop_cr
8308 options are now no-ops, retained only for backwards compatibility. The
8309 following line terminators are recognized: LF CRLF CR. However, special
8310 processing applies to CR:
8312 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8313 nor a local message in the state where . is a terminator.
8315 (ii) If a bare CR is encountered in a header line, an extra space is added
8316 after the line terminator so as not to end the header. The reasoning
8317 behind this is that bare CRs in header lines are most likely either
8318 to be mistakes, or people trying to play silly games.
8320 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8321 was being incorrectly given as 18 bytes larger than it should have been.
8322 This is a VOB (very old bug).
8324 60. This may never have affected anything current, but just in case it has:
8325 When the local host is found other than at the start of a list of hosts,
8326 the local host, those with the same MX, and any that follow, are discarded.
8327 When the list in question was part of a longer list of hosts, the following
8328 hosts (not currently being processed) were also being discarded. This no
8329 longer happens. I'm not sure if this situation could ever has previously
8332 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8335 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8336 header. This is something that is recommended in a new Internet Draft, and
8337 is something that is documented as being done by Sendmail. There are two
8338 possible values. For messages generated by the autoreply transport, Exim
8341 Auto-Submitted: auto-replied
8343 whereas for all other generated messages (e.g. bounces) it adds
8345 Auto-Submitted: auto-generated
8347 63. The "personal" condition in filters now includes a test for the
8348 Auto-Submitted: header. If it contains the string "auto-" the message it
8349 not considered personal.
8351 64. Added rcpt_include_affixes as a generic transport option.
8353 65. Added queue_only_override (default true).
8355 66. Added the syslog_duplication option.
8357 67. If what should have been the first header line of a message consisted of
8358 a space followed by a colon, Exim was mis-interpreting it as a header line.
8359 It isn't of course - it is syntactically invalid and should therefore be
8360 treated as the start of the message body. The misbehaviour could have
8361 caused a number of strange effects, including loss of data in subsequent
8362 header lines, and spool format errors.
8364 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8365 client host had authenticated. This control can now be exercised by an ACL
8366 for more flexibility.
8368 69. By default, callouts do not happen when testing with -bh. There is now a
8369 variant, -bhc, which does actually run the callout code, including
8370 consulting and updating the callout cache.
8372 70. Added support for saslauthd authentication, courtesy of Alexander
8375 71. If statvfs() failed on the spool or log directories while checking their
8376 size for availability, Exim confusingly gave the error "space shortage".
8377 Furthermore, in debugging mode it crashed with a floating point exception.
8378 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8379 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8380 serious problem, Exim now writes to the main and panic logs when this
8381 happens, with details of the failure. It then refuses to accept the
8382 incoming message, giving the message "spool directory problem" or "log
8383 directory problem" with a 421 code for SMTP messages.
8385 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8386 0, 1, and 2 exist, because some OS complain for execs without them (see
8387 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8388 descriptors. However, the code omitted to check that the open succeeded,
8389 causing mysterious errors if for some reason the permissions on /dev/null
8390 got screwed. Now Exim writes a message to the main and panic logs, and
8391 bombs out if it can't open /dev/null.
8393 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8394 interact so that it is all more flexible. It is supposed to remain
8395 backwards compatible. Also added extra_local_interfaces.
8397 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8398 to bomb out with an assertion failure - to the client this appears as a
8399 connection drop. This problem occurs in the part of the code that was taken
8400 from the Samba project. Fortunately, the assertion is in a very simple
8401 function, so I have fixed this by reproducing the function inline in the
8402 one place where it is called, and arranging for authentication to fail
8403 instead of killing the process with assert().
8405 75. The SPA client code was not working when the server requested OEM rather
8406 than Unicode encoding.
8408 76. Added code to make require_files with a specific uid setting more usable in
8409 the case where statting the file as root fails - usually a non-root-mounted
8410 NFS file system. When this happens and the failure is EACCES, Exim now
8411 forks a subprocess and does the per-uid checking as the relevant uid.
8413 77. Added process_log_path.
8415 78. If log_file_path was not explicitly set, a setting of check_log_space or
8416 check_log_inodes was ignored.
8418 79. If a space check for the spool or log partitions fails, the incident is now
8419 logged. Of course, in the latter case the data may get lost...
8421 80. Added the %p formatting code to string_format() so that it can be used to
8422 print addresses in debug_print(). Adjusted all the address printing in the
8423 debugging in store.c to use %p rather than %d.
8425 81. There was a concern that a line of code in smtp_in.c could overflow a
8426 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8427 initially expressed, the concern was not well-founded, because trailing
8428 spaces are removed early. However, if the trailing spaces were followed by
8429 a NULL, they did not get removed, so the overflow was possible. Two fixes
8432 (a) I re-wrote the offending code in a cleaner fashion.
8433 (b) If an incoming SMTP command contains a NULL character, it is rejected
8436 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8437 runs initgroups(), so that if the Exim user is in any additional groups,
8438 they will be used during message reception.
8444 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob