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.
173 JH/33 Fix free for live variable $value created by a ${run ...} expansion.
174 Although not seen, this could have resulted in a SIGSEGV.
180 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
181 after reception to before a subsequent reception. This should
182 mean slightly faster delivery, and also confirmation of reception
185 JH/02 Move from using the pcre library to pcre2. The former is no longer
186 being developed or supported (by the original developer).
188 JH/03 Constification work in the filters module required a major version
189 bump for the local-scan API. Specifically, the "headers_charset"
190 global which is visible via the API is now const and may therefore
191 not be modified by local-scan code.
193 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
194 sendfile() didi not account for the way the ClamAV driver code called it.
196 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
197 time check was being done for every character; replace that with one
200 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
201 sent was prefixed with a length byte.
203 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
204 RFC 5321. Previously Dovecot (at least) would log errors during
207 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
208 in 4.61 without a reason logged. Binaries will be bigger, which might
209 matter on diskspace-constrained systems, but debug is easier.
211 JH/09 Fix macro-definition during "-be" expansion testing. The move to
212 write-protected store for macros had not accounted for these runtime
213 additions; fix by removing this protection for "-be" mode.
215 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
216 handing out large-numbered file descriptors, violating the usual Unix
217 assumption (and required by Posix) that the lowest possible number will be
218 allocated by the kernel when a new one is needed. In the daemon, and any
219 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
220 are not useable for FD_SET() [and hence select()] and overwrite the stack.
221 Assorted crashes happen.
223 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
224 main-section options or in a connect ACL, the value from the first ever
225 connection was never replaced for subsequent connections. Found by
228 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
229 though only once PCRE2 was introduced: the memory accounting used under
230 debug offset allocations by an int, giving a hard trap in early startup.
231 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
233 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
234 with underbars is given. The write-protection of configuration introduced
235 in 4.95 trapped when normalisation was applied to an option not needing
238 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
240 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
241 failing to close the certificates directory, every hour or any time it
244 JH/16 Debugging initiated by an ACL control now continues through into routing
245 and transport processes. Previously debugging stopped any time Exim
246 re-execs, or for processing a queued message.
248 JH/17 The "expand" debug selector now gives more detail, specifically on the
249 result of expansion operators and items.
251 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
252 bad comparison between the option value and the name of the file to
253 be included was done, and a mismatch was wrongly identified.
254 4.88 to 4.95 are affected.
256 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
258 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
259 supplied as the db file-pair basename. Previously, if a directory
260 path was given, for example via the autoreply "once" option, the DB
261 file.pag and file.dir files would be created in that directory's
264 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
265 "taint" log_selector. These were previously deprecated.
267 JH/22 Fix static address-list lookups to properly return the matched item.
268 Previously only the domain part was returned.
270 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
271 the call into OpenSSL to send a TLS Close was being repeated; this
272 resulted in the library waiting for the peer's Close. If that was never
273 sent we waited forever. Fix by tracking send calls.
275 JH/24 The ${run} expansion item now expands its command string elements after
276 splitting. Previously it was before; the new ordering makes handling
277 zero-length arguments simpler. The old ordering can be obtained by
278 appending a new option "preexpand", after a comma, to the "run".
280 JH/25 Taint-check exec arguments for transport-initiated external processes.
281 Previously, tainted values could be used. This affects "pipe", "lmtp" and
282 "queryprogram" transport, transport-filter, and ETRN commands.
283 The ${run} expansion is also affected: in "preexpand" mode no part of
284 the command line may be tainted, in default mode the executable name
287 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
288 the the facility was not passed across execs, and only the first message
289 passed over a connection could use BDAT; any further ones using DATA.
291 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
292 uses $sending_ip_address and an interface is specified.
293 Previously any use of the local address in the EHLO name disabled
294 PIPECONNECT, the common case being to use the rDNS of it.
296 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
297 resumption. Previously verify failed because no certificate status is
298 passed on the wire for the restarted session. Fix by using the recorded
299 ocsp status of the stored session for the new connection.
301 JH/29 TLS resumption: the key for session lookup in the client now includes
302 more info that a server could potentially use in configuring a TLS
303 session, avoiding oferring mismatching sessions to such a server.
304 Previously only the server IP was used.
306 JH/30 Fix string_copyn() for limit greater than actual string length.
307 Previously the copied amount was the limit, which could result in a
308 overlapping memcpy for newly allocated destination soon after a
309 source string shorter than the limit. Found/investigated by KM.
311 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
312 close; it may be needed for a subsequent connection. This caused a
313 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
315 JH/32 Fix CHUNKING for a second message on a connection when the first was
316 rejected. Previously we did not reset the chunking-offered state, and
317 erroneously rejected the BDAT command. Investigation help from
320 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
321 an empty address. Previously the expansion returned an error.
323 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
324 proxy. Previously these were misparsed, leading to paniclog entries.
330 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
331 related applications. Previously an "H" was used where available info
332 says that "M" should be, so change to match.
334 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
335 as arguments, so an implementation trying to copy these into a local
336 buffer was taking a taint-enforcement trap. Fix by using dynamically
337 created buffers. Similar fix for radius expansion condition.
339 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
340 reasonable, eg. to count headers. Fix by using dynamically created
341 buffers rather than a local. Do similar fixes for ACL actions "dcc",
342 "log_reject_target", "malware" and "spam"; the arguments are expanded
343 so could be handling tainted values.
345 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
346 broken the (no-op) support for this sendmail command. Restore it
347 to doing nothing, silently, and returning good status.
349 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
350 record path was given (or the default used) without a leading directory
351 path, an error occurred on trying to open it. Use the transport's working
354 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
355 transport. Previously it was the name on the DNS A-record; use instead
356 the head of the CNAME chain leading there (if there is one). This seems
357 to align better with RFC 6125.
359 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
360 smtp_accept_max_per_host allocated resources which were not released
361 when the limit was exceeded. This eventually crashed the daemon. Fix
362 by adding a release action in that path.
364 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
365 expanded; previously using tainted values was rejected. Fix by using
366 dynamically-created buffers.
368 JH/09 Relax restrictions on ACL verify condition needing access to message
369 headers. Previously they were only permitted in data and non-smtp ACLs;
370 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
371 not_blind, header_sender and header_names_ascii verification.
373 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
374 Previously a macro used one argument twice; when called with the
375 argument as an expression having side-effects, incorrect operation
376 resulted. Use an inlineable function.
378 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
379 held open for a verify callout. Previously this wan not accounted for
380 and a corrupt onward SMTP conversation resulted.
382 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
383 Previously they were embedded naively in the constructed address; when
384 needed, strip the quoting and quote the entire local_part.
385 Also make the inbound_srs expansion condition handle quoting.
387 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
388 excluded, not matching the documentation.
390 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
391 was given for the sqlite_dbfile a trap resulted.
393 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
394 "name" argument resulted in a trap. There is no reason to disallow such;
395 this was a coding error.
397 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
398 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
399 spent suspended, ignoring the POSIX definition. Previously we assumed
400 it did and a constant offset from real time could be used as a correction.
401 Change to using the same clock source for the start-of-message and the
402 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
403 exists, just to get a clock slightly more aligned to reality.
405 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
406 RFC says it is optional some validators care. The missing char was not
407 intended but triggered by a line-wrap alignment. Discovery and fix by
408 Guillaume Outters, hacked on by JH.
410 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
411 name being quoted was tainted a trap would be taken. Fix by using
412 dynamically created buffers. The routine could have been called by a
413 rewrite with the "h" flag, by using the "-F" command-line option, or
414 by using a "name=" option on a control=submission ACL modifier.
416 JH/19 SPF: change the Authentication-Results expansion component to give
417 smtp.helo when the sender domain is empty. Previously it gave
420 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
421 not in 127.0.0.0/8 to help in spotting list domains taken over by a
422 domain-parking registrar.
424 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
425 Previously when a whitespace character was specified it was not inserted
426 after removing the newline.
428 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
429 the domain part of the recipient address. This overrides any tls_sni
430 option set, which was previously used.
432 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
435 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
436 is_tainted() had an off-by-one error in the overenthusiastic direction.
437 Find and fix by Gavan. Although NetBSD is not a supported platform for
438 4.94 this bug could affect other platforms.
440 PP/01 Fix default prime selection to be consistent.
441 One path used ike23 still, instead of exim.dev.20160529.3; now both
442 execution flows will use the same DH primes (currently
443 exim.dev.20160529.3).
445 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
446 option in smtp transport, to match the documentation. Previously
447 verification was not being done in some cases where it should have been.
449 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
450 than one server was defined and depending on the platform memory layout
451 details, an internal consistency trap could be hit while walking the list
454 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
455 files. The value is used by the authresults expansion item. Previously
456 if this was used in a router or transport, a crash could result.
458 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
459 have a sufficient incidence of EINTR returns from select that an
460 interaction with other operations done by the main daemon loop exposed
461 a bug in the error-handling. This was benign apart from the log
464 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
465 for consistency with delivered "=>" and failed "**" lines. While we're
466 there, handle PRX and TFO.
468 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
469 applied. This resulted, if any header-line rewrite rules were configured,
470 in a panic-log triggerable by sending a message with a long address in
471 a header. Fix by increasing the arbitrary limit to larger than a single
472 (dewrapped) 5322 header line maximum size.
474 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
475 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
476 RFC 6648 which deprecates X- options in protocols as a general practice.
477 Changeover between the implementations is handled by the mechanisms
480 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
481 a remote which uses callout/hold. Previously the local was queued.
483 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
486 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
487 log line. The asterisk is now consistently appended to the remote IP
488 (and port, if given), and will also be provided on defer and fail log
489 lines. Previously it could be placed on the local IP if that was being
490 logged, and was only provided on delivery lines.
492 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
494 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
495 authenticator client_send option. Previously the next char, after a pair
496 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
497 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
498 documentation. There is still no way to get a leading ^ immediately
499 after a NUL (ie. for the password of a PLAIN method authenticator.
501 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
502 files. For bad sizes read, delete the record and whine to paniclog.
504 JH/38 When logging an AUTH failure, as server, do not include sensitive
505 information. Previously, the credentials would be included if given
506 as part of the AUTH command line and an ACL denied authentication.
508 JH/39 Bug 2691: fix $local_part_data. When the matching list element
509 referred to a file, bad data was returned. This likely also affected
512 JH/40 The gsasl authenticator now supports caching of the salted password
513 generated by the client-side implementation. This required the addition
514 of a new variable: $auth4.
516 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
517 left undeleted; the attempt to re-create it then failed - resulting in
518 the usual "SIGHUP tp have daemon reload configuration" to not work.
519 This affected any platform not supporting "abstract" Unix-domain
520 sockets (i.e. not Linux).
522 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
523 recipients" response to RCPT in a later response, with a 250. The
524 previous coding assumed this would not happen, and under PIPELINING
525 would result in both lost and duplicate recipients for a message.
527 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
528 Previously the weighting was incorrectly applied. Similar fix for socks
529 proxies. Found and fixed by Heiko Schlichting.
531 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
532 not handle sub-lists included using the +namedlist syntax. While
533 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
534 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
537 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
538 Previously the permanent pool was used, so the sore could not be freed.
539 This meant a connection with many messages would use continually-growing
542 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
543 per-pool so as not to waste too much space. Previously a constant size
544 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
545 DOS attacks harder. The cost is wasted memory use in the larger blocks.
547 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
548 same space repeatedly, and a smaller process footprint.
550 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
551 for GnuTLS. Previously the name started "%s" which, while not a bug,
552 looked as if if might be one.
554 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
555 "continued connection") make the $proxy_* variables available. Previously
556 the information was not passed across the exec() call for subsequent
557 transport executions. This also mean that the log lines for the
558 messages can show the proxy information.
560 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
561 receive time. With modern systems the difference is significant.
562 The historical behaviour can be restored by disabling (a new) log_selector
563 "queue_time_exclusive".
565 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
566 content) were specifically tested for. Now, also cover expansions
567 resulting in ACL names and inline ACL content.
569 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
570 making it unusable in complex expressions.
572 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
573 fail with a taint-check on the spoolfile name, and leave the message
576 HS/01 Enforce absolute PID file path name.
578 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
580 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
582 PP/02 Bug 2643: Correct TLS DH constants.
583 A missing NUL termination in our code-generation tool had led to some
584 incorrect Diffie-Hellman constants in the Exim source.
585 Reported by kylon94, code-gen tool fix by Simon Arlott.
587 PP/03 Impose security length checks on various command-line options.
588 Fixes CVE-2020-SPRSS reported by Qualys.
590 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
591 better. Reported by Qualys.
593 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
594 providing a particularly obnoxious sender full name.
597 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
599 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
602 PP/08 Change default for recipients_max from unlimited to 50,000.
604 PP/09 Fix security issue with too many recipients on a message (to remove a
605 known security problem if someone does set recipients_max to unlimited,
606 or if local additions add to the recipient list).
607 Fixes CVE-2020-RCPTL reported by Qualys.
609 PP/10 Fix security issue in SMTP verb option parsing
610 Fixes CVE-2020-EXOPT reported by Qualys.
612 PP/11 Fix security issue in BDAT state confusion.
613 Ensure we reset known-good where we know we need to not be reading BDAT
614 data, as a general case fix, and move the places where we switch to BDAT
615 mode until after various protocol state checks.
616 Fixes CVE-2020-BDATA reported by Qualys.
618 HS/03 Die on "/../" in msglog file names
620 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
621 the Exim runtime user are allowed to create files.
623 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
626 QS/03 When reading the output from interpreted forward files we do not
627 pass the pipe between the parent and the interpreting process to
628 executed child processes (if any).
630 QS/04 Always die if requested from internal logging, even is logging is
633 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
634 the API; compilation noo longer completes with DMARC support included.
635 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
636 been reported on other platforms.
638 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
640 JH/56 Make the majority of info read from config files readonly, for defence-in-
641 depth against exploits. Suggestion by Qualys.
642 Not supported on Solaris 10.
644 JH/57 Fix control=fakreject for a custom message containing tainted data.
645 Previously this resulted in a log complaint, due to a re-expansion present
646 since fakereject was originally introduced.
648 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
649 modifier was given, a loop resulted.
651 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
652 pipelined SMTP command followed the BDAT LAST then it would be
653 incorrectly treated as part of the message body, causing a verification
656 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
657 require looser validation rules than those for 821-level addresses,
658 which only permit IP addresses.
664 JH/01 Avoid costly startup code when not strictly needed. This reduces time
665 for some exim process initialisations. It does mean that the logging
666 of TLS configuration problems is only done for the daemon startup.
668 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
670 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
671 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
674 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
675 DKIM signing is being done. Previously a transport_filter would always
676 disable CHUNKING, falling back to traditional DATA.
678 JH/05 Regard command-line recipients as tainted.
680 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
682 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
683 PAM library frees one of the arguments given to it, despite the
684 documentation. Therefore a plain malloc must be used.
686 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
687 on-stack buffers were used, resulting in a taint trap when DSN information
688 copied from a received message was written into the buffer.
690 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
691 the ordering of its ARC headers. This caused a crash.
693 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
694 a new record was being constructed with information from the peer, a trap
697 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
698 installation would get error messages from DMARC verify, when it hit the
699 nonexistent file indicated by the default. Distros wanting DMARC enabled
700 should both provide the file and set the option.
701 Also enforce no DMARC verification for command-line sourced messages.
703 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
704 could, depending on the platform, hang at the STARTTLS response.
706 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
707 message on a connection. Previously if one message had ARC headers and
708 the following one did not, a crash could result when adding an
709 Authentication-Results: header.
711 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
712 Exim main code and Exim-related utities. The introduction of taint
713 tracking also did many adjustments to string handling. Since then, eximon
714 frequently terminated with an assert failure.
716 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
717 check for 452 responses. This slightly helps the inefficieny of doing
718 a large alias-expansion into a recipient-limited target. The max_rcpt
719 transport option still applies (and at the current default, will override
720 the new feature). The check is done for either cause of synch, and forces
721 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
722 connection. The new facility is not tunable at this time.
724 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
725 library live data was being used, so the results became garbage. Make
726 copies while it is still usable.
728 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
729 on delivery deferred (==) and failed (**) lines (if a delivery was
730 attemtped). Previously it was only on completion (=>) lines.
732 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
733 for the expansion of the server_scram_iter and server_scram_salt options.
735 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
736 are now specifically given a NO_DATA response without hitting the system
737 resolver. The library goes on to do the now-standard TXT lookup.
738 Use of dnsdb lookups is not affected.
740 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
741 only retrieve the errormessage once. Previously two calls to dlerror()
742 were used, and the second one (for mainlog/paniclog) retrieved null
745 JH/20 Taint checking: disallow use of tainted data for
746 - the appendfile transport file and directory options
747 - the pipe transport command
748 - the autoreply transport file, log and once options
749 - file names used by the redirect router (including filter files)
751 - paths used by single-key lookups
752 Previously this was permitted.
754 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
755 adjusted the size of a major service buffer; this failed because the
756 buffer was in use at the time. Change to a compile-time increase in the
757 buffer size, when this authenticator is compiled into exim.
759 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
760 previous fast-mode was untenable in the face of glibs using mmap to
761 support larger malloc requests.
763 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
764 New values supported, if defined on system where compiled:
765 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
766 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
768 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
769 running a limited number of proceses in parallel, a benefit is gained. The
770 amount varies with the platform hardware and load. The use of the option
771 queue_run_in_order means we cannot do this, as ordering becomes
774 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
775 had introduced a string-copy (for ensuring NUL-termination) which was not
776 appropriate for that case, which can include embedded NUL bytes in the
777 block of data. Investigation showed the copy to actually be needless, the
778 data being length-specified.
780 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
781 done during a receiving connection, and both used TLS, global info was
782 used rather than per-connection info for tracking the state of data
783 queued for transmission. This could result in a connection hang.
785 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
786 Previously, when delivering serveral messages down a single connection
787 only the first would provide a SIZE. This was due to the size information
788 not being properly tracked.
790 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
791 TAI (at 37 seconds currently), pretend to be in UTC for time-related
792 expansion and logging. Previously, spurious values such as a future
793 minute could be seen.
795 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
796 it could crash from a null-deref. This could also affect the
797 ${addresses: } operator and ${readsock } item.
799 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
800 message following a mime one, the variable was not reset.
802 JH/30 When an pipelined-connect fails at the first response, assume incorrect
803 cached capability (perhaps the peer reneged?) and immediately retry in
806 JH/31 Fix spurious detection of timeout while writing to transport filter.
808 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
809 an attempt to copy the string was made before checking it.
811 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
812 taint of the lookup key was maintained; we now regard the presence in the
813 filesystem as sufficient validation.
815 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
816 argument is supplied.
818 JH/35 The dsearch lookup now requires that the directory is an absolute path.
819 Previously this was not checked, and nonempty relative paths made an
820 access under Exim's current working directory.
822 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
823 Previously no event was raised.
825 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
826 parameter supplied by the sender MAIL FROM command. Previously it was
827 ignored, and only the check_spool_space option value for the required
830 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
831 the size of the signing public-key. Previously it was instead giving
832 the size of the signature hash.
834 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
835 the default. See the (new) dkim_verify_min_keysizes option.
837 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
838 and an ACL use a lookup for checking either the local_part or domain,
839 stale data could be accessed. Ensure that variable references are
840 dropped between messages.
842 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
843 by the client was not checked as pointing within response data before
844 being used. A malicious client could thus cause an out-of-bounds read and
845 possibly gain authentication. Fix by adding the check.
847 JH/42 Internationalisation: change the default for downconversion in the smtp
848 transport to be "if needed". Previously it was "as previously set" for
849 the message, which usually meant "if needed" for message-submission but
850 "no" for everything else. However, MTAs have been seen using SMTPUTF8
851 even when the envelope addresses did not need it, resulting in forwarding
852 failures to non-supporting MTAs. A downconvert in such cases will be
853 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
854 transport. The change does mean that addresses needing conversion will
855 be converted when previously a delivery failure would occur.
857 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
858 response was received it would be used unchecked in a fail-DSN, violating
859 standards on line-length limits. Truncate if needed.
861 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
862 doesn't work. (Additionally add a new main config option to configure the
869 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
870 side, to decode a TLS 1.3 packet capture.
872 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
873 Previously the default library behaviour applied, sending two, each in
876 JH/03 Debug output for ACL now gives the config file name and line number for
879 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
881 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
883 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
884 buffer overrun for (non-chunking) other transports.
886 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
887 TLS1.3, means that a server rejecting a client certificate is not visible
888 to the client until the first read of encrypted data (typically the
889 response to EHLO). Add detection for that case and treat it as a failed
890 TLS connection attempt, so that the normal retry-in-clear can work (if
891 suitably configured).
893 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
894 and/or domain. Found and fixed by Jason Betts.
896 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
897 configuration). If a CNAME target was not a wellformed name pattern, a
900 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
901 the OS reports them interleaved with other addresses.
903 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
904 used both for input and for a verify callout, both encrypted, SMTP
905 responses being sent by the server could be lost. This resulted in
906 dropped connections and sometimes bounces generated by a peer sending
909 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
910 string. Previously it was possible to cause undefined behaviour in a
911 library routine (usually a crash). Found by "zerons".
913 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
916 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
917 API was removed, so update to use the newer ones.
919 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
920 any timeout set, is taking a long time. Previously we would hang on to a
921 rotated logfile "forever" if the input was arriving with long gaps
922 (a previous attempt to fix addressed lack, for a long time, of initial
925 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
926 shared (NFS) environment. The length of the tempfile name is now
927 4 + 16 ("hdr.$message_exim_id") which might break on file
928 systems which restrict the file name length to lower values.
931 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
932 shared (NFS) environment.
934 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
935 did for all versions <4.90). Notably -M, -m, --invert, -I may be
938 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
939 on some platforms for bit 31.
941 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
942 to changes apparently associated with TLS1.3 handling some of the APIs
943 previously used were either nonfunctional or inappropriate. Strings
944 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
945 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
946 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
947 This affects log line X= elements, the $tls_{in,out}_cipher variables,
948 and the use of specific cipher names in the encrypted= ACL condition.
950 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
952 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
953 verification result was not updated unless hosts_require_ocsp applied.
955 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
956 queue_list_requires_admin set to false, non-admin users were denied the
959 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
960 directory-of-certs mode. Previously they were advertised despite the
963 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
964 A single TCP connection by a client will now hold a TLS connection open
965 for multiple message deliveries, by default. Previously the default was to
968 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
969 default. If built with the facility, DANE will be used. The facility
970 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
972 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
973 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
974 must be defined and you must still, unless you define DISABLE_TLS, manage
975 the the include-dir and library-file requirements that go with that
976 choice. Non-TLS builds are still supported.
978 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
981 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
982 default. If the platform supports and has the facility enabled, it will
983 be requested on all coneections.
985 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
986 controlled by the build-time option SUPPORT_PIPE_CONNECT.
988 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
990 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
991 success-DSN messages. Previously the From: header was always the default
992 one for these; the option was ignored.
994 JH/28 Fix the timeout on smtp response to apply to the whole response.
995 Previously it was reset for every read, so a teergrubing peer sending
996 single bytes within the time limit could extend the connection for a
997 long time. Credit to Qualsys Security Advisory Team for the discovery.
999 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1000 delivery address, which leaked information of the results of local
1001 forwarding. Change to the original envelope recipient address, per
1004 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1005 requested. Previously not bounce was generated and a log entry of
1006 error ignored was made.
1008 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1010 JH/32 Introduce a general tainting mechanism for values read from the input
1011 channel, and values derived from them. Refuse to expand any tainted
1012 values, to catch one form of exploit.
1014 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1015 was unused and the unexpanded text used for the test. Found and
1016 fixed by Ruben Jenster.
1018 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1019 an attempt to use a TLS library read routine dereffed a nul pointer,
1022 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1023 them in our smtp response.
1025 JH/36 Have the general router option retry_use_local_part default to true when
1026 any of the restrictive preconditions are set (to anything). Previously it
1027 was only for check_local user. The change removes one item of manual
1028 configuration which is required for proper retries when a remote router
1029 handles a subset of addresses for a domain.
1031 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1032 link count into consideration.
1034 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1035 caused the extension of big_buffer, the following lines were ignored.
1037 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1038 accordance with RFC 2308. Previously there was no expiry, so a longlived
1039 receive process (eg. due to ACL delays) versus a short SOA value could
1042 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1044 JH/39 Promote DMARC support to mainline.
1046 JH/40 Bug 2452: Add a References: header to DSNs.
1048 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1049 parameters. The relevant library call is documented as "Deprecated: This
1050 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1051 3.6.0, DH parameters are negotiated following RFC7919."
1053 HS/06 Change the default of dnssec_request_domains to "*"
1055 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1056 carried on and emitted a BDAT command, even when PIPELINING was not
1059 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1060 buffer was used for the filename, resulting in a trap when tainted
1061 arguments (eg. $domain) were used.
1063 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1064 recommended to avoid a possible server-load attack. The feature can be
1065 re-enabled via the openssl_options main cofiguration option.
1067 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1068 for version 4.90 - adding a "more data" boolean to the arguments.
1069 Bumped the ABI version number also, this having been missed previously;
1070 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1071 respect of usage of smtp_printf() by either local_scan code or libraries
1072 accessed via the ${dlfunc } expansion item. Both will need coding
1073 adjustment for any calls to smtp_printf() to match the new function
1074 signature; a FALSE value for the new argument is always safe.
1076 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1077 the file-offset (which the Linux syscall does, and exim expects); this
1078 resulted in an indefinite loop.
1080 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1081 to do ARC verification. The Authentication-Results: header line added
1082 by the configuration then had no ARC item.
1084 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1085 an incorrect mode was used for reading the result, resulting in it being
1092 JH/01 Remove code calling the customisable local_scan function, unless a new
1093 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1095 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1096 non-signal-safe functions being used.
1098 JH/03 Bug 2269: When presented with a received message having a stupidly large
1099 number of DKIM-Signature headers, disable DKIM verification to avoid
1100 a resource-consumption attack. The limit is set at twenty.
1102 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1103 report of oldest_pass in ${authres } in consequence, and separate out
1104 some descriptions of reasons for verification fail.
1106 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1107 files in the spool were present and unlocked. A queue-runner could spot
1108 them, resulting in a duplicate delivery. Fix that by doing the unlock
1109 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1110 add more error-checking on spoolfile handling while that code is being
1113 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1114 No known attacks, no CVE, this is defensive hardening.
1116 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1117 a queue-runner could start a delivery while other operations were ongoing.
1118 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1119 Found and investigated by Tim Stewart. Fix by using the open message data
1120 file handle rather than opening another, and not locally closing it (which
1121 releases a lock) for that case, while creating the temporary .eml format
1122 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1124 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1125 $sender_verify_failure/$recipient_verify_failure to "random".
1127 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1130 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1131 Previously this would segfault.
1133 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1136 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1137 like zero, since the resolver should be doing this for us, But we need one
1138 as a CNAME but no MX presence gets the CNAME returned; we need to check
1139 that doesn't point to an MX to declare it "no MX returned" rather than
1140 "error, loop". A new main option is added so the older capability of
1141 following some limited number of chain links is maintained.
1143 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1145 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1146 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1147 now-deprecated earlier definitions used only specified the range up to TLS
1148 1.2 (in the older-version library docs).
1150 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1152 JH/15 Rework TLS client-side context management. Stop using a global, and
1153 explicitly pass a context around. This enables future use of TLS for
1154 connections to service-daemons (eg. malware scanning) while a client smtp
1155 connection is using TLS; with cutthrough connections this is quite likely.
1157 JH/16 Fix ARC verification to do AS checks in reverse order.
1159 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1161 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1162 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1163 Previously the "utf8" would be re-prepended for every additional message.
1165 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1166 Previously thery were accepted, resulting in issues when attempting to
1167 forward messages to a non-supporting MTA.
1169 PP/02 Let -n work with printing macros too, not just options.
1171 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1172 one parent address was copied, and bogus data was used at delivery-logging
1173 time. Either a crash (after delivery) or bogus log data could result.
1174 Discovery and analysis by Tim Stewart.
1176 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1177 Previously if the string ended mid-character, we did not insert the
1178 promised '?' replacement.
1180 PP/04 Documentation: current string operators work on bytes, not codepoints.
1182 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1183 should pack well giving a smaller memory footprint so better caching and
1184 therefore performance. Group the declarations where this can't be done so
1185 that the byte-sized flag variables are not interspersed among pointer
1186 variables, giving a better chance of good packing by the compiler.
1188 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1189 non-null, to avoid issues with sites running BATV. Previously reports were
1190 sent with an empty envelope sender so looked like bounces.
1192 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1193 The ignore_error flag wasn't being returned from the filter subprocess so
1194 was not set for later routers. Investigation and fix by Matthias Kurz.
1196 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1197 and a msg:complete for the whole, when a message is manually removed using
1198 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1200 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1201 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1202 hope that is portable enough.
1204 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1205 requiring a known-CA anchor certificate; make it now rely entirely on the
1206 TLSA as an anchor. Checking the name on the leaf cert against the name
1207 on the A-record for the host is still done for TA (but not for EE mode).
1209 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1210 included in delivery lines for non-proxied connections, when compiled with
1211 SUPPORT_SOCKS and running with proxy logging enabled.
1213 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1214 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1215 move the existing event to fire before the normal logging of message
1216 failure so that custom logging is bracketed by normal logging.
1218 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1219 msg:fail:internal event. Developement by Matthias Kurz.
1221 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1222 far too small for todays use of crypto signatures stored there. Go all
1223 the way to the max DNS message size of 64kB, even though this might be
1224 overmuch for IOT constrained device use.
1226 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1227 copy a string over itself. The library routine is documented as not
1228 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1230 JH/32 For main options check_spool_space and check_inode_space, where the
1231 platform supports 64b integers, support more than the previous 2^31 kB
1232 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1233 the previous G, M, k.
1235 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1236 $authenticated_fail_id variable on authentication failure. Previously
1239 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1240 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1241 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1242 GNUTLS_SEC_PARAM_MEDIUM.
1244 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1245 side. Previously we would continue as if no SNI had been received.
1247 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1248 "<" character, which should be a list-separator specification, we walked
1249 off past the nul-terimation.
1251 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1252 causes) even when the retry time is not yet met. Previously they were
1253 not, meaning that when (say) an account was over-quota and temp-rejecting,
1254 and multiple senders' messages were queued, only one sender would get
1255 notified on each configured delay_warning cycle.
1257 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1259 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1260 cipher-suites, an error can be left on the stack even for a succeeding
1261 accept; this results in impossible error messages when a later operation
1264 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1265 return error codes indicating retry. Under TLS1.3 this becomes required.
1267 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1268 it only wrote the new authenticators, resulting in a lack of tracking of
1269 peer changes of ESMTP extensions until the next cache flush.
1271 JH/41 Fix the loop reading a message header line to check for integer overflow,
1272 and more-often against header_maxsize. Previously a crafted message could
1273 induce a crash of the recive process; now the message is cleanly rejected.
1275 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1276 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1282 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1283 When redis_servers is set to a list of > 1 element, and the Redis servers
1284 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1285 case of MOVED into a DEFER case instead, thus moving the query onto the
1286 next server in the list. For a cluster of N elements, all N servers must
1287 be defined in redis_servers.
1289 GF/02 Catch and remove uninitialized value warning in exiqsumm
1290 Check for existence of @ARGV before looking at $ARGV[0]
1292 JH/01 Replace the store_release() internal interface with store_newblock(),
1293 which internalises the check required to safely use the old one, plus
1294 the allocate and data copy operations duplicated in both (!) of the
1295 extant use locations.
1297 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1298 modifier. This matches the restriction on the commandline.
1300 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1301 Previously only the last row was returned.
1303 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1304 we assumed that tags in the header were well-formed, and parsed the
1305 element content after inspecting only the first char of the tag.
1306 Assumptions at that stage could crash the receive process on malformed
1309 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1310 While running the DKIM ACL we operate on the Permanent memory pool so that
1311 variables created with "set" persist to the DATA ACL. Also (at any time)
1312 DNS lookups that fail create cache records using the Permanent pool. But
1313 expansions release any allocations made on the current pool - so a dnsdb
1314 lookup expansion done in the DKIM ACL releases the memory used for the
1315 DNS negative-cache, and bad things result. Solution is to switch to the
1316 Main pool for expansions.
1317 While we're in that code, add checks on the DNS cache during store_reset,
1318 active in the testsuite.
1319 Problem spotted, and debugging aided, by Wolfgang Breyha.
1321 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1322 When none of the hosts presented to a transport match an already-open
1323 connection, close it and proceed with the list. Previously we would
1324 queue the message. Spotted by Lena with Yahoo, probably involving
1327 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1328 Previously a spurious "250 OK id=" response was appended to the proper
1331 JH/08 The "support for" informational output now, which built with Content
1332 Scanning support, has a line for the malware scanner interfaces compiled
1333 in. Interface can be individually included or not at build time.
1335 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1336 by the template makefile "src/EDITME". The "STREAM" support for an older
1337 ClamAV interface method is removed.
1339 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1340 rows affected is given instead).
1342 JH/11 The runtime Berkeley DB library version is now additionally output by
1343 "exim -d -bV". Previously only the compile-time version was shown.
1345 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1346 SMTP connection. Previously, when one had more recipients than the
1347 first, an abortive onward connection was made. Move to full support for
1348 multiple onward connections in sequence, handling cutthrough connection
1349 for all multi-message initiating connections.
1351 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1352 routers. Previously, a multi-recipient message would fail to match the
1353 onward-connection opened for the first recipient, and cause its closure.
1355 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1356 a timeout on read on a GnuTLS initiating connection, resulting in the
1357 initiating connection being dropped. This mattered most when the callout
1358 was marked defer_ok. Fix to keep the two timeout-detection methods
1361 JH/15 Relax results from ACL control request to enable cutthrough, in
1362 unsupported situations, from error to silently (except under debug)
1363 ignoring. This covers use with PRDR, frozen messages, queue-only and
1366 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1368 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1369 metadata, resulting in a crash in free().
1371 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1372 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1373 Broken also in d185889f4, with init system revamp.
1375 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1376 Previously we abruptly closed the connection after reading a malware-
1377 found indication; now we go on to read the "scan ok" response line,
1380 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1381 ACL. Previously, a crash would result.
1383 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1384 macro text after a replacement (previously it was only once per line) and
1385 by skipping builtin macros when searching for an uppercase lead character.
1387 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1388 for the build is renamed.
1390 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1391 was allocated for every new TLS startup, meaning one per message. Fix
1392 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1394 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1395 reported the original. Fix to report (as far as possible) the ACL
1396 result replacing the original.
1398 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1399 OpenSSL. Certificate information is loaded for every new TLS startup,
1400 and the resources needed to be freed.
1402 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1404 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1407 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1408 DATA response info to the (existing) per-recipient response info for
1409 the "C=" log element. It can have useful tracking info from the
1410 destination system. Patch from Simon Arlott.
1412 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1413 length value. Previously this would segfault.
1415 HS/02 Support Avast multiline protoocol, this allows passing flags to
1416 newer versions of the scanner.
1418 JH/28 Ensure that variables possibly set during message acceptance are marked
1419 dead before release of memory in the daemon loop. This stops complaints
1420 about them when the debug_store option is enabled. Discovered specifically
1421 for sender_rate_period, but applies to a whole set of variables.
1422 Do the same for the queue-runner and queue-list loops, for variables set
1423 from spool message files. Do the same for the SMTP per-message loop, for
1424 certain variables indirectly set in ACL operations.
1426 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1427 as a multi-recipient message from a mailinglist manager). The coding had
1428 an arbitrary cutoff number of characters while checking for more input;
1429 enforced by writing a NUL into the buffer. This corrupted long / fast
1430 input. The problem was exposed more widely when more pipelineing of SMTP
1431 responses was introduced, and one Exim system was feeding another.
1432 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1433 receiving system, and refused recipients seen by the sending system
1434 (propating to people being dropped from mailing lists).
1435 Discovered and pinpointed by David Carter.
1437 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1438 replaced by the ${authresults } expansion.
1440 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1442 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1443 allows proper process termination in container environments.
1445 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1446 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1448 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1449 and "err_perm", deprecated since 4.83 when the RFC-defined words
1450 "temperror" and "permerror" were introduced.
1452 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1453 transport-filters or DKIM-signing. The restriction was lost in the
1454 consolidation of verify-callout and delivery SMTP handling.
1455 Extend the restriction to also cover ARC-signing.
1457 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1458 in defer=pass mode supply a 450 to the initiator. Previously the message
1461 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1462 tls_require_ciphers is used as before.
1464 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1465 "pass_unscanned". Only tmpfails from the scanner are written to
1466 the paniclog, as they may require admin intervention (permission
1467 denied, license issues). Other scanner errors (like decompression
1468 bombs) do not cause a paniclog entry.
1470 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1471 Previously it was possible to log spurious information in receive log
1474 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1475 triggered odd behaviour from Outlook Express clients.
1477 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1480 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1481 since the IETF WG has not yet settled on that versus the original
1482 "bare" representation.
1484 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1485 Previously the millisecond value corrupted the output.
1486 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1487 corrupted the output.
1493 JH/01 Rework error string handling in TLS interface so that the caller in
1494 more cases is responsible for logging. This permits library-sourced
1495 string to be attached to addresses during delivery, and collapses
1496 pairs of long lines into single ones.
1498 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1499 during configuration. Wildcards are allowed and expanded.
1501 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1502 This permits better logging.
1504 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1505 do not close it down and have the child transport start it up again on
1506 the passed-on TCP connection. Instead, proxy the child (and any
1507 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1508 affected: the continued delivery log lines do not have any DNSSEC, TLS
1509 Certificate or OCSP information. TLS cipher information is still logged.
1511 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1512 identical IP addresses on different listening ports. Will also affect
1515 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1516 add noisy ifdef guards to special-case this sillyness.
1517 Patch from Bernd Kuhls.
1519 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1520 than 255 are no longer allowed.
1522 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1523 Disable the session-cache too, which might reduce our load. Since we
1524 currrectly use a new context for every connection, both as server and
1525 client, there is no benefit for these.
1526 GnuTLS appears to not support tickets server-side by default (we don't
1527 call gnutls_session_ticket_enable_server()) but client side is enabled
1528 by default on recent versions (3.1.3 +) unless the PFS priority string
1531 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1532 <https://reproducible-builds.org/specs/source-date-epoch/>.
1534 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1535 the check for any unsuccessful recipients did not notice the limit, and
1536 erroneously found still-pending ones.
1538 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1539 MSG_MORE. Only in-clear (not on TLS connections).
1541 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1542 transport-filter is involved we can buffer the headers in memory for
1543 creating the signature, and read the spool data file once for the
1544 signature and again for transmission.
1546 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1547 4.77 as the kernel support then wasn't solid, having issues in 64bit
1548 mode. Now, it's been long enough. Add support for FreeBSD also.
1550 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1551 case where the routing stage had gathered several addresses to send to
1552 a host before calling the transport for the first, we previously failed
1553 to close down TLS in the old transport process before passing the TCP
1554 connection to the new process. The new one sent a STARTTLS command
1555 which naturally failed, giving a failed delivery and bloating the retry
1556 database. Investigation and fix prototype from Wolfgang Breyha.
1558 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1559 false-negatives in the check that the sender had not preempted a response
1560 or prompt from Exim (running as a server), due to that code's lack of
1561 awareness of the SMTP input buffering.
1563 PP/04 Add commandline_checks_require_admin option.
1564 Exim drops privileges sanely, various checks such as -be aren't a
1565 security problem, as long as you trust local users with access to their
1566 own account. When invoked by services which pass untrusted data to
1567 Exim, this might be an issue. Set this option in main configuration
1568 AND make fixes to the calling application, such as using `--` to stop
1571 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1572 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1573 responses to those, into a single TLS record each way (this usually means
1574 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1577 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1578 affects you only if you're dancing at the edge of the param size limits.
1579 If you are, and this message makes sense to you, then: raise the
1580 configured limit or use OpenSSL 1.1. Nothing we can do for older
1583 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1584 cmdline element to get the documented default one. Previously it was
1587 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1588 are unsuable for non-message-specific reasons - eg. connection timeout,
1589 banner-time rejection.
1591 JH/16 Fix logging of delivery remote port, when specified by router, under
1594 PP/06 Repair manualroute's ability to take options in any order, even if one
1595 is the name of a transport.
1598 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1600 JH/17 Change the list-building routines interface to use the expanding-string
1601 triplet model, for better allocation and copying behaviour.
1603 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1604 Previously it was constructed the first time a possibly-matching string
1605 was met in the configuration file input during startup; now it is done
1608 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1609 compatible one, to avoid the (poorly documented) possibility of a config
1610 file in the working directory redirecting the DB files, possibly correpting
1611 some existing file. CVE-2017-10140 assigned for BDB.
1613 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1614 cache-hot. Previously, although the result was properly cached, the
1615 initial verify call returned a defer.
1617 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1618 the main verify for receipient in uncached-mode.
1620 JH/22 Retire historical build files to an "unsupported" subdir. These are
1621 defined as "ones for which we have no current evidence of testing".
1623 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1624 if present. Previously it was ignored.
1626 JH/24 Start using specified-initialisers in C structure init coding. This is
1627 a C99 feature (it's 2017, so now considered safe).
1629 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1630 if was a fixed-sized field and bitmask ops via macros; it is now more
1633 PP/07 GitHub PR 56: Apply MariaDB build fix.
1634 Patch provided by Jaroslav Škarvada.
1636 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1637 during Coverity cleanups [4.87 JH/47]
1638 Diagnosis and fix provided by Michael Fischer v. Mollard.
1640 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1641 the right size to place the terminating semicolon on its own folded
1642 line, the header hash was calculated to an incorrect value thanks to
1643 the (relaxed) space the fold became.
1645 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1646 and confused the parent.
1648 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1649 which could crash as a result. This could lead to undeliverable messages.
1651 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1654 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1655 matters on fast-turnover and PID-randomising systems, which were getting
1656 out-of-order delivery.
1658 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1659 a possibly-overlapping copy. The symptom was that "Remote host closed
1660 connection in response to HELO" was logged instead of the actual 4xx
1663 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1664 Previously only that bufferd was discarded, resulting in SYMTP command
1667 JH/32 DKIM: when a message has multiple signatures matching an identity given
1668 in dkim_verify_signers, run the dkim acl once for each. Previously only
1669 one run was done. Bug 2189.
1671 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1672 "panic the current process" to "deliberately defer". The panic log is
1673 still written with the problem list name; the mail and reject logs now
1674 get a temp-reject line for the message that was being handled, saying
1675 something like "domains check lookup or other defer". The SMTP 451
1676 message is still "Temporary local problem".
1678 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1679 A crafted sequence of BDAT commands could result in in-use memory beeing
1680 freed. CVE-2017-16943.
1682 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1683 from SMTP input. Previously it was always done; now only done for DATA
1684 and not BDAT commands. CVE-2017-16944.
1686 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1687 to the message (such as an overlong header line). Previously this was
1688 not done and we did not exit BDAT mode. Followon from the previous item
1689 though a different problem.
1695 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1696 than -2003 did; needs libidn2 in addition to libidn.
1698 JH/02 The path option on a pipe transport is now expanded before use.
1700 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1701 Patch provided by "Björn", documentation fix added too.
1703 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1704 missing a wire-to-host endian conversion.
1706 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1707 close after a BDAT command line could be taken as a following command,
1708 giving a synch failure. Fix by only checking for synch immediately
1709 before acknowledging the chunk.
1711 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1712 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1713 macro. Patches provided by Josh Soref.
1715 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1716 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1717 by IANA as a service extension. However, John Klensin suggests that we
1720 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1721 the dkim code may be unix-mode line endings rather than smtp wire-format
1722 CRLF, so prepend a CR to any bare LF.
1724 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1725 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1727 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1728 the first were themselves being wrongly included in the feed into dkim
1729 processing; with most chunk sizes in use this resulted in an incorrect
1730 body hash calculated value.
1732 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1733 DKIM signature block, for verification. Although advised against by
1734 standards it is specifically not ruled illegal.
1736 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1738 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1739 missing a body hash (the bh= tag).
1741 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1742 It seems that HAProxy sends the Proxy Protocol information in clear and
1743 only then does a TLS startup, so do the same.
1745 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1746 TCP connections (such as for Spamd) unless the daemon successfully set
1747 Fast Open mode on its listening sockets. This fixes breakage seen on
1748 too-old kernels or those not configured for Fast Open, at the cost of
1749 requiring both directions being enabled for TFO, and TFO never being used
1750 by non-daemon-related Exim processes.
1752 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1753 endings, at least on the first header line. Try to canonify any that get
1754 past that check, despite the cost.
1756 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1757 now limited to an arbitrary five deep, while parsing addresses with the
1758 strip_excess_angle_brackets option enabled.
1760 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1761 instead leave the unprompted TLS handshake in socket buffer for the
1762 TLS library to consume.
1764 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1766 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1768 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1769 region in one operation, for speed, and this leaves assigned pointers
1770 dangling. Add checks run only under the testsuite which checks all
1771 variables at a store-reset and panics on a dangling pointer; add code
1772 explicitly nulling out all the variables discovered. Fixes one known
1773 bug: a transport crash, where a dangling pointer for $sending_ip_address
1774 originally assigned in a verify callout, is re-used.
1776 PP/06 Drop '.' from @INC in various Perl scripts.
1778 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1780 PP/08 Reduce a number of compilation warnings under clang; building with
1781 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1782 should be warning-free.
1784 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1786 HS/01 Fix portability problems introduced by PP/08 for platforms where
1787 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1789 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1790 chunk. This allows us to accept broken chunked messages. We need a more
1791 general solution here.
1793 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1794 already-broken messages in the queue.
1796 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1798 JH/19 Fix reference counting bug in routing-generated-address tracking.
1804 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1805 supports it and a size is available (ie. the sending peer gave us one).
1807 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1808 years of being deprecated). The replacements are the ACLs
1809 acl_smtp_mime and acl_not_smtp_mime.
1811 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1812 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1813 or even an in-clear connection were permitted. Now, if the host lookup
1814 was dnssec and dane was requested then the host is only used if the
1815 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1816 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1817 if one fails this test.
1818 This means that a poorly-configured remote DNS will make it incommunicado;
1819 but it protects against a DNS-interception attack on it.
1821 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1822 non-noisy when a race steals the message being considered.
1824 JH/05 If main configuration option tls_certificate is unset, generate a
1825 self-signed certificate for inbound TLS connections.
1827 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1828 in rewrites and routers.
1830 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1831 and logged a warning sing 4.83; now they are a configuration file error.
1833 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1834 (lacking @domain). Apply the same qualification processing as RCPT.
1836 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1838 JH/10 Support ${sha256:} applied to a string (as well as the previous
1841 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1842 a cutthrough deliver is pending, as we always want to make a connection.
1843 This also avoids re-routing the message when later placing the cutthrough
1844 connection after a verify cache hit.
1845 Do not update it with the verify result either.
1847 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1848 when routing results in more than one destination address.
1850 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1851 signing (which inhibits the cutthrough capability). Previously only
1852 the presence of an option was tested; now an expansion evaluating as
1853 empty is permissible (obviously it should depend only on data available
1854 when the cutthrough connection is made).
1856 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1857 the relevant preceding SMTP command did not note the pipelining mode.
1859 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1860 Previously they were not counted.
1862 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1863 as one having no matching records. Previously we deferred the message
1864 that needed the lookup.
1866 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1867 distinguished as "(=".
1869 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1870 for missing MX records. Previously it only worked for missing A records.
1872 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1874 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1875 after the data-go-ahead and data-ack. Patch from Jason Betts.
1877 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1878 even for a "none" policy. Patch from Tony Meyer.
1880 JH/22 Fix continued use of a connection for further deliveries. If a port was
1881 specified by a router, it must also match for the delivery to be
1884 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1885 When one of the recipients of a message was unsuitable for the connection
1886 (has no matching addresses), we lost track of needing to mark it
1887 deferred. As a result mail would be lost.
1889 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1891 JH/25 Decoding ACL controls is now done using a binary search; the source code
1892 takes up less space and should be simpler to maintain. Merge the ACL
1893 condition decode tables also, with similar effect.
1895 JH/26 Fix problem with one_time used on a redirect router which returned the
1896 parent address unchanged. A retry would see the parent address marked as
1897 delivered, so not attempt the (identical) child. As a result mail would
1900 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1901 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1902 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1905 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1906 Main config options check_{log,spool}_{inodes,space} are now
1907 100 inodes, 10MB unless set otherwise in the configuration.
1909 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1910 Previously it only applied to the main-section connection policy
1913 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1915 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1916 by me. Added RFC7919 DH primes as an alternative.
1918 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1919 are not in the system include path.
1921 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1922 GnuTLS, when a session startup failed (eg because the client disconnected)
1923 Exim did stdio operations after fclose. This was exposed by a recent
1924 change which nulled out the file handle after the fclose.
1926 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1927 signed directly by the cert-signing cert, rather than an intermediate
1928 OCSP-signing cert. This is the model used by LetsEncrypt.
1930 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1932 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1933 an incoming connection.
1935 HS/02 Bug 1802: Do not half-close the connection after sending a request
1938 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1939 fallback to "prime256v1".
1941 JH/34 SECURITY: Use proper copy of DATA command in error message.
1942 Could leak key material. Remotely exploitable. CVE-2016-9963.
1948 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1949 and 3.4.4 - once the server is enabled to respond to an OCSP request
1950 it does even when not requested, resulting in a stapling non-aware
1951 client dropping the TLS connection.
1953 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1954 support variable-length bit vectors. No functional change.
1956 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1957 The I= interface field on outgoing lines is now after the H= remote
1958 host field, same as incoming lines. There is a separate
1959 outgoing_interface log selector which allows you to disable the
1962 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1963 If not running log_selector +smtp_connection the mainlog would be held
1964 open indefinitely after a "too many connections" event, including to a
1965 deleted file after a log rotate. Leave the per net connection logging
1966 leaving it open for efficiency as that will be quickly detected by the
1967 check on the next write.
1969 HS/01 Bug 1671: Fix post transport crash.
1970 Processing the wait-<transport> messages could crash the delivery
1971 process if the message IDs didn't exist for some reason. When
1972 using 'split_spool_directory=yes' the construction of the spool
1973 file name failed already, exposing the same netto behaviour.
1975 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1976 mime_regex ACL conditions.
1978 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1979 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1980 to HELO, local diagnostic string.
1982 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1983 log line to debug. Even when configured with a tls authenticator many
1984 client connections are expected to not authenticate in this way, so
1985 an authenticate fail is not an error.
1987 HS/02 Add the Exim version string to the process info. This way exiwhat
1988 gives some more detail about the running daemon.
1990 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1991 matter for fast-change records such as DNSBLs.
1993 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1994 retry record, even if constant. There may be multiple transports with
1995 different interface settings and the retry behaviour needs to be kept
1998 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2000 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2002 JH/10 Bug 840: fix log_defer_output option of pipe transport
2004 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2005 hosts_randomize. This matters a lot when combined with mua_wrapper.
2007 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2008 ${quote_pgsql:<string>} operator.
2010 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2012 JH/14 Change header file name for freeradius-client. Relevant if compiling
2013 with Radius support; from the Gentoo tree and checked under Fedora.
2015 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2017 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2018 option transport_filter, meaning no filtering.
2020 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2022 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2023 defaults to "*" (all hosts). The variable is now available when not built
2024 with TLS, default unset, mainly to enable keeping the testsuite sane.
2025 If a server certificate is not supplied (via tls_certificate) an error is
2026 logged, and clients will find TLS connections fail on startup. Presumably
2027 they will retry in-clear.
2028 Packagers of Exim are strongly encouraged to create a server certificate
2029 at installation time.
2031 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2032 with the $config_file variable.
2034 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2035 in transport context, after the attempt, and per-recipient. The latter type
2036 is per host attempted. The event data is the error message, and the errno
2037 information encodes the lookup type (A vs. MX) used for the (first) host,
2038 and the trailing two digits of the smtp 4xx response.
2040 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2041 to write to mainlog (or rejectlog, paniclog) in the window between file
2042 creation and permissions/ownership being changed. Particularly affects
2043 installations where exicyclog is run as root, rather than exim user;
2044 result is that the running daemon panics and dies.
2046 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2048 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2049 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2050 "pri" and "weight". Note that the previous implicit priority given by the
2051 list order is no longer honoured.
2053 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2054 for DKIM processing.
2056 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2057 by defining SUPPORT_SOCKS.
2059 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2060 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2061 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2062 variables are renamed to proxy_{local,external}_{address,port}.
2064 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2065 for a build by defining SUPPORT_I18N
2067 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2068 of the query string, and make ${quote_redis:} do that quoting.
2070 JH/29 Move Events support from Experimental to mainline, enabled by default
2071 and removable for a build by defining DISABLE_EVENT.
2073 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2075 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2076 cached by the daemon.
2078 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2079 by defining LOOKUP_REDIS. The libhiredis library is required.
2081 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2082 keys are given for lookup.
2084 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2085 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2086 only supported when built with TLS support. The PolarSSL SHA routines
2087 are still used when the TLS library is too old for convenient support.
2089 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2090 openssl_options), for security. OpenSSL forces this from version 1.1.0
2091 server-side so match that on older versions.
2093 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2094 allocation for $value could be released as the expansion processing
2095 concluded, but leaving the global pointer active for it.
2097 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2098 and to use the domains and local_parts ACL conditions.
2100 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2101 incorrectly not doubled on cutthrough transmission, hence seen as a
2102 body-termination at the receiving system - resulting in truncated mails.
2103 Commonly the sender saw a TCP-level error, and retransmitted the message
2104 via the normal store-and-forward channel. This could result in duplicates
2105 received - but deduplicating mailstores were liable to retain only the
2106 initial truncated version.
2108 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2110 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2112 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2113 we're in there, support oversigning also; bug 1309.
2115 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2117 HS/04 Add support for keep_environment and add_environment options.
2119 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2120 either intentional arithmetic overflow during PRNG, or testing config-
2123 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2124 delivery resulted in actual delivery. Cancel cutthrough before DATA
2127 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2128 rejects a recipient: pass the reject to the originator.
2130 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2131 Many were false-positives and ignorable, but it's worth fixing the
2134 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2135 for the new environment-manipulation done at startup. Move the routines
2136 from being local to tls.c to being global via the os.c file.
2138 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2139 an extract embedded as result-arg for a map, the first arg for extract
2140 is unavailable so we cannot tell if this is a numbered or keyed
2141 extraction. Accept either.
2147 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2150 JH/02 The smtp transport option "multi_domain" is now expanded.
2152 JH/03 The smtp transport now requests PRDR by default, if the server offers
2155 JH/04 Certificate name checking on server certificates, when exim is a client,
2156 is now done by default. The transport option tls_verify_cert_hostnames
2157 can be used to disable this per-host. The build option
2158 EXPERIMENTAL_CERTNAMES is withdrawn.
2160 JH/05 The value of the tls_verify_certificates smtp transport and main options
2161 default to the word "system" to access the system default CA bundle.
2162 For GnuTLS, only version 3.0.20 or later.
2164 JH/06 Verification of the server certificate for a TLS connection is now tried
2165 (but not required) by default. The verification status is now logged by
2166 default, for both outbound TLS and client-certificate supplying inbound
2169 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2172 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2173 Status Notification (bounce) messages are now MIME format per RFC 3464.
2174 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2175 under the control of the dsn_advertise_hosts option, and routers may
2176 have a dsn_lasthop option.
2178 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2179 default, modifiable by a malware= option. The list separator for
2180 the options can now be changed in the usual way. Bug 68.
2182 JH/10 The smtp_receive_timeout main option is now expanded before use.
2184 JH/11 The incoming_interface log option now also enables logging of the
2185 local interface on delivery outgoing connections.
2187 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2188 if the interface and destination host and port all match.
2190 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2193 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2194 Patch from Andrew Lewis.
2196 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2197 now supports optional time-restrictions, weighting, and priority
2198 modifiers per server. Patch originally by <rommer@active.by>.
2200 JH/16 The spamd_address main option now supports a mixed list of local
2201 and remote servers. Remote servers can be IPv6 addresses, and
2202 specify a port-range.
2204 JH/17 Bug 68: The spamd_address main option now supports an optional
2205 timeout value per server.
2207 JH/18 Bug 1581: Router and transport options headers_add/remove can
2208 now have the list separator specified.
2210 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2213 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2216 JH/21 Support for the A6 type of dns record is withdrawn.
2218 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2219 rather than the verbs used.
2221 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2222 from 255 to 1024 chars.
2224 JH/24 Verification callouts now attempt to use TLS by default.
2226 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2227 are generic router options now. The defaults didn't change.
2229 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2230 Original patch from Alexander Shikoff, worked over by JH.
2232 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2233 files if ZCAT_COMMAND is not executable.
2235 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2237 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2239 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2240 Normally benign, it bites when the pair was led to by a CNAME;
2241 modern usage is to not canonicalize the domain to a CNAME target
2242 (and we were inconsistent anyway for A-only vs AAAA+A).
2244 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2246 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2247 when evaluating $sender_host_dnssec.
2249 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2250 $sender_helo_dnssec variable.
2252 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2254 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2256 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2258 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2259 documented as working, but never had. Support all but $spam_report.
2261 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2262 added for tls authenticator.
2264 HS/03 Add perl_taintmode main config option
2270 TL/01 When running the test suite, the README says that variables such as
2271 no_msglog_check are global and can be placed anywhere in a specific
2272 test's script, however it was observed that placement needed to be near
2273 the beginning for it to behave that way. Changed the runtest perl
2274 script to read through the entire script once to detect and set these
2275 variables, reset to the beginning of the script, and then run through
2276 the script parsing/test process like normal.
2278 TL/02 The BSD's have an arc4random API. One of the functions to induce
2279 adding randomness was arc4random_stir(), but it has been removed in
2280 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2281 function when detected.
2283 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2284 cause callback expansion.
2286 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2287 syntax errors in an expansion can be treated as a string instead of
2288 logging or causing an error, due to the internal use of bool_lax
2289 instead of bool when processing it.
2291 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2292 server certificates when making smtp deliveries.
2294 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2296 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2298 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2300 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2301 Merged patch from Sebastian Wiedenroth.
2303 JH/05 Fix results-pipe from transport process. Several recipients, combined
2304 with certificate use, exposed issues where response data items split
2305 over buffer boundaries were not parsed properly. This eventually
2306 resulted in duplicates being sent. This issue only became common enough
2307 to notice due to the introduction of connection certificate information,
2308 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2310 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2311 size buffer was used, resulting in syntax errors when an expansion
2314 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2315 version 3.3.6 or later.
2317 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2318 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2319 both become "event_action", the variables become $event_name, $event_data
2320 and $event_defer_errno. There is a new variable $verify_mode, usable in
2321 routers, transports and related events. The tls:cert event is now also
2322 raised for inbound connections, if the main configuration event_action
2325 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2326 early OCSP support, but no stapling (appears to be less than 1.0.0).
2328 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2329 server certificate names available under the smtp transport option
2330 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2333 JH/10 Time-related extraction expansions from certificates now use the main
2334 option "timezone" setting for output formatting, and are consistent
2335 between OpenSSL and GnuTLS compilations. Bug 1541.
2337 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2338 encoded parameter in the incoming message. Bug 1558.
2340 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2341 include certificate info, eximon was claiming there were spoolfile
2344 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2346 JH/14 Log delivery-related information more consistently, using the sequence
2347 "H=<name> [<ip>]" wherever possible.
2349 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2350 are problematic for Debian distribution, omit them from the release
2353 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2355 JH/16 Fix string representation of time values on 64bit time_t architectures.
2358 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2359 output list separator was used.
2364 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2365 checkers that were complaining about end of non-void function with no
2368 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2369 This was a regression introduced in 4.83 by another bugfix.
2371 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2373 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2374 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2380 TF/01 Correctly close the server side of TLS when forking for delivery.
2382 When a message was received over SMTP with TLS, Exim failed to clear up
2383 the incoming connection properly after forking off the child process to
2384 deliver the message. In some situations the subsequent outgoing
2385 delivery connection happened to have the same fd number as the incoming
2386 connection previously had. Exim would try to use TLS and fail, logging
2387 a "Bad file descriptor" error.
2389 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2390 utilities have not been installed.
2392 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2393 temporary space as the ACL may create new global variables.
2395 TL/01 LDAP support uses per connection or global context settings, depending
2396 upon the detected version of the libraries at build time.
2398 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2399 to extract and use the src ip:port in logging and expansions as if it
2400 were a direct connection from the outside internet. PPv2 support was
2401 updated based on HAProxy spec change in May 2014.
2403 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2405 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2406 Properly escape header and check for NULL return.
2408 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2411 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2413 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2414 characters in header names, implemented as a verify condition.
2415 Contributed by Michael Fischer v. Mollard.
2417 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2418 results permerror and temperror. Previous values are deprecated but
2419 still accepted. In a future release, err_perm and err_temp will be
2420 completely removed, which will be a backward incompatibility if the
2421 ACL tests for either of these two old results. Patch contributed by
2422 user bes-internal on the mailing list.
2424 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2426 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2427 selectors, in both main and reject logs.
2429 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2432 JH/07 Add malware type "sock" for talking to simple daemon.
2434 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2436 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2437 routers/transports under cutthrough routing.
2439 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2440 numbers. Touch up "bool" conditional to keep the same definition.
2442 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2444 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2446 JH/12 Expand items in router/transport headers_add or headers_remove lists
2447 individually rather than the list as a whole. Bug 1452.
2449 Required for reasonable handling of multiple headers_ options when
2450 they may be empty; requires that headers_remove items with embedded
2451 colons must have them doubled (or the list-separator changed).
2453 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2454 view the policy declared in the DMARC record. Currently, $dmarc_status
2455 is a combined value of both the record presence and the result of the
2458 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2460 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2461 dnslookup router and the smtp transport (applying to the forward
2464 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2465 of ldap servers used for a specific lookup. Patch provided by Heiko
2468 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2469 New variable $lookup_dnssec_authenticated for observability.
2471 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2472 Patch submitted by Lars Timman.
2474 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2476 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2477 Requires trusted mode and valid format message id, aborts otherwise.
2478 Patch contributed by Heiko Schlichting.
2480 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2481 certextract with support for various fields. Bug 1358.
2483 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2484 is requested by default, modifiable by smtp transport option
2487 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2488 operate on certificate variables to give certificate fingerprints
2489 Also new ${sha256:cert_variable}.
2491 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2493 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2494 Christian Aistleitner.
2496 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2498 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2499 file. Patch from Wolfgang Breyha.
2501 JH/25 Expand the coverage of the delivery $host and $host_address to
2502 client authenticators run in verify callout. Bug 1476.
2504 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2505 align with daemon_smtp_ports. Bug 72.
2507 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2508 support and error reporting did not work properly.
2510 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2511 and is readable. Patch from Andrew Colin Kissa.
2513 TL/14 Enhance documentation of ${run expansion and how it parses the
2514 commandline after expansion, particularly in the case when an
2515 unquoted variable expansion results in an empty value.
2517 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2519 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2520 by renaming away from it. Addresses GH issue 15, reported by
2523 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2525 TL/15 SECURITY: prevent double expansion in math comparison functions
2526 (can expand unsanitized data). Not remotely exploitable.
2533 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2535 PP/02 Make -n do something, by making it not do something.
2536 When combined with -bP, the name of an option is not output.
2538 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2541 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2542 $sender_host_name and config options to manage this, and basic check
2545 PP/05 DSCP support for outbound connections and control modifier for inbound.
2547 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2548 (Only plugin which currently uses this is kerberos4, which nobody should
2549 be using, but we should make it available and other future plugins might
2550 conceivably use it, even though it would break NAT; stuff *should* be
2551 using channel bindings instead).
2553 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2554 name; added for Sendmail compatibility; requires admin caller.
2555 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2556 just ignore it); requires trusted caller.
2557 Also parse but ignore: -Ac -Am -X<logfile>
2560 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2562 TL/02 Add +smtp_confirmation as a default logging option.
2564 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2565 Patch by Magnus Holmgren from 2007-02-20.
2567 TL/04 Bugzilla 1281 - Spec typo.
2568 Bugzilla 1283 - Spec typo.
2569 Bugzilla 1290 - Spec grammar fixes.
2571 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2573 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2575 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2576 mailing list by by Dmitry Isaikin.
2578 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2580 JH/02 Support "G" suffix to numbers in ${if comparisons.
2582 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2584 NM/01 Bugzilla 1197 - Spec typo
2585 Bugzilla 1196 - Spec examples corrections
2587 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2589 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2590 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2593 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2594 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2596 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2597 "acl {{name}{arg}...}", and optional args on acl condition
2600 JH/05 Permit multiple router/transport headers_add/remove lines.
2602 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2604 JH/07 Avoid using a waiting database for a single-message-only transport.
2605 Performance patch from Paul Fisher. Bugzilla 1262.
2607 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2610 JH/09 Add $headers_added variable, with content from use of ACL modifier
2611 add_header (but not yet added to the message). Bugzilla 199.
2613 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2614 Pulled from Bugzilla 817 by Wolfgang Breyha.
2616 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2618 (nb: this is the same fix as in Exim 4.80.1)
2620 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2623 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2626 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2628 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2629 advertise SMTP AUTH mechanism to us, instead of a generic
2630 protocol violation error. Also, make Exim more robust to bad
2631 data from the Dovecot auth socket.
2633 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2635 When a queue runner is handling a message, Exim first routes the
2636 recipient addresses, during which it prunes them based on the retry
2637 hints database. After that it attempts to deliver the message to
2638 any remaining recipients. It then updates the hints database using
2641 So if a recipient address works intermittently, it can get repeatedly
2642 deferred at routing time. The retry hints record remains fresh so the
2643 address never reaches the final cutoff time.
2645 This is a fairly common occurrence when a user is bumping up against
2646 their storage quota. Exim had some logic in its local delivery code
2647 to deal with this. However it did not apply to per-recipient defers
2648 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2650 This change adds a proper retry rule check during routing so that the
2651 final cutoff time is checked against the message's age. We only do
2652 this check if there is an address retry record and there is not a
2653 domain retry record; this implies that previous attempts to handle
2654 the address had the retry_use_local_parts option turned on. We use
2655 this as an approximation for the destination being like a local
2656 delivery, as in LMTP.
2658 I suspect this new check makes the old local delivery cutoff check
2659 redundant, but I have not verified this so I left the code in place.
2661 TF/02 Correct gecos expansion when From: is a prefix of the username.
2663 Test 0254 submits a message to Exim with the header
2667 When I ran the test suite under the user fanf2, Exim expanded
2668 the header to contain my full name, whereas it should have added
2669 a Resent-Sender: header. It erroneously treats any prefix of the
2670 username as equal to the username.
2672 This change corrects that bug.
2674 GF/01 DCC debug and logging tidyup
2675 Error conditions log to paniclog rather than rejectlog.
2676 Debug lines prefixed by "DCC: " to remove any ambiguity.
2678 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2680 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2681 Bug spotted by Jeremy Harris; was flawed since initial commit.
2682 Would have resulted in OCSP responses post-SNI triggering an Exim
2683 NULL dereference and crash.
2685 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2687 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2688 Bug detection, analysis and fix by Samuel Thibault.
2689 Bugzilla 1331, Debian bug #698092.
2691 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2693 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2694 Server implementation by Todd Lyons, client by JH.
2695 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2696 config variable "prdr_enable" controls whether the server
2697 advertises the facility. If the client requests PRDR a new
2698 acl_data_smtp_prdr ACL is called once for each recipient, after
2699 the body content is received and before the acl_smtp_data ACL.
2700 The client is controlled by both of: a hosts_try_prdr option
2701 on the smtp transport, and the server advertisement.
2702 Default client logging of deliveries and rejections involving
2703 PRDR are flagged with the string "PRDR".
2705 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2706 fclose(). Diagnosis by Todd Lyons.
2708 PP/17 Update configure.default to handle IPv6 localhost better.
2709 Patch by Alain Williams (plus minor tweaks).
2712 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2713 This is now consistent with GnuTLS, and is now documented: the
2714 previous undocumented portable approach to treating the option as
2715 unset was to force an expansion failure. That still works, and
2716 an empty string is now equivalent.
2718 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2719 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2720 not performing validation itself.
2722 PP/20 Added force_command boolean option to pipe transport.
2723 Patch from Nick Koston, of cPanel Inc.
2725 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2728 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2730 PP/21 Fix eximon continuous updating with timestamped log-files.
2731 Broken in a format-string cleanup in 4.80, missed when I repaired the
2732 other false fix of the same issue.
2733 Report and fix from Heiko Schlichting.
2736 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2737 Report from Prashanth Katuri.
2739 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2740 It's SecureTransport, so affects any MacOS clients which use the
2741 system-integrated TLS libraries, including email clients.
2743 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2744 using a MIME ACL for non-SMTP local injection.
2745 Report and assistance in diagnosis by Warren Baker.
2747 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2749 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2751 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2752 last id that failed so it may be referenced in subsequent ACL's.
2754 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2757 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2758 ldap library initialization, allowing self-signed CA's to be
2759 used. Also properly sets require_cert option later in code by
2760 using NULL (global ldap config) instead of ldap handle (per
2761 session). Bug diagnosis and testing by alxgomz.
2763 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2764 the src/util/ subdirectory.
2766 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2767 renamed to Transport Post Delivery Action by Jeremy Harris, as
2770 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2771 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2772 redis_servers = needs to be configured which will be used by the redis
2773 lookup. Patch from Warren Baker, of The Packet Hub.
2775 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2777 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2778 hostname or reverse DNS when processing a host list. Used suggestions
2779 from multiple comments on this bug.
2781 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2783 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2784 Missed a few lines, added it to make the runtest require no keyboard
2787 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2788 contains upper case chars. Make router use caseful_local_part.
2790 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2791 support when GnuTLS has been built with p11-kit.
2797 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2799 This, or similar/improved, will also be change PP/11 of 4.82.
2805 PP/01 Handle short writes when writing local log-files.
2806 In practice, only affects FreeBSD (8 onwards).
2807 Bugzilla 1053, with thanks to Dmitry Isaikin.
2809 NM/01 Bugzilla 949 - Documentation tweak
2811 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2814 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2816 PP/02 Implemented gsasl authenticator.
2818 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2820 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2821 `pkg-config foo` for cflags/libs.
2823 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2824 with rest of GSASL and with heimdal_gssapi.
2826 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2827 `pkg-config foo` for cflags/libs for the TLS implementation.
2829 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2830 properties get this fed in as external SSF. A number of robustness
2831 and debugging improvements to the cyrus_sasl authenticator.
2833 PP/08 cyrus_sasl server now expands the server_realm option.
2835 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2836 Patch by Jeremy Harris.
2838 PP/10 Added dbmjz lookup type.
2840 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2842 PP/12 MAIL args handles TAB as well as SP, for better interop with
2843 non-compliant senders.
2844 Analysis and variant patch by Todd Lyons.
2846 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2847 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2848 Patch from Dirk Mueller <dmueller@suse.com>
2850 PP/13 tls_peerdn now print-escaped for spool files.
2851 Observed some $tls_peerdn in wild which contained \n, which resulted
2852 in spool file corruption.
2854 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2855 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2856 or write after TLS renegotiation, which otherwise led to messages
2859 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2860 as a tracking header (ie: a signed header comes before the signature).
2861 Patch from Wolfgang Breyha.
2863 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2864 comma-sep list; embedded commas doubled.
2866 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2868 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2870 Report and patch from Dmitry Banschikov.
2872 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2873 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2874 failures. We appear to now support TLS1.1+ with Exim.
2876 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2877 lets Exim select keys and certificates based upon TLS SNI from client.
2878 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2879 before an outbound SMTP session. New log_selector, +tls_sni.
2881 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2882 NULL dereference. Report and patch from Alun Jones.
2884 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2885 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2886 Not seeing resolver debug output on NetBSD, but suspect this is a
2887 resolver implementation change.
2889 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2890 Left warnings. Added "eximon gdb" invocation mode.
2892 PP/21 Defaulting "accept_8bitmime" to true, not false.
2894 PP/22 Added -bw for inetd wait mode support.
2896 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2897 locate the relevant includes and libraries. Made this the default.
2899 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2900 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2902 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2903 This may cause build issues on older platforms.
2905 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2906 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2907 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2908 Added SNI support via GnuTLS too.
2909 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2911 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2913 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2914 Applied second patch from Janne, implementing suggestion to default
2915 multiple-strings-in-record handling to match SPF spec.
2917 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2919 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2920 read-only, out of scope).
2921 Patch from Wolfgang Breyha, report from Stuart Northfield.
2923 PP/29 Fix three issues highlighted by clang analyser static analysis.
2924 Only crash-plausible issue would require the Cambridge-specific
2925 iplookup router and a misconfiguration.
2926 Report from Marcin Mirosław.
2928 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2930 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2931 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2932 As part of this, removing so much warning spew let me fix some minor
2933 real issues in debug logging.
2935 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2936 assignment on my part. Fixed.
2938 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2939 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2940 Janne Snabb (who went above and beyond: thank you).
2942 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2943 string otherwise requires a connection and a bunch more work and it's
2944 relatively easy to get wrong. Should also expose TLS library linkage
2947 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2948 64-bit ${eval} (JH/03).
2950 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2951 GNU libc to support some of the 64-bit stuff, should not lead to
2952 conflicts. Defined before os.h is pulled in, so if a given platform
2953 needs to override this, it can.
2955 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2956 protection layer was required, which is not implemented.
2957 Bugzilla 1254, patch from Wolfgang Breyha.
2959 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2960 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2961 tls_dhparam take prime identifiers. Also unbreak combination of
2962 OpenSSL+DH_params+TLSSNI.
2964 PP/39 Disable SSLv2 by default in OpenSSL support.
2970 PP/01 Solaris build fix for Oracle's LDAP libraries.
2971 Bugzilla 1109, patch from Stephen Usher.
2973 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2975 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2978 TF/02 Fix a couple more cases where we did not log the error message
2979 when unlink() failed. See also change 4.74-TF/03.
2981 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2982 lock up or crash if it happened to be inside a call to libc when it
2983 got a SIGUSR1 from exiwhat.
2985 The SIGUSR1 handler appends the current process status to the process
2986 log which is later printed by exiwhat. It used to use the general
2987 purpose logging code to do this, but several functions it calls are
2988 not safe for signals.
2990 The new output code in the SIGUSR1 handler is specific to the process
2991 log, and simple enough that it's easy to inspect for signal safety.
2992 Removing some special cases also simplifies the general logging code.
2993 Removing the spurious timestamps from the process log simplifies
2996 TF/04 Improved ratelimit ACL condition.
2998 The /noupdate option has been deprecated in favour of /readonly which
2999 has clearer semantics. The /leaky, /strict, and /readonly update modes
3000 are mutually exclusive. The update mode is no longer included in the
3001 database key; it just determines when the database is updated. (This
3002 means that when you upgrade Exim will forget old rate measurements.)
3004 Exim now checks that the per_* options are used with an update mode that
3005 makes sense for the current ACL. For example, when Exim is processing a
3006 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3007 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3008 must specify per_mail/readonly. If you omit the update mode it defaults to
3009 /leaky where that makes sense (as before) or /readonly where required.
3011 The /noupdate option is now undocumented but still supported for
3012 backwards compatibility. It is equivalent to /readonly except that in
3013 ACLs where /readonly is required you may specify /leaky/noupdate or
3014 /strict/noupdate which are treated the same as /readonly.
3016 A useful new feature is the /count= option. This is a generalization
3017 of the per_byte option, so that you can measure the throughput of other
3018 aggregate values. For example, the per_byte option is now equivalent
3019 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3021 The per_rcpt option has been generalized using the /count= mechanism
3022 (though it's more complicated than the per_byte equivalence). When it is
3023 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3024 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3025 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3026 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3027 Note that using per_rcpt with a non-readonly update mode in more than
3028 one ACL will cause the recipients to be double-counted. (The per_mail
3029 and per_byte options don't have this problem.)
3031 The handling of very low rates has changed slightly. If the computed rate
3032 is less than the event's count (usually one) then this event is the first
3033 after a long gap. In this case the rate is set to the same as this event's
3034 count, so that the first message of a spam run is counted properly.
3036 The major new feature is a mechanism for counting the rate of unique
3037 events. The new per_addr option counts the number of different
3038 recipients that someone has sent messages to in the last time period. It
3039 behaves like per_rcpt if all the recipient addresses are different, but
3040 duplicate recipient addresses do not increase the measured rate. Like
3041 the /count= option this is a general mechanism, so the per_addr option
3042 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3043 example, measure the rate that a client uses different sender addresses
3044 with the options per_mail/unique=$sender_address. There are further
3045 details in the main documentation.
3047 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3049 TF/06 Removed a few PCRE remnants.
3051 TF/07 Automatically extract Exim's version number from tags in the git
3052 repository when doing development or release builds.
3054 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3055 Bugzilla 879. Patch from Paul Fisher.
3057 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3058 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3061 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3063 PP/05 Variable $av_failed, true if the AV scanner deferred.
3064 Bugzilla 1078. Patch from John Horne.
3066 PP/06 Stop make process more reliably on build failure.
3067 Bugzilla 1087. Patch from Heiko Schlittermann.
3069 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3070 Bugzilla 1089. Patch from Heiko Schlittermann.
3072 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3073 Bugzilla 1131. Patch from Holger Weiß.
3075 PP/09 Handle IPv6 addresses with SPF.
3076 Bugzilla 860. Patch from Wolfgang Breyha.
3078 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3080 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3083 PP/11 match_* no longer expand right-hand-side by default.
3084 New compile-time build option, EXPAND_LISTMATCH_RHS.
3085 New expansion conditions, "inlist", "inlisti".
3087 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3089 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3091 PP/14 fix log_write() format string regression from TF/03.
3092 Bugzilla 1152. Patch from Dmitry Isaikin.
3098 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3100 PP/02 Harmonised TLS library version reporting; only show if debugging.
3101 Layout now matches that introduced for other libraries in 4.74 PP/03.
3103 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3105 PP/04 New "dns_use_edns0" global option.
3107 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3110 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3111 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3113 TK/01 Updated PolarSSL code to 0.14.2.
3114 Bugzilla 1097. Patch from Andreas Metzler.
3116 PP/07 Catch divide-by-zero in ${eval:...}.
3117 Fixes bugzilla 1102.
3119 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3122 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3123 format-string attack -- SECURITY: remote arbitrary code execution.
3125 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3126 time unintentionally subject to list matching rules, letting the header
3127 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3128 arbitrary string expansion). This allowed for information disclosure.
3130 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3131 INT_MIN/-1 -- value coerced to INT_MAX.
3137 NM/01 Workaround for PCRE version dependency in version reporting
3140 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3141 This fixes portability to compilers other than gcc, notably
3142 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3144 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3145 makefiles for portability to HP-UX and POSIX correctness.
3147 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3148 Also via indented variable definition in the Makefile.
3149 (Debugging by Oliver Heesakkers).
3151 PP/02 Restore caching of spamd results with expanded spamd_address.
3152 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3154 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3155 Improves build reliability. Fix from: Frank Elsner
3157 NM/02 Fix wide character breakage in the rfc2047 coding
3158 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3160 NM/03 Allow underscore in dnslist lookups
3161 Fixes bug 1026. Patch from Graeme Fowler
3163 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3164 Code patches from Adam Ciarcinski of NetBSD.
3166 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3169 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3170 is logged, to avoid truncation. Patch from John Horne.
3172 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3173 Patch from Jakob Hirsch.
3175 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3176 SQL string expansion failure details.
3177 Patch from Andrey Oktyabrski.
3179 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3180 Patch from Simon Arlott.
3182 PP/09 New lookups functionality failed to compile on old gcc which rejects
3183 extern declarations in function scope.
3184 Patch from Oliver Fleischmann
3186 PP/10 Use sig_atomic_t for flags set from signal handlers.
3187 Check getgroups() return and improve debugging.
3188 Fixed developed for diagnosis in bug 927 (which turned out to be
3191 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3192 Patch from Mark Zealey.
3194 PP/12 Bugzilla 1056: Improved spamd server selection.
3195 Patch from Mark Zealey.
3197 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3198 Based on patch from Heiko Schlittermann.
3200 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3201 Patch from Uwe Doering, sign-off by Michael Haardt.
3203 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3204 variable declaration deep within a block. Bug and patch from
3207 PP/15 lookups-Makefile IRIX compatibility coercion.
3209 PP/16 Make DISABLE_DKIM build knob functional.
3211 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3212 Patch by Simon Arlott
3214 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3215 variable argument macros. Our copy now differs from upstream.
3221 TF/01 Failure to get a lock on a hints database can have serious
3222 consequences so log it to the panic log.
3224 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3225 controlled using the smtp_confirmation log selector.
3227 TF/03 Include the error message when we fail to unlink a spool file.
3229 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3230 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3231 for maintaining out-of-tree patches for some time.
3233 PP/01 Bugzilla 139: Documentation and portability issues.
3234 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3235 Handle per-OS dynamic-module compilation flags.
3237 PP/02 Let /dev/null have normal permissions.
3238 The 4.73 fixes were a little too stringent and complained about the
3239 permissions on /dev/null. Exempt it from some checks.
3240 Reported by Andreas M. Kirchwitz.
3242 PP/03 Report version information for many libraries, including
3243 Exim version information for dynamically loaded libraries. Created
3244 version.h, now support a version extension string for distributors
3245 who patch heavily. Dynamic module ABI change.
3247 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3248 privilege escalation vulnerability whereby the Exim run-time user
3249 can cause root to append content of the attacker's choosing to
3252 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3255 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3256 If dropping privileges for untrusted macros, we disabled normal logging
3257 on the basis that it would fail; for the Exim run-time user, this is not
3258 the case, and it resulted in successful deliveries going unlogged.
3259 Fixed. Reported by Andreas Metzler.
3265 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3266 only prepend for the Resent-* case. Fixes regression introduced in
3267 Exim 4.70 by NM/22 for Bugzilla 607.
3269 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3270 increasing numbers of administrators be bitten by this.
3272 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3274 PP/03 Bugzilla 994: added openssl_options main configuration option.
3276 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3278 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3280 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3282 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3283 without a peer certificate, leading to a segfault because of an
3284 assumption that peers always have certificates. Be a little more
3285 paranoid. Problem reported by Martin Tscholak.
3287 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3288 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3289 NB: ClamAV planning to remove STREAM in "middle of 2010".
3290 CL also introduces -bmalware, various -d+acl logging additions and
3291 more caution in buffer sizes.
3293 PP/09 Implemented reverse_ip expansion operator.
3295 PP/10 Bugzilla 937: provide a "debug" ACL control.
3297 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3299 PP/12 Bugzilla 973: Implement --version.
3301 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3303 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3305 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3307 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3308 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3309 ignore trailing whitespace.
3311 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3313 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3316 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3317 Notification from Dr Andrew Aitchison.
3319 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3320 ExtendedDetectionInfo response format.
3321 Notification from John Horne.
3323 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3326 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3327 XSL and documented dependency on system catalogs, with examples of how
3330 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3333 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3334 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3335 configuration file which is writeable by the Exim user or group.
3337 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3338 of configuration files to cover files specified with the -C option if
3339 they are going to be used with root privileges, not just the default
3342 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3343 option (effectively making it always true).
3345 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3346 files to be used while preserving root privileges.
3348 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3349 that rogue child processes cannot use them.
3351 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3352 run-time user, instead of root.
3354 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3355 Exim run-time user without dropping privileges.
3357 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3358 result string, instead of calling string_vformat() twice with the same
3361 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3362 for other users. Others should always drop root privileges if they use
3363 -C on the command line, even for a whitelisted configure file.
3365 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3367 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3373 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3374 $data_path, and $header_path variables; fixed documentation bugs and
3377 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3378 exipick to access non-standard spools, including the "frozen" queue
3381 NM/01 Bugzilla 965: Support mysql stored procedures.
3382 Patch from Alain Williams
3384 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3386 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3387 Patch from Andreas Metzler
3389 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3390 Patch from Kirill Miazine
3392 NM/05 Bugzilla 671: Added umask to procmail example.
3394 JJ/03 installed exipick 20100323.0, fixing doc bug
3396 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3397 directory. Notification and patch from Dan Rosenberg.
3399 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3401 TK/02 Improve log output when DKIM signing operation fails.
3403 MH/01 Treat the transport option dkim_domain as a colon separated
3404 list, not as a single string, and sign the message with each element,
3405 omitting multiple occurences of the same signer.
3407 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3408 Bugzilla 985, 986. Patch by Simon Arlott
3410 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3411 Patch by Simon Arlott
3413 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3414 MBX locking. Notification from Dan Rosenberg.
3420 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3422 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3424 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3426 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3428 NM/04 Bugzilla 915: Flag broken perl installation during build.
3434 TK/01 Added patch by Johannes Berg that expands the main option
3435 "spamd_address" if it starts with a dollar sign.
3437 TK/02 Write list of recipients to X-Envelope-Sender header when building
3438 the mbox-format spool file for content scanning (suggested by Jakob
3441 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3442 (http://www.dcc-servers.net/) support via dccifd. Activated by
3443 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3445 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3446 by Mark Daniel Reidel <mr@df.eu>.
3448 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3449 When building exim an external PCRE library is now needed -
3450 PCRE is a system library on the majority of modern systems.
3451 See entry on PCRE_LIBS in EDITME file.
3453 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3454 conversation. Added nologin parameter to request.
3455 Patch contributed by Kirill Miazine.
3457 TF/01 Do not log submission mode rewrites if they do not change the address.
3459 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3461 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3462 log files in place. Contributed by Roberto Lima.
3464 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3466 TF/03 Bugzilla 615: When checking the local_parts router precondition
3467 after a local_part_suffix or local_part_prefix option, Exim now
3468 does not use the address's named list lookup cache, since this
3469 contains cached lookups for the whole local part.
3471 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3472 Robert Millan. Documentation is in experimental-spec.txt.
3474 TF/04 Bugzilla 668: Fix parallel build (make -j).
3476 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3478 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3479 Patch provided by Jan Srzednicki.
3481 TF/05 Leading white space used to be stripped from $spam_report which
3482 wrecked the formatting. Now it is preserved.
3484 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3485 that they are available at delivery time.
3487 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3489 TF/08 TLS error reporting now respects the incoming_interface and
3490 incoming_port log selectors.
3492 TF/09 Produce a more useful error message if an SMTP transport's hosts
3493 setting expands to an empty string.
3495 NM/06 Bugzilla 744: EXPN did not work under TLS.
3496 Patch provided by Phil Pennock.
3498 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3499 Patch provided by Richard Godbee.
3501 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3502 acl_smtp_notquit, added index entry.
3504 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3505 Patch provided by Eugene Bujak.
3507 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3508 accept(). Patch provided by Maxim Dounin.
3510 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3511 Patch provided by Phil Pennock.
3513 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3515 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3516 Patch provided by Brad "anomie" Jorsch.
3518 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3519 Patch provided by Dean Brooks.
3521 TK/05 Add native DKIM support (does not depend on external libraries).
3523 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3524 Patch provided by Graeme Fowler.
3526 NM/16 Bugzilla 851: Documentation example syntax fix.
3528 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3530 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3533 NM/19 Bugzilla 745: TLS version reporting.
3534 Patch provided by Phil Pennock.
3536 NM/20 Bugzilla 167: bool: condition support.
3537 Patch provided by Phil Pennock.
3539 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3540 clients. Patch provided by Phil Pennock.
3542 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3543 Patch provided by Brad "anomie" Jorsch.
3545 NM/23 Bugzilla 687: Fix misparses in eximstats.
3546 Patch provided by Heiko Schlittermann.
3548 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3549 Patch provided by Heiko Schlittermann.
3551 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3552 plus update to original patch.
3554 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3556 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3557 Patch provided by David Brownlee.
3559 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3561 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3563 NM/30 Bugzilla 888: TLS documentation bugfixes.
3565 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3567 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3568 Unlike the original bugzilla I have changed all shell scripts in src tree.
3570 NM/33 Bugzilla 898: Transport filter timeout fix.
3571 Patch by Todd Rinaldo.
3573 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3574 Patch by Serge Demonchaux.
3576 NM/35 Bugzilla 39: Base64 decode bug fixes.
3577 Patch by Jakob Hirsch.
3579 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3581 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3583 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3585 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3591 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3592 ALT-N's libdkim that I have put here:
3593 http://duncanthrax.net/exim-experimental/
3595 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3596 were called 'true' and it seems that C99 defines that as a reserved
3597 keyword to be used with 'bool' variable types. That means you could
3598 not include C99-style headers which use bools without triggering
3599 build errors in sieve.c.
3601 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3602 as mailq or other aliases. Changed the --help handling significantly
3603 to do whats expected. exim_usage() emits usage/help information.
3605 SC/01 Added the -bylocaldomain option to eximstats.
3607 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3609 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3611 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3617 PH/01 Another patch from the Sieve maintainer.
3619 PH/02 When an IPv6 address is converted to a string for single-key lookup
3620 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3621 dots are used instead of colons so that keys in lsearch files need not
3622 contain colons. This was done some time before quoting was made available
3623 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3624 (notated using the quote facility) so as to distinguish them from IPv4
3625 keys. This meant that lookups for IP addresses in host lists did not work
3626 for iplsearch lookups.
3628 This has been fixed by arranging for IPv6 addresses to be expressed with
3629 colons if the lookup type is iplsearch. This is not incompatible, because
3630 previously such lookups could never work.
3632 The situation is now rather anomalous, since one *can* have colons in
3633 ordinary lsearch keys. However, making the change in all cases is
3634 incompatible and would probably break a number of configurations.
3636 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3639 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3640 conversion specification without a maximum field width, thereby enabling
3641 a rogue spamd server to cause a buffer overflow. While nobody in their
3642 right mind would setup Exim to query an untrusted spamd server, an
3643 attacker that gains access to a server running spamd could potentially
3644 exploit this vulnerability to run arbitrary code as the Exim user.
3646 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3647 $primary_hostname instead of what libspf2 thinks the hosts name is.
3649 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3650 a directory entry by the name of the lookup key. Previously, if a
3651 symlink pointed to a non-existing file or a file in a directory that
3652 Exim lacked permissions to read, a lookup for a key matching that
3653 symlink would fail. Now it is enough that a matching directory entry
3654 exists, symlink or not. (Bugzilla 503.)
3656 PH/03 The body_linecount and body_zerocount variables are now exported in the
3659 PH/04 Added the $dnslist_matched variable.
3661 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3662 This means they are set thereafter only if the connection becomes
3665 PH/06 Added the client_condition to authenticators so that some can be skipped
3666 by clients under certain conditions.
3668 PH/07 The error message for a badly-placed control=no_multiline_responses left
3669 "_responses" off the end of the name.
3671 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3673 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3674 (without spaces) instead of just copying the configuration text.
3676 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3678 PH/11 Added $max_received_linelength.
3680 PH/12 Added +ignore_defer and +include_defer to host lists.
3682 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3683 way in which PCRE > 7.0 is built.
3685 PH/14 Implemented queue_only_load_latch.
3687 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3688 MAIL command. The effect was to mangle the value on 64-bit systems.
3690 PH/16 Another patch from the Sieve maintainer.
3692 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3694 PH/18 If a system quota error occurred while trying to create the file for
3695 a maildir delivery, the message "Mailbox is full" was not appended to the
3696 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3697 only to a quota excession during the actual writing of the file.
3699 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3700 characters?) which causes problems in log lines. The DN values are now
3701 passed through string_printing() before being added to log lines.
3703 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3704 and InterBase are left for another time.)
3706 PH/21 Added message_body_newlines option.
3708 PH/22 Guard against possible overflow in moan_check_errorcopy().
3710 PH/23 POSIX allows open() to be a macro; guard against that.
3712 PH/24 If the recipient of an error message contained an @ in the local part
3713 (suitably quoted, of course), incorrect values were put in $domain and
3714 $local_part during the evaluation of errors_copy.
3720 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3721 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3724 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3725 issue a MAIL command.
3727 PH/02 In an ACL statement such as
3729 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3731 if a client was not listed at all, or was listed with a value other than
3732 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3733 the condition was not true (as it should be), so access was not denied.
3734 The bug was that the ! inversion was incorrectly passed on to the second
3735 item. This has been fixed.
3737 PH/03 Added additional dnslists conditions == and =& which are different from
3738 = and & when the dns lookup returns more than one IP address.
3740 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3741 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3743 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3744 FSYNC, which compiles an option called disable_fsync that allows for
3745 bypassing fsync(). The documentation is heavily laced with warnings.
3747 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3749 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3750 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3751 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3752 including adding "make clean"; (3) Added -fPIC when compiling the test
3753 dynamically loaded module, to get rid of a warning.
3755 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3756 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3757 = 0s. The bug is otherwise harmless.
3759 PH/07 There was a bug in the dovecot authenticator such that the value of
3760 $auth1 could be overwritten, and so not correctly preserved, after a
3761 successful authentication. This usually meant that the value preserved by
3762 the server_setid option was incorrect.
3764 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3766 PH/09 Installed PCRE release 7.0.
3768 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3769 run for batched SMTP input. It is now run at the start of every message
3770 in the batch. While fixing this I discovered that the process information
3771 (output by running exiwhat) was not always getting set for -bs and -bS
3772 input. This is fixed, and it now also says "batched" for BSMTP.
3774 PH/11 Added control=no_pipelining.
3776 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3777 patch, slightly modified), and move the expansion of helo_data till after
3778 the connection is made in the smtp transport (so it can use these
3781 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3783 PH/14 Added log_selector = +pid.
3785 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3787 PH/16 Add ${if forany and ${if forall.
3789 PH/17 Added dsn_from option to vary the From: line in DSNs.
3791 PH/18 Flush SMTP output before performing a callout, unless control =
3792 no_callout_flush is set.
3794 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3795 was true (the default) a successful delivery failed to delete the retry
3796 item, thus causing premature timeout of the address. The bug is now
3799 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3801 PH/21 Long custom messages for fakedefer and fakereject are now split up
3802 into multiline responses in the same way that messages for "deny" and
3803 other ACL rejections are.
3805 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3806 with slight modification.
3808 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3809 draft, changing the syntax and factoring some duplicate code".
3811 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3812 for deliveries of the second and subsequent messages over the same SMTP
3815 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3816 ${reduce, with only minor "tidies".
3818 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3820 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3821 expansion side effects.
3823 PH/27 When a message times out after an over-quota error from an Exim-imposed
3824 quota, the bounce message says "mailbox is full". This message was not
3825 being given when it was a system quota that was exceeded. It now should
3828 MH/03 Made $recipients available in local_scan(). local_scan() already has
3829 better access to the recipient list through recipients_list[], but
3830 $recipients can be useful in postmaster-provided expansion strings.
3832 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3833 in the case of a MAIL command with additional options following the
3834 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3835 were accidentally chopped off.
3837 PH/29 SMTP synchronization checks are implemented when a command is read -
3838 there is a check that no more input is waiting when there shouldn't be
3839 any. However, for some commands, a delay in an ACL can mean that it is
3840 some time before the response is written. In this time, more input might
3841 arrive, invalidly. So now there are extra checks after an ACL has run for
3842 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3843 pipelining has not been advertised.
3845 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3847 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3848 returned in the SMTP response when smtp_return_error_details was false.
3849 This has been fixed.
3851 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3852 instead of the C I/O that was originally supplied, because problems were
3853 reported on Solaris.
3855 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3856 Exim which did not show up earlier: it was assuming that a call to
3857 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3858 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3859 was a macro that became an assignment, so it seemed to work. This has
3860 changed to a proper function call with a void return, hence the compile
3861 error. Exim's code has been fixed.
3863 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3866 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3868 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3870 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3871 because it turns out to be too big or there is a timeout) there is no
3872 "Completed" line in the log. When some messages of this type were
3873 selected by exigrep, they were listed as "not completed". Others were
3874 picked up by some special patterns. I have improved the selection
3875 criteria to be more general.
3877 PH/38 The host_find_failed option in the manualroute router can now be set
3878 to "ignore", to completely ignore a host whose IP address cannot be
3879 found. If all hosts are ignored, the behaviour is controlled by the new
3880 host_all_ignored option.
3882 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3883 homing or because of multiple MX records with /mx) generated more than
3884 one IP address, and the following item turned out to be the local host,
3885 all the secondary addresses of the first item were incorrectly removed
3886 from the list, along with the local host and any following hosts (which
3887 is what is supposed to happen).
3889 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3890 whoever called Exim into the -H file. In the case of the daemon it was
3891 behaving confusingly. When first started, it used values for whoever
3892 started the daemon, but after a SIGHUP it used the Exim user (because it
3893 calls itself on a restart). I have changed the code so that it now always
3896 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3897 message are rejected with the same error (e.g. no authentication or bad
3898 sender address), and a DATA command is nevertheless sent (as can happen
3899 with PIPELINING or a stupid MUA), the error message that was given to the
3900 RCPT commands is included in the rejection of the DATA command. This is
3901 intended to be helpful for MUAs that show only the final error to their
3904 PH/42 Another patch from the Sieve maintainer.
3906 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3907 Eximstats - Fixed some broken HTML links and added missing column headers
3909 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3910 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3912 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3914 PH/43 Yet another patch from the Sieve maintainer.
3916 PH/44 I found a way to check for a TCP/IP connection going away before sending
3917 the response to the final '.' that terminates a message, but only in the
3918 case where the client has not sent further data following the '.'
3919 (unfortunately, this is allowed). However, in many cases there won't be
3920 any further data because there won't be any more messages to send. A call
3921 to select() can be used: if it shows that the input is "ready", there is
3922 either input waiting, or the socket has been closed. An attempt to read
3923 the next input character can distinguish the two cases. Previously, Exim
3924 would have sent an OK response which the client would never have see.
3925 This could lead to message repetition. This fix should cure that, at
3926 least in a lot of common cases.
3928 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3929 advertised in response to EHLO.
3935 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3936 fixed by 4.65/MH/01 (is this a record?) are fixed:
3938 (i) An empty string was always treated as zero by the numeric comparison
3939 operators. This behaviour has been restored.
3941 (ii) It is documented that the numeric comparison operators always treat
3942 their arguments as decimal numbers. This was broken in that numbers
3943 starting with 0 were being interpreted as octal.
3945 While fixing these problems I realized that there was another issue that
3946 hadn't been noticed. Values of message_size_limit (both the global option
3947 and the transport option) were treated as octal if they started with 0.
3948 The documentation was vague. These values are now always treated as
3949 decimal, and I will make that clear in the documentation.
3955 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3956 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3959 MH/01 Don't check that the operands of numeric comparison operators are
3960 integers when their expansion is in "skipping" mode (fixes bug
3961 introduced by 4.64-PH/07).
3963 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3964 child addresses, Exim now panics and dies. Previously, because the count
3965 is held in a short int, deliveries were likely to be lost. As such a
3966 large number of recipients for a single message is ridiculous
3967 (performance will be very, very poor), I have chosen to impose a limit
3968 rather than extend the field.
3974 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3975 leftover -K file (the existence of which was triggered by #402).
3976 While we were at it, introduced process PID as part of the -K
3977 filename. This should rule out race conditions when creating
3980 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3981 processing considerably. Previous code took too long for large mails,
3982 triggering a timeout which in turn triggers #401.
3984 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3985 in the DK code in transports.c. sendfile() is not really portable,
3986 hence the _LINUX specificness.
3988 TF/01 In the add_headers option to the mail command in an Exim filter,
3989 there was a bug that Exim would claim a syntax error in any
3990 header after the first one which had an odd number of characters
3993 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3994 callout verification, Exim cached a "reject" for the entire domain. This
3995 is correct for most verifications, but it is not correct for a recipient
3996 verification with use_sender or use_postmaster set, because in that case
3997 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3998 case of MAIL FROM:<> rejection from other early rejections (e.g.
3999 rejection of HELO). When verifying a recipient using a non-null MAIL
4000 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4001 Whatever the result of the callout, the value of the domain cache is
4002 left unchanged (for any other kind of callout, getting as far as trying
4003 RCPT means that the domain itself is ok).
4005 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4008 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4009 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4010 session, the daemon ignores ECONNECT errors and logs others; it now
4011 ignores EPIPE as well.
4013 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4014 (quoted-printable decoding).
4016 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4017 later the small subsequent patch to fix an introduced bug.
4019 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4021 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4023 PH/08 An error is now given if message_size_limit is specified negative.
4025 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4026 to be given (somewhat) arbitrary names.
4028 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4031 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4032 miscellaneous code fixes
4034 PH/10 Added the log_reject_target ACL modifier to specify where to log
4037 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4038 hostname. This is wrong, because it relates to the incoming message (and
4039 probably the interface on which it is arriving) and not to the outgoing
4040 callout (which could be using a different interface). This has been
4041 changed to use the value of the helo_data option from the smtp transport
4042 instead - this is what is used when a message is actually being sent. If
4043 there is no remote transport (possible with a router that sets up host
4044 addresses), $smtp_active_hostname is used.
4046 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4047 tweaks were necessary in order to get it to work (see also 21 below):
4048 (a) The code assumed that strncpy() returns a negative number on buffer
4049 overflow, which isn't the case. Replaced with Exim's string_format()
4051 (b) There were several signed/unsigned issues. I just did the minimum
4052 hacking in of casts. There is scope for a larger refactoring.
4053 (c) The code used strcasecmp() which is not a standard C function.
4054 Replaced with Exim's strcmpic() function.
4055 (d) The code set only $1; it now sets $auth1 as well.
4056 (e) A simple test gave the error "authentication client didn't specify
4057 service in request". It would seem that Dovecot has changed its
4058 interface. Fortunately there's a specification; I followed it and
4059 changed what the client sends and it appears to be working now.
4061 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4064 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4065 address A is aliased to B and C, where B exists and C does not. Without
4066 -v the output is "A verified" because verification stops after a
4067 successful redirection if more than one address is generated. However,
4068 with -v the child addresses are also verified. Exim was outputting "A
4069 failed to verify" and then showing the successful verification for C,
4070 with its parentage. It now outputs "B failed to verify", showing B's
4071 parentage before showing the successful verification of C.
4073 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4074 look up a TXT record in a specific list after matching in a combined
4077 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4078 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4079 they consult the DNS. I had assumed they would set it the way they
4080 wanted; and indeed my experiments on Linux seem to show that in some
4081 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4082 To be on the safe side, however, I have now made the interface to
4083 host_find_byname() similar to host_find_bydns(), with an argument
4084 containing the DNS resolver options. The host_find_byname() function now
4085 sets these options at its start, just as host_find_bydns() does. The smtp
4086 transport options dns_qualify_single and dns_search_parents are passed to
4087 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4088 of host_find_byname() use the default settings of RES_DEFNAMES
4089 (qualify_single) but not RES_DNSRCH (search_parents).
4091 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4092 spool_read_header() do less string testing, by means of a preliminary
4093 switch on the second character of optional "-foo" lines. (This is
4094 overdue, caused by the large number of possibilities that now exist.
4095 Originally there were few.) While I was there, I also converted the
4096 str(n)cmp tests so they don't re-test the leading "-" and the first
4097 character, in the hope this might squeeze out yet more improvement.
4099 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4100 flag allowing group syntax was set by the header_syntax check but not
4101 turned off, possible causing trouble later; (2) The flag was not being
4102 set at all for the header_verify test, causing "group"-style headers to
4103 be rejected. I have now set it in this case, and also caused header_
4104 verify to ignore an empty address taken from a group. While doing this, I
4105 came across some other cases where the code for allowing group syntax
4106 while scanning a header line wasn't quite right (mostly, not resetting
4107 the flag correctly in the right place). These bugs could have caused
4108 trouble for malformed header lines. I hope it is now all correct.
4110 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4111 with the "reply" argument non-NULL. The code, however (which originally
4112 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4113 but it didn't always do it. This confused somebody who was copying the
4114 code for some other use. I have removed all the tests.
4116 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4117 feature that was used to support insecure browsers during the U.S. crypto
4118 embargo. It requires special client support, and Exim is probably the
4119 only MTA that supported it -- and would never use it because real RSA is
4120 always available. This code has been removed, because it had the bad
4121 effect of slowing Exim down by computing (never used) parameters for the
4122 RSA_EXPORT functionality.
4124 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4125 authenticator to fail if there's a tab character in the incoming data
4126 (there should never be unless someone is messing about, as it's supposed
4127 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4128 if the connection is using TLS or if the remote IP is the same as the
4129 local IP, and the "valid-client-cert option" if a client certificate has
4132 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4133 authenticators. This can be used for authorization after authentication
4134 succeeds. (In the case of plaintext, it servers for both authentication
4137 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4138 if any retry times were supplied.
4140 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4141 connection, where there is no client IP address to check. In this
4142 situation, the verify now always succeeds.
4144 PH/25 Applied John Jetmore's -Mset patch.
4146 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4148 PH/27 In a string expansion for a processed (not raw) header when multiple
4149 headers of the same name were present, leading whitespace was being
4150 removed from all of them, but trailing whitespace was being removed only
4151 from the last one. Now trailing whitespace is removed from each header
4152 before concatenation. Completely empty headers in a concatenation (as
4153 before) are ignored.
4155 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4156 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4158 PH/29 [Removed. This was a change that I later backed out, and forgot to
4159 correct the ChangeLog entry (that I had efficiently created) before
4160 committing the later change.]
4162 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4163 address errors (4xx response to RCPT) over the same connection as other
4164 messages routed to the same hosts. Such deliveries are always "forced",
4165 so retry times are not inspected. This resulted in far too many retries
4166 for the affected addresses. The effect occurred only when there were more
4167 hosts than the hosts_max_try setting in the smtp transport when it had
4168 the 4xx errors. Those hosts that it had tried were not added to the list
4169 of hosts for which the message was waiting, so if all were tried, there
4170 was no problem. Two fixes have been applied:
4172 (i) If there are any address or message errors in an SMTP delivery, none
4173 of the hosts (tried or untried) are now added to the list of hosts
4174 for which the message is waiting, so the message should not be a
4175 candidate for sending over the same connection that was used for a
4176 successful delivery of some other message. This seems entirely
4177 reasonable: after all the message is NOT "waiting for some host".
4178 This is so "obvious" that I'm not sure why it wasn't done
4179 previously. Hope I haven't missed anything, but it can't do any
4180 harm, as the worst effect is to miss an optimization.
4182 (ii) If, despite (i), such a delivery is accidentally attempted, the
4183 routing retry time is respected, so at least it doesn't keep
4184 hammering the server.
4186 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4187 in ${readsocket because some servers need this prod.
4189 PH/32 Added some extra debug output when updating a wait-xxx database.
4191 PH/33 The hint "could be header name not terminated by colon", which has been
4192 given for certain expansion errors for a long time, was not being given
4193 for the ${if def:h_colon_omitted{... case.
4195 PH/34 The spec says: "With one important exception, whenever a domain list is
4196 being scanned, $domain contains the subject domain." There was at least
4197 one case where this was not true.
4199 PH/35 The error "getsockname() failed: connection reset by peer" was being
4200 written to the panic log as well as the main log, but it isn't really
4201 panic-worthy as it just means the connection died rather early on. I have
4202 removed the panic log writing for the ECONNRESET error when getsockname()
4205 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4206 runs only) independently of the message's sender address. This meant
4207 that, if the 4xx error was in fact related to the sender, a different
4208 message to the same recipient with a different sender could confuse
4209 things. In particular, this can happen when sending to a greylisting
4210 server, but other circumstances could also provoke similar problems.
4211 I have changed the default so that the retry time for these errors is now
4212 based a combination of the sender and recipient addresses. This change
4213 can be overridden by setting address_retry_include_sender=false in the
4216 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4217 remote server are returned as part of bounce messages. This was not
4218 happening for LMTP over a pipe (the lmtp transport), but now it is the
4219 same for both kinds of LMTP.
4221 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4222 in header lines that were in fact CNAMEs. This is no longer the case.
4224 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4225 and queue runs started by the daemon processed all messages. This has
4226 been fixed so that -R and -S can now usefully be given with -q<time>.
4228 PH/40 Import PCRE release 6.7 (fixes some bugs).
4230 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4232 PH/42 Give an error if -q is specified more than once.
4234 PH/43 Renamed the variables $interface_address and $interface_port as
4235 $received_ip_address and $received_port, to make it clear that these
4236 values apply to message reception, and not to the outgoing interface when
4237 a message is delivered. (The old names remain recognized, of course.)
4239 PH/44 There was no timeout on the connect() call when using a Unix domain
4240 socket in the ${readsocket expansion. There now is.
4242 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4243 be meaningful with "accept".
4245 SC/01 Eximstats V1.43
4246 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4248 SC/02 Eximstats V1.44
4249 Use a glob alias rather than an array ref in the generated
4250 parser. This improves both readability and performance.
4252 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4253 Collect SpamAssassin and rejection statistics.
4254 Don't display local sender or destination tables unless
4255 there is data to show.
4256 Added average volumes into the top table text output.
4258 SC/04 Eximstats V1.46
4259 Collect data on the number of addresses (recipients)
4260 as well as the number of messages.
4262 SC/05 Eximstats V1.47
4263 Added 'Message too big' to the list of mail rejection
4264 reasons (thanks to Marco Gaiarin).
4266 SC/06 Eximstats V1.48
4267 Mainlog lines which have GMT offsets and are too short to
4268 have a flag are now skipped.
4270 SC/07 Eximstats V1.49 (Alain Williams)
4271 Added the -emptyok flag.
4273 SC/08 Eximstats V1.50
4274 Fixes for obtaining the IP address from reject messages.
4276 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4277 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4278 whitespace changes from 4.64-PH/27
4280 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4283 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4284 are found, allow negative numbers in numeric criteria)
4286 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4288 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4289 to match changes made in 4.64-PH/43
4291 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4293 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4294 "domain" is actually an IP address, and force a failure. This locks out
4295 those revolvers/nameservers that support "A-for-A" lookups, in
4296 contravention of the specifications.
4298 PH/48 When a host name was looked up from an IP address, and the subsequent
4299 forward lookup of the name timed out, the host name was left in
4300 $sender_host_name, contrary to the specification.
4302 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4303 restricted to single-key lookups, Exim was not diagnosing an error if
4304 * or *@ was used with a query-style lookup.
4306 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4308 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4309 long ago, but noone interested enough thought of it. Let's just say that
4310 the "1.1" means that there are some new functions that weren't there at
4311 some point in the past.
4313 PH/51 Error processing for expansion failure of helo_data from an smtp
4314 transport during callout processing was broken.
4316 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4317 tested/used via the -bh/-bhc/-bs options.
4319 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4320 bug, fixed in subsequent PCRE releases).
4322 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4323 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4325 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4331 SC/01 Use a glob alias rather than an array ref in eximstats generated
4332 parser. This improves both readability and performance.
4334 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4335 Don't display local sender or destination tables in eximstats unless
4336 there is data to show.
4337 Added average volumes into the eximstats top table text output.
4339 SC/03 Collect data on the number of addresses (recipients) as well
4340 as the number of messages in eximstats.
4342 TF/01 Correct an error in the documentation for the redirect router. Exim
4343 does (usually) call initgroups() when daemonizing.
4345 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4346 with consistent privilege compared to when running as a daemon.
4348 TF/03 Note in the spec that $authenticated_id is not set for local
4349 submissions from trusted users.
4351 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4352 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4354 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4355 by adding some example configuration directives to the default
4356 configuration file. A little bit of work is required to uncomment the
4357 directives and define how usernames and passwords are checked, but
4358 there is now a framework to start from.
4360 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4361 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4362 without this. I don't know how relevant this is to other LDAP libraries.
4364 PH/02 Add the verb name to the "unknown ACL verb" error.
4366 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4368 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4370 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4371 directory not expanded when it should be if an expanded home directory
4372 was set for the address (which is overridden by the transport).
4374 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4377 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4378 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4379 because it is too late at that time, and has no effect.
4381 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4382 security issue with \' (bugzilla #107). I could not use the
4383 PQescapeStringConn() function, because it needs a PGconn value as one of
4386 PH/08 When testing addresses using -bt, indicate those final addresses that
4387 are duplicates that would not cause an additional delivery. At least one
4388 person was confused, thinking that -bt output corresponded to deliveries.
4389 (Suppressing duplicates isn't a good idea as you lose the information
4390 about possibly different redirections that led to the duplicates.)
4392 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4393 systems where poll() doesn't work, in particular OS X.
4395 PH/10 Added more information to debugging output for retry time not reached.
4397 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4398 operations in malware.c.
4400 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4403 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4404 syslog_duplication set false, log lines that would normally be written
4405 both the the main log and to the reject log were not written to syslog at
4408 PH/14 In the default configuration, change the use of "message" in ACL warn
4409 statements to "add_header".
4411 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4412 not followed by a command (e.g. "seen endif").
4414 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4415 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4418 PH/17 Added extra conditions to the default value of delay_warning_condition
4422 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4423 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4424 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4427 The Auto-Submitted: and various List- headers are standardised, whereas I
4428 don't think Precedence: ever was.
4430 PH/18 Refactored debugging code in route_finduser() to show more information,
4431 in particular, the error code if getpwnam() issues one.
4433 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4434 This is apparently needed in addition to the PH/07 change above to avoid
4435 any possible encoding problems.
4437 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4438 but not after initializing Perl.
4440 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4441 output them only if debugging. By default they are written stderr,
4442 apparently, which is not desirable.
4444 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4447 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4450 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4452 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4453 authenticated or an ident call has been made. Suppress the default
4454 values for $authenticated_id and $authenticated_sender (but permit -oMai
4455 and -oMas) when testing with -bh.
4457 PH/24 Re-jigged the order of the tests in the default configuration so that the
4458 tests for valid domains and recipients precede the DNS black list and CSA
4459 tests, on the grounds that those ones are more expensive.
4461 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4462 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4463 This bug exists in every version of Exim that I still have, right back to
4466 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4467 However, an attempt to turn on case-sensitivity in a regex key by
4468 including (?-i) didn't work because the subject string was already
4469 lowercased, and the effects were non-intuitive. It turns out that a
4470 one-line patch can be used to allow (?-i) to work as expected.
4476 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4477 other effects) broke the use of negated acl sub-conditions.
4479 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4482 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4483 "Deny" causes Exim to reject the incoming connection with a 554 error.
4484 Unfortunately, if there is a major crisis, such as a disk failure,
4485 tcp-wrappers gives "deny", whereas what one would like would be some
4486 kind of temporary error. A kludge has been added to help with this.
4487 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4488 554 error is used if errno is still zero or contains ENOENT (which occurs
4489 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4492 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4494 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4495 errors. Otherwise a message that provokes a temporary error (when other
4496 messages do not) can cause a whole host to time out.
4498 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4499 addresses were routed directly to files or pipes from a redirect router.
4500 File deliveries just didn't batch; pipe deliveries might have suffered
4503 PH/06 A failure to get a lock for a hints database would erroneously always say
4504 "Failed to get write lock", even when it was really a read lock.
4506 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4507 of 0600. This has been changed to use the value of the lockfile_mode
4508 option (which defaults to 0600).
4510 PH/08 Applied small patch from the Sieve maintainer.
4512 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4513 folder from quota calculations, a direct delivery into this folder messed
4514 up the contents of the maildirsize file. This was because the regex was
4515 used only to exclude .Trash (or whatever) when the size of the mailbox
4516 was calculated. There was no check that a delivery was happening into an
4517 excluded directory. This bug has been fixed by ignoring all quota
4518 processing for deliveries into excluded directories.
4520 PH/10 Added the maildirfolder_create_regex option to appendfile.
4526 PH/01 The code for finding all the local interface addresses on a FreeBSD
4527 system running IPv6 was broken. This may well have applied to all BSD
4528 systems, as well as to others that have similar system calls. The broken
4529 code found IPv4 interfaces correctly, but gave incorrect values for the
4530 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4531 that it would not match correctly against @[] and not recognize the IPv6
4534 PH/02 The ipliteral router was not recognizing addresses of the form user@
4535 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4537 PH/03 Added disable_ipv6.
4539 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4540 decoded form, because it is most often used to construct To: headers
4541 lines in autoreplies, and the decoded form may well be syntactically
4542 invalid. However, $reply_address has leading white space removed, and all
4543 newlines turned into spaces so that the autoreply transport does not
4546 PH/05 If group was specified without a user on a router, and no group or user
4547 was specified on a transport, the group from the router was ignored.
4549 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4550 for visible compile-time settings that can be used to change these
4551 numbers, for those that want even more. Backwards compatibility with old
4552 spool files has been maintained. However, going back to a previous Exim
4553 release will lost any variables that are in spool files.
4555 PH/07 Two small changes when running in the test harness: increase delay when
4556 passing a TCP/IP connection to a new process, in case the original
4557 process has to generate a bounce, and remove special handling of
4558 127.0.0.2 (sic), which is no longer necessary.
4560 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4561 be the same on different OS.
4563 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4566 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4567 whether --show-vars was specified or not
4569 JJ/02 exipick: Added support for new ACL variable spool format introduced
4572 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4573 syntactically invalid From: or Reply-to: line, and a filter used this to
4574 generate an autoreply, and therefore failed to obtain an address for the
4575 autoreply, Exim could try to deliver to a non-existent relative file
4576 name, causing unrelated and misleading errors. What now happens is that
4577 it logs this as a hard delivery error, but does not attempt to create a
4580 PH/11 The exinext utility has a -C option for testing purposes, but although
4581 the given file was scanned by exinext itself; it wasn't being passed on
4582 when Exim was called.
4584 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4585 an end-of-file indication when reading a command response.
4587 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4588 compiled. In many other places in Exim, IPv6 addresses are always
4589 recognized, so I have changed this. It also means that IPv4 domain
4590 literals of the form [IPV4:n.n.n.n] are now always recognized.
4592 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4593 used if the router is not running as root, for example, when verifying at
4594 ACL time, or when using -bh. The debugging output from this situation was
4595 non-existent - all you got was a failure to exec. I have made two
4598 (a) Failures to set uid/gid, the current directory, or a process leader
4599 in a subprocess such as that created by queryprogram now generate
4600 suitable debugging output when -d is set.
4602 (b) The queryprogram router detects when it is not running as root,
4603 outputs suitable debugging information if -d is set, and then runs
4604 the subprocess without attempting to change uid/gid.
4606 PH/15 Minor change to Makefile for building test_host (undocumented testing
4609 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4610 additional section of a DNS packet that returns MX or SRV records.
4611 Instead, it always explicitly searches for A/AAAA records. This avoids
4612 major problems that occur when a DNS server includes only records of one
4613 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4614 fixed another bug: if SRV records were looked up and the corresponding
4615 address records were *not* found in the additional section, the port
4616 values from the SRV records were lost.
4618 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4619 using the correct key (the original address) when searching the retry
4620 rules in order to find which one to use for generating the retry hint.
4622 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4623 adding the default one. Similarly, if it contains a Reply-To: header, the
4624 errors_reply_to option, if set, is not used.
4626 PH/19 When calculating a retry time, Exim used to measure the "time since
4627 failure" by looking at the "first failed" field in the retry record. Now
4628 it does not use this if it is later than than the arrival time of the
4629 message. Instead it uses the arrival time. This makes for better
4630 behaviour in cases where some deliveries succeed, thus re-setting the
4631 "first failed" field. An example is a quota failure for a huge message
4632 when small messages continue to be delivered. Without this change, the
4633 "time since failure" will always be short, possible causing more frequent
4634 delivery attempts for the huge message than are intended.
4635 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4637 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4638 $1, $2, $3) because the numerical variables can be reset during some
4639 expansion items (e.g. "match"), thereby losing the authentication data.
4641 PH/21 Make -bV show the size of off_t variables so that the test suite can
4642 decide whether to run tests for quotas > 2G.
4644 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4645 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4646 filecount value is greater than 2G or if a quota value is greater than 2G
4647 on a system where the size of off_t is not greater than 4, a panic error
4650 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4651 never match. The debug and -bh output now contains an explicit error
4652 message indicating a malformed IPv4 address or mask.
4654 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4655 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4656 PH/23 above applies.
4658 PH/25 Do not write to syslog when running in the test harness. The only
4659 occasion when this arises is a failure to open the main or panic logs
4660 (for which there is an explicit test).
4662 PH/26 Added the /no_tell option to "control=freeze".
4664 PH/27 If a host name lookup failed very early in a connection, for example, if
4665 the IP address matched host_lookup and the reverse lookup yielded a name
4666 that did not have a forward lookup, an error message of the form "no IP
4667 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4668 could be logged. Now it outputs the IP address instead of "NULL".
4670 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4671 allows the sender and the authenticated sender to be set when
4672 submitting a message from within Exim. Since child_open_exim() is
4673 documented for local_scan(), the new function should be too.
4675 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4676 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4677 results in an empty string is now treated as unset.
4679 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4681 PH/31 Added sender_verify_fail logging option.
4683 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4684 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4685 tidied the source and removed it altogether.
4687 PH/33 When a queue run was abandoned because the load average was too high, a
4688 log line was always written; now it is written only if the queue_run log
4689 selector is set. In addition, the log line for abandonment now contains
4690 information about the queue run such as the pid. This is always present
4691 in "start" and "stop" lines but was omitted from the "abandon" line.
4693 PH/34 Omit spaces between a header name and the colon in the error message that
4694 is given when verify = headers_syntax fails (if there are lots of them,
4695 the message gets confusing).
4697 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4698 names, as there are now some PTR records that contain slashes. This check
4699 is only to protect against broken name servers that fall over on strange
4700 characters, so the fact that it applies to all lookups doesn't matter.
4702 PH/36 Now that the new test suite is complete, we can remove some of the
4703 special code in Exim that was needed for the old test suite. For example,
4704 sorting DNS records because real resolvers return them in an arbitrary
4705 order. The new test suite's fake resolver always returns records in the
4708 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4709 bounces) except when queue_only is set, to avoid logging races between
4710 the different processes.
4712 PH/38 Panic-die if .include specifies a non-absolute path.
4714 PH/39 A tweak to the "H" retry rule from its user.
4716 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4717 a label. They prevented compilation on older perls.
4719 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4720 a warning to be raised on newish perls.
4722 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4723 on queue. Changes to match documented behaviour of showing count of
4724 messages matching specified criteria.
4726 PH/40 Changed the default ident timeout from 30s to 5s.
4728 PH/41 Added support for the use of login_cap features, on those BSD systems
4729 that have them, for controlling the resources used by pipe deliveries.
4731 PH/42 The content-scanning code uses fopen() to create files in which to put
4732 message data. Previously it was not paying any attention to the mode of
4733 the files. Exim runs with umask(0) because the rest of the code creates
4734 files with open(), and sets the required mode explicitly. Thus, these
4735 files were ending up world-writeable. This was not a big issue, because,
4736 being within the spool directory, they were not world-accessible. I have
4737 created a function called modefopen, which takes an additional mode
4738 argument. It sets umask(777), creates the file, chmods it to the required
4739 mode, then resets the umask. All the relevant calls to fopen() in the
4740 content scanning code have been changed to use this function.
4742 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4743 to 24 hours. This avoids potential overflow problems when processing G
4744 and H retry rules. I suspect nobody ever tinkers with this value.
4746 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4748 PH/45 When the plaintext authenticator is running as a client, the server's
4749 challenges are checked to ensure they are valid base64 strings. By
4750 default, the authentication attempt is cancelled if an invalid string is
4751 received. Setting client_ignore_invalid_base64 true ignores these errors.
4752 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4753 they are received. Thus, the responses can be made to depend on the
4754 challenges. If an invalid string is ignored, an empty string is placed in
4757 PH/46 Messages that are created by the autoreply transport now contains a
4758 References: header, in accordance with RFCs 2822 and 3834.
4760 PH/47 Added authenticated_sender_force to the smtp transport.
4762 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4764 PH/49 Installed latest patch from the Sieve maintainer.
4766 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4767 was also set, the appendfile transport was unnecessarily scanning a
4768 directory of message files (e.g. for maildir delivery) to find the count
4769 of files (along with the size), even though it did not need this
4770 information. It now does the scan only if it needs to find either the
4771 size of the count of files.
4773 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4775 PH/52 Two bugs concerned with error handling when the smtp transport is
4778 (i) Exim was not creating retry information for temporary errors given
4779 for individual recipients after the DATA command when the smtp transport
4780 was used in LMTP mode. This meant that they could be retried too
4781 frequently, and not timed out correctly.
4783 (ii) Exim was setting the flag that allows error details to be returned
4784 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4785 recipients that were returned after the DATA command.
4787 PH/53 This is related to PH/52, but is more general: for any failing address,
4788 when detailed error information was permitted to be returned to the
4789 sender, but the error was temporary, then after the final timeout, only
4790 "retry timeout exceeded" was returned. Now it returns the full error as
4791 well as "retry timeout exceeded".
4793 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4794 do this, and (what is worse) MTAs that accept it.
4796 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4797 will now be deprecated.
4799 PH/56 New os.c-cygwin from the Cygwin maintainer.
4801 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4802 formats (previously only available in exim formats via -bpr, -bpru,
4803 and -bpra. Now also available in native and exiqgrep formats)
4805 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4806 with very large, slow to parse queues
4808 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4810 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4812 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4813 responses to authentication challenges, though it was showing the
4814 challenges; (ii) I've removed the CR characters from the debug output for
4817 PH/58 Allow for the insertion of a newline as well as a space when a string
4818 is turned into more than one encoded-word during RFC 2047 encoding. The
4819 Sieve code now uses this.
4821 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4822 data_4xx, lost_connection, tls_required.
4824 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4825 message was being sent as an SMTP response.
4827 PH/61 Add -l and -k options to exicyclog.
4829 PH/62 When verifying, if an address was redirected to one new address, so that
4830 verification continued, and the new address failed or deferred after
4831 having set something in $address_data, the value of $address_data was not
4832 passed back to the ACL. This was different to the case when no
4833 redirection occurred. The value is now passed back in both cases.
4835 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4836 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4837 use login_cap.h, so on its own it isn't the distinguishing feature. The
4838 new name refers directly to the setclassresources() function.
4840 PH/65 Added configuration files for NetBSD3.
4842 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4844 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4845 is preferred over IPv4.
4847 PH/68 The bounce_return_message and bounce_return_body options were not being
4848 honoured for bounces generated during the reception of non-SMTP messages.
4849 In particular, this applied to messages rejected by the ACL. This bug has
4850 been fixed. However, if bounce_return_message is true and bounce_return_
4851 body is false, the headers that are returned for a non-SMTP message
4852 include only those that have been read before the error was detected.
4853 (In the case of an ACL rejection, they have all been read.)
4855 PH/69 The HTML version of the specification is now built in a directory called
4856 spec_html instead of spec.html, because the latter looks like a path with
4857 a MIME-type, and this confuses some software.
4859 PH/70 Catch two compiler warnings in sieve.c.
4861 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4862 function verify_get_ident() calls ip_connect() to connect a socket, but
4863 if the "connect()" function timed out, ip_connect() used to close the
4864 socket. However, verify_get_ident() also closes the socket later, and in
4865 between Exim writes to the log, which may get opened at this point. When
4866 the socket was closed in ip_connect(), the log could get the same file
4867 descriptor number as the socket. This naturally causes chaos. The fix is
4868 not to close the socket in ip_connect(); the socket should be closed by
4869 the function that creates it. There was only one place in the code where
4870 this was missing, in the iplookup router, which I don't think anybody now
4871 uses, but I've fixed it anyway.
4873 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4874 well as to direct DNS lookups. Otherwise the handling of names in host
4875 lists is inconsistent and therefore confusing.
4881 PH/01 Two changes to the default runtime configuration:
4883 (1) Move the checks for relay_from_hosts and authenticated clients from
4884 after to before the (commented out) DNS black list checks.
4886 (2) Add control=submission to the relay_from_hosts and authenticated
4887 clients checks, on the grounds that messages accepted by these
4888 statements are most likely to be submissions.
4890 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4892 (1) Generate an error if the third argument for the ${prvs expansion is
4895 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4898 (3) Reset the variables that are obtained from the first argument of
4899 ${prvscheck and used in the second argument before leaving the code,
4900 because their memory is reclaimed, so using them afterwards may do
4903 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4904 one (it's much easier than Tom thought :-).
4906 (5) Because of (4), we can now allow for the use of $prvscheck_result
4907 inside the third argument.
4909 PH/03 For some reason, the default setting of PATH when running a command from
4910 a pipe transport was just "/usr/bin". I have changed it to
4913 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4914 anything to be listed in the output from -bV.
4916 PH/05 When a filter generated an autoreply, the entire To: header line was
4917 quoted in the delivery log line, like this:
4919 => >A.N.Other <ano@some.domain> <original@ddress> ...
4921 This has been changed so that it extracts the operative address. There
4922 may be more than one such address. If so, they are comma-separated, like
4925 => >ano@some.domain,ona@other.domain <original@ddress> ...
4927 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4928 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4929 not being looked up in the rDNS to get a host name, Exim was showing the
4930 IP address twice in Received: lines, even though the IP addresses were
4931 identical. For example:
4933 Received: from [1.2.3.4] (helo=[1.2.3.4])
4935 However, if the real host name was known, it was omitting the HELO data
4936 if it matched the actual IP address. This has been tidied up so that it
4937 doesn't show the same IP address twice.
4939 PH/07 When both +timestamp and +memory debugging was on, the value given by
4940 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4941 called by the debug printing, thereby overwriting the timestamp buffer.
4942 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4944 PH/08 When the original message was included in an autoreply transport, it
4945 always said "this is a copy of the message, including all the headers",
4946 even if body_only or headers_only was set. It now gives an appropriate
4949 PH/09 Applied a patch from the Sieve maintainer which:
4951 o fixes some comments
4952 o adds the (disabled) notify extension core
4953 o adds some debug output for the result of if/elsif tests
4954 o points to the current vacation draft in the documentation
4955 and documents the missing references header update
4959 o fixes a bug in processing the envelope test (when testing
4960 multiple envelope elements, the last element determined the
4963 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4964 Electronic Mail") by including:
4966 Auto-submitted: auto-generated
4968 in the messages that it generates (bounce messages and others, such as
4969 warnings). In the case of bounce messages for non-SMTP messages, there was
4970 also a typo: it was using "Auto_submitted" (underscore instead of
4971 hyphen). Since every message generated by Exim is necessarily in response
4972 to another message, thes have all been changed to:
4974 Auto-Submitted: auto-replied
4976 in accordance with these statements in the RFC:
4978 The auto-replied keyword:
4980 - SHOULD be used on messages sent in direct response to another
4981 message by an automatic process,
4983 - MUST NOT be used on manually-generated messages,
4985 - MAY be used on Delivery Status Notifications (DSNs) and Message
4986 Disposition Notifications (MDNs),
4988 - MUST NOT be used on messages generated by automatic or periodic
4989 processes, except for messages which are automatic responses to
4992 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4993 to the default Received: header definition.
4995 PH/12 Added log selector acl_warn_skipped (default on).
4997 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4998 variables because (a) they are in the wrong storage pool and (b) even if
4999 they were copied, it wouldn't work properly because of the caching.
5001 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5002 checking when decoding. Apparently there are clients that generate
5003 overlong encoded strings. Why am I not surprised?
5005 PH/15 If the first argument of "${if match_address" was not empty, but did not
5006 contain an "@" character, Exim crashed. Now it writes a panic log message
5007 and treats the condition as false.
5009 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5011 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5012 extension "envelope-auth". The code is finished and in agreement with
5013 other implementations, but there is no documentation so far and in fact,
5014 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5015 not changing the active code.
5017 Print executed "if" and "elsif" statements when debugging is used. This
5018 helps a great deal to understand what a filter does.
5020 Document more things not specified clearly in RFC3028. I had all this
5021 sorted out, when out of a sudden new issues came to my mind. Oops."
5023 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5026 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5027 canonical form (as documented). However, after a host name lookup from
5028 the IP address, check_host() was doing a simple string comparison with
5029 addresses acquired from the DNS when checking that the found name did
5030 have the original IP as one of its addresses. Since any found IPv6
5031 addresses are likely to be in abbreviated form, the comparison could
5032 fail. Luckily, there already exists a function for doing the comparison
5033 by converting both addresses to binary, so now that is used instead of
5034 the text comparison.
5036 PH/20 There was another similar case to PH/19, when a complete host name was
5037 given in a host list; looking up its IP address could give an abbreviated
5038 form, whereas the current host's name might or might not be abbreviated.
5039 The same fix has been applied.
5045 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5046 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5049 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5050 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5052 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5054 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5055 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5056 Also, for "Auto-Submitted", treat anything other than "no" as
5057 non-personal, in accordance with RFC 3834. (Previously it treated
5058 anything starting "auto-" as non-personal.)
5060 TF/01 The control=submission/name=... option had a problem with syntax
5061 errors if the name included a slash character. The /name= option
5062 now slurps the rest of the string, so it can include any characters
5063 but it must come last in the list of options (after /sender_retain
5066 PH/05 Some modifications to the interface to the fake nameserver for the new
5074 TK/01 Added the "success_on_redirect" address verification option. See
5075 NewStuff for rationale and an example.
5077 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5079 PH/02 Patch to exigrep to allow it to work on syslog lines.
5081 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5082 fread() to skip over the body file's header line, because in Cygwin the
5083 header line is locked and is inaccessible.
5085 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5086 co-exist for some time) to make it clear that it is the Exim ID that is
5087 referenced, not the Message-ID: header line.
5089 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5090 string_format() function, because snprintf() does not exist on all
5093 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5094 ${stat: expansion item.
5096 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5097 protocol synchronization error", to keep the pedants happy.
5099 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5100 well as for IRIX systems, when gcc is being used. See the host.c source
5103 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5105 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5108 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5109 transport and to the smtp transport in LMTP mode.
5111 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5113 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5114 run a filter in a subprocess. This could lead to confusion in subsequent
5115 lookups in the parent process. There should also be a search_tidyup() at
5116 the end of the subprocess.
5118 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5119 only if the host matched helo_try_verify_hosts, which caused the
5120 verification to occur when the EHLO/HELO command was issued. The ACL just
5121 tested the remembered result. Now, if a previous verification attempt has
5122 not happened, "verify = helo" does it there and then.
5124 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5126 TK/03 Fix log output including CR from clamd.
5128 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5129 exist provoked a memory error which could cause a segfault.
5131 PH/15 Installed PCRE 6.2
5133 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5135 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5136 of the problem. Specifically, suggested +O2 rather than +O1 for the
5139 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5141 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5142 record, the port was not correctly logged when the outgoing_port log
5143 selector was set (it logged the transort's default port).
5145 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5146 fallback_hosts, and "hosts" in the smtp transport.
5148 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5149 host errors such as "Connection refused".
5151 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5152 authentication with radiusclient 0.4.9:
5154 - Error returned from rc_read_config was caught wrongly
5155 - Username/password not passed on to radius server due to wrong length.
5157 The presumption is that some radiusclient API changes for 4.51/PH/17
5158 were not taken care of correctly. The code is still untested by me (my
5159 Linux distribution still has 0.3.2 of radiusclient), but it was
5160 contributed by a Radius user.
5162 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5163 expanding the "port" option of the smtp transport.
5165 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5166 while reading a MIME header. Thanks to Tom Hughes for a patch.
5168 PH/24 Include config.h inside local_scan.h so that configuration settings are
5171 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5172 that in an ACL for RCPT (for example), you can examine exactly what was
5175 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5176 commands, but it was not correctly comparing the address with the actual
5177 client host address. Thus, it would show the EHLO address in Received:
5178 header lines when this was not necessary.
5180 PH/27 Added the % operator to ${eval:}.
5182 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5183 it should be ignoring failures (because with -C, for example, it has lost
5184 privilege). It wasn't ignoring creation failures other than "already
5187 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5190 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5191 because an input line was too long, either on its own, or by virtue of
5192 too many continuations, the temporary file was not being removed, and the
5193 return code was incorrect.
5195 PH/31 Missing "BOOL" in function definition in filtertest.c.
5197 PH/32 Applied Sieve patches from the maintainer.
5199 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5201 PH/33 Added "verify = not_blind".
5203 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5204 Local/Makefile (with some defaults set). These are used in built scripts
5205 such as exicyclog, but they have never been used in the exim_install
5206 script (though there are many overriding facilities there). I have
5207 arranged that the exim_install script now takes note of these two
5210 PH/35 Installed configuration files for Dragonfly.
5212 PH/36 When a locally submitted message by a trusted user did not contain a
5213 From: header, and the sender address was obtained from -f or from an SMTP
5214 MAIL command, and the trusted user did not use -F to supply a sender
5215 name, $originator_name was incorrectly used when constructing a From:
5216 header. Furthermore, $originator_name was used for submission mode
5217 messages from external hosts without From: headers in a similar way,
5218 which is clearly wrong.
5220 PH/37 Added control=suppress_local_fixups.
5222 PH/38 When log_selector = +received_sender was set, and the addition of the
5223 sender made the log line's construction buffer exactly full, or one byte
5224 less than full, an overflow happened when the terminating "\n" was
5227 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5228 when the result of a list match is failure because a DNS lookup failed.
5230 PH/40 RM_COMMAND is now used in the building process.
5232 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5233 the "build-* directories that it finds.
5235 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5236 address in a domain literal was a prefix of an interface address.
5238 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5239 when verifying a sender address, unless rewrite_headers is false.
5241 PH/44 Wrote a long comment about why errors_to addresses are verified as
5242 recipients, not senders.
5244 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5245 the ratelimit ACL was added.
5247 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5249 PH/46 Added extra information about PostgreSQL errors to the error string.
5251 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5252 avoiding the need to install special zones in a real server. This is
5253 backwards compatible; if it can't find the fake resolver, it drops back.
5254 Thus, both old and new test suites can be run.
5256 TF/02 Added util/ratelimit.pl
5258 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5259 clock is set back in time.
5261 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5262 Candler <B.Candler@pobox.com>.
5264 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5265 OK for addresses that are the result of redirections.
5267 PH/48 A number of further additions for the benefit of the new test suite,
5268 including a fake gethostbyname() that interfaces to the fake DNS resolver
5271 TF/06 The fix for widen_domains has also been applied to qualify_single and
5272 search_parents which are the other dnslookup options that can cause
5275 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5278 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5280 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5281 service" if it encountered an MX record with an empty target hostname.
5282 The message is now "an MX or SRV record indicated no SMTP service".
5284 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5285 if the DNS of the sending site is misconfigured. This is quite a
5286 common situation. This change restores the behaviour of treating a
5287 helo verification defer as a failure.
5289 PH/51 If self=fail was set on a router, the bounce message did not include the
5290 actual error message.
5296 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5298 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5299 command itself ended in error, the underlying message about the transport
5300 filter timeout was being overwritten with the pipe command error. Now the
5301 underlying error message should be appended to the second error message.
5303 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5305 PH/02 Reduce the amount of output that "make" produces by default. Full output
5306 can still be requested.
5308 PH/03 The warning log line about a condition test deferring for a "warn" verb
5309 was being output only once per connection, rather than after each
5310 occurrence (because it was using the same function as for successful
5311 "warn" verbs). This seems wrong, so I have changed it.
5313 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5314 it should not have, which might have caused a crash in the right
5315 circumstances, but probably never did.
5317 PH/04 Installed a modified version of Tony Finch's patch to make submission
5318 mode fix the return path as well as the Sender: header line, and to
5319 add a /name= option so that you can make the user's friendly name appear
5322 TF/03 Added the control = fakedefer ACL modifier.
5324 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5325 Mark Lowes for thorough testing.
5327 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5329 TK/03 Merged latest SRS patch from Miles Wilton.
5331 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5332 with the definition in sysexits.h (which is #included earlier).
5333 Fortunately, Exim does not actually use EX_OK. The code used to try to
5334 preserve the sysexits.h value, by assuming that macro definitions were
5335 scanned for macro replacements. I have been disabused of this notion,
5336 so now the code just undefines EX_OK before #including unistd.h.
5338 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5339 in the smtp transport. When a block could not be written in a single
5340 write() function, the timeout was being re-applied to each part-write.
5341 This seems wrong - if the receiver was accepting one byte at a time it
5342 would take for ever. The timeout is now adjusted when this happens. It
5343 doesn't have to be particularly precise.
5345 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5346 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5348 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5349 without a domain if the check to <postmaster@domain> fails.
5351 SC/01 Eximstats: added -xls and the ability to specify output files
5352 (patch written by Frank Heydlauf).
5354 SC/02 Eximstats: use FileHandles for outputting results.
5356 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5358 SC/04 Eximstats: fixed display of large numbers with -nvr option
5360 SC/05 Eximstats: fixed merging of reports with empty tables.
5362 SC/06 Eximstats: added the -include_original_destination flag
5364 SC/07 Eximstats: removed tabs and trailing whitespace.
5366 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5368 TK/06 MBOX spool code: Add real "From " MBOX separator line
5369 so the .eml file is really in mbox format (even though
5370 most programs do not really care). Patch from Alex Miller.
5372 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5373 The latter is generated from $received_to and is only set if the
5374 message has one envelope recipient. SA can use these headers,
5375 obviously out-of-the-box. Patch from Alex Miller.
5377 PH/08 The ${def test on a variable was returning false if the variable's
5378 value was "0", contrary to what the specification has always said!
5379 The result should be true unless the variable is empty.
5381 PH/09 The syntax error of a character other than { following "${if
5382 def:variable_name" (after optional whitespace) was not being diagnosed.
5383 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5384 accidental colon was present, for example, could give incorrect results.
5386 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5387 result is used (not including appendfile, where other changes are about
5390 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5391 This involved changing a lot of size variables from int to off_t. It
5392 should work with maildirs and everything.
5394 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5395 spamd dying while we are connected to it.
5397 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5400 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5401 function for BDB 4.3.
5403 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5405 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5406 the sources for PCRE have been re-arranged and more files are now
5409 PH/15 The code I had for printing potentially long long variables in PH/11
5410 above was not the best (it lost precision). The length of off_t variables
5411 is now inspected at build time, and an appropriate printing format (%ld
5412 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5413 to be "long long int" or "long int". This is needed for the internal
5414 formatting function string_vformat().
5416 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5417 the configuration file to be ":syslog", then the script "guesses" where
5418 the logs files are, rather than using the compiled in default. In our
5419 case the guess is not the same as the compiled default, so the script
5420 suddenly stopped working when I started to use syslog. The patch checks
5421 to see if log_file_path is "". If so, it attempts to read it from exim
5422 with no configuration file to get the compiled in version, before it
5423 falls back to the previous guessing code."
5425 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5426 implementing BATV in an Exim configuration. See NewStuff for the gory
5429 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5430 Makefile that are specific to HP-UX.
5432 PH/18 If the "use_postmaster" option was set for a recipient callout together
5433 with the "random" option, the postmaster address was used as the MAIL
5434 FROM address for the random test, but not for the subsequent recipient
5435 test. It is now used for both.
5437 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5438 patch removes a few documentation additions to RFC 3028, because the
5439 latest draft now contains them. It adds the new en;ascii-case comparator
5440 and a new error check for 8bit text in MIME parts. Comparator and
5441 require names are now matched exactly. I enabled the subaddress
5442 extension, but it is not well tested yet (read: it works for me)."
5444 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5445 rework some of the code of TK/09 above to avoid the hardwired use of
5446 "%lld" and "long long". Replaced the call to snprintf() with a call to
5449 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5450 records point to non-existent hosts", "retry timeout exceeded", and
5451 "retry time not reached for any host after a long failure period".
5453 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5454 experimental DomainKeys support:
5456 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5457 (2) On an error such as an illegally used "control", the wrong name for
5458 the control was given.
5460 These problems did NOT occur unless DomainKeys support was compiled.
5462 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5464 PH/24 Added ${if match_ip condition.
5466 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5467 databases so that it will be absolutely obvious if a crash occurs in the
5468 DB library. This is a regular occurrence (often caused by mis-matched
5471 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5472 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5473 picked up on a user's system that detects such things. There doesn't seem
5474 to be a gcc warning option for this - only an attribute that has to be
5475 put on the function's prototype. It seems that in Fedora Core 4 they have
5476 set this on a number of new functions. No doubt there will be more in due
5479 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5480 specify a transport. However, if an address that was verified by such a
5481 router was the subject of a callout, Exim crashed because it tried to
5482 read the rcpt_include_affixes from the non-existent transport. Now it
5483 just assumes that the setting of that option is false. This bug was
5484 introduced by 4.51/PH/31.
5486 PH/28 Changed -d+all to exclude +memory, because that information is very
5487 rarely of interest, but it makes the output a lot bigger. People tend to
5488 do -d+all out of habit.
5490 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5491 code in os-type was giving problems when libc.so lives in lib64, like on
5494 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5495 aren't the modern standard, and it seems that some systems' include files
5496 don't always have them. Exim was already checking for some of the newer
5497 ones like T_AAAA, and defining it itself. I've added checks for all the
5498 record types that Exim uses.
5500 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5501 not automatically generating a new one, as it is supposed to. This
5502 prevented TLS from working. If the file did exist, but contained invalid
5503 data, a new version was generated, as expected. It was only the case of a
5504 non-existent file that was broken.
5506 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5507 with a change in libdomainkeys > 0.64.
5509 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5510 from DNS. If the selector record carries the flag, it now has
5511 precedence over the domain-wide flag.
5513 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5515 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5516 the use of a port name that isn't defined in /etc/services) occurred, the
5517 message was deferred as in a normal delivery, and thus remained on the
5518 spool, instead of being failed because of the mua_wrapper setting. This
5519 is now fixed, and I tidied up some of the mua_wrapper messages at the
5522 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5523 the messages in a hash of arrays rather than using individual hashes.
5524 This is a bit cleaner and results in dramatic memory savings, albeit
5525 at a slight CPU cost.
5527 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5528 as requested by Marc Sherman.
5530 SC/10 Eximstats: added histograms for user specified patterns as requested
5533 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5535 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5536 fopen() in the content-scanning modules that did not already have it.
5542 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5543 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5545 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5547 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5549 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5550 to test Sieve filters that use "vacation".
5552 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5553 that changes the way the GnuTLS parameters are stored in the cache file.
5554 The new format can be generated externally. For backward compatibility,
5555 if the data in the cache doesn't make sense, Exim assumes it has read an
5556 old-format file, and it generates new data and writes a new file. This
5557 means that you can't go back to an older release without removing the
5560 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5561 work if there are any delivery delays because "one_time" forces the
5562 parent to be marked "delivered", so its unseen clone is never tried
5563 again. For this reason, Exim now forbids the simultaneous setting of
5566 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5567 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5568 entry if you want to know the details. The fix, however, neglected to
5569 consider the case when local delivery batching is involved. The test for
5570 "previously delivered" was not happening when checking to see if an
5571 address could be batched with a previous (undelivered) one; under
5572 certain circumstances this could lead to multiple deliveries to the same
5575 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5576 in its include files, and this causes problems building Exim.
5578 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5579 header_syntax) but Exim was just ignoring anything given after a slash.
5580 In particular, this caused confusion with an attempt to use "verify =
5581 reverse_host_lookup/defer_ok". An error is now given when options are
5582 supplied for verify items that do not have them. (Maybe reverse_host_
5583 lookup should have a defer_ok option, but that's a different point.)
5585 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5586 defined by RFC 821) to 2048, because there were problems with some AUTH
5587 commands, and RFC 1869 says the size should be increased for extended
5588 SMTP commands that take arguments.
5590 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5593 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5594 "unknown" error; now it says that the functionality isn't in the binary.
5596 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5597 an address' error message when a string expansion fails (syntax or
5598 whatever). Otherwise the password may appear in the log. Following change
5599 PH/42 below, there is no longer a chance of it appearing in a bounce
5602 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5604 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5605 of its addresses was ever tried. (Bugzilla bug #2.)
5607 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5608 the result incorrectly in the debug output. (It correctly added a newline
5609 to what was transported.)
5611 TF/01 Added $received_time.
5613 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5614 commented out examples of how to interface to a virus scanner and to
5615 SpamAssassin. Also added commented examples of av_scanner and
5616 spamd_address settings.
5618 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5619 and controls are allowed in which ACLs. There were a couple of minor
5620 errors. Some of the entries in the conditions table (which is a table of
5621 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5622 negation of where the condition IS allowed.
5624 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5626 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5627 header file does not have a version number, so I've had to invent a new
5628 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5629 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5630 radiusclient), but it was contributed by a Radius user.
5632 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5633 files or directories, for OpenSSL.
5635 PH/19 When an Exim process that is running as root has to create an Exim log
5636 file, it does so in a subprocess that runs as exim:exim so as to get the
5637 ownership right at creation (otherwise, other Exim processes might see
5638 the file with the wrong ownership). There was no test for failure of this
5639 fork() call, which would lead to the process getting stuck as it waited
5640 for a non-existent subprocess. Forks do occasionally fail when resources
5641 run out. I reviewed all the other calls to fork(); they all seem to check
5644 PH/20 When checking for unexpected SMTP input at connect time (before writing
5645 the banner), Exim was not dealing correctly with a non-positive return
5646 from the read() function. If the client had disconnected by this time,
5647 the result was a log entry for a synchronization error with an empty
5648 string after "input=" when read() returned zero. If read() returned -1
5649 (an event I could not check), uninitialized data bytes were printed.
5650 There were reports of junk text (parts of files, etc) appearing after
5653 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5655 PH/22 Added support for macro redefinition, and (re)definition in between
5656 driver and ACL definitions.
5658 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5659 forgetting to use the resulting value; it was using the unexpanded value.
5661 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5662 hadn't been configured. The fix is from Juergen Kreileder, who
5663 understands it better than I do:
5665 "Here's what I see happening with three configured cyrus_sasl
5666 authenticators configured (plain, login, cram-md5):
5668 On startup auth_cyrus_sasl_init() gets called for each of these.
5669 This means three calls to sasl_listmech() without a specified mech_list.
5670 => SASL tests which mechs of all available mechs actually work
5671 => three warnings about OTP not working
5672 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5674 With the patch, sasl_listmech() also gets called three times. But now
5675 SASL's mech_list option is set to the server_mech specified in the the
5676 authenticator. Or in other words, the answer from sasl_listmech()
5677 gets limited to just the mech you're testing for (which is different
5679 => the return list contains just 'plain' or 'login', 'cram-md5' or
5680 nothing depending on the value of ob->server_mech.
5682 I've just tested the patch: Authentication still works fine,
5683 unavailable mechs specified in the exim configuration are still
5684 caught, and the auth.log warnings about OTP are gone."
5686 PH/25 When debugging is enabled, the contents of the command line are added
5687 to the debugging output, even when log_selector=+arguments is not
5690 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5691 answer is "GNU", and only if the return is "GNU/something" is the answer
5694 PH/27 $acl_verify_message is now set immediately after the failure of a
5695 verification in an ACL, and so is available in subsequent modifiers. In
5696 particular, the message can be preserved by coding like this:
5698 warn !verify = sender
5699 set acl_m0 = $acl_verify_message
5701 Previously, $acl_verify_message was set only while expanding "message"
5702 and "log_message" when a very denied access.
5704 PH/28 Modified OS/os.c-Linux with
5706 -#ifndef OS_LOAD_AVERAGE
5707 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5709 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5710 nomenclature these days.)
5712 PH/29 Installed patch from the Sieve maintainer that adds the options
5713 sieve_useraddress and sieve_subaddress to the redirect router.
5715 PH/30 In these circumstances:
5716 . Two addresses routed to the same list of hosts;
5717 . First host does not offer TLS;
5718 . First host accepts first address;
5719 . First host gives temporary error to second address;
5720 . Second host offers TLS and a TLS session is established;
5721 . Second host accepts second address.
5722 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5723 and peerdn, if requested) that were in fact used only for the second
5726 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5727 attention to any local part prefix or suffix that was matched by the
5728 router that accepted the address. It now behaves in the same way as it
5729 does for delivery: the affixes are removed from the local part unless
5730 rcpt_include_affixes is set on the transport.
5732 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5733 timeout during the DATA phase of an incoming message.
5735 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5736 applied a patch sanctioned by the Sieve maintainer.
5738 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5739 the uid or gid is negative. A case of a negative gid caused this to be
5740 noticed. The fix allows for either to be negative.
5742 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5743 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5746 PH/36 The change PH/12 above was broken. Fixed it.
5748 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5749 the grounds that routing the same address twice would always produce the
5750 same answer. This might have been true once, but it is certainly no
5751 longer true now. Routing a child address may depend on the previous
5752 routing that produced that child. Some complicated redirection strategies
5753 went wrong when messages had multiple recipients, and made Exim's
5754 behaviour dependent on the order in which the addresses were given.
5756 I have moved the duplicate checking until after the routing is complete.
5757 Exim scans the addresses that are assigned to local and remote
5758 transports, and removes any duplicates. This means that more work will be
5759 done, as duplicates will always all be routed, but duplicates are
5760 presumably rare, so I don't expect this is of any significance.
5762 For deliveries to pipes, files, and autoreplies, the duplicate checking
5763 still happens during the routing process, since they are not going to be
5766 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5767 It corrects a timeout issue with spamd. This is Ian's comment: "The
5768 background is that sometimes spamd either never reads data from a
5769 connection it has accepted, or it never writes response data. The exiscan
5770 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5771 blindly assumes that writes won't block so it may never time out."
5773 PH/39 Allow G after quota size as well as K and M.
5775 PH/40 The value set for $authenticated_id in an authenticator may not contain
5776 binary zeroes or newlines because the value is written to log lines and
5777 to spool files. There was no check on this. Now the value is run through
5778 the string_printing() function so that such characters are converted to
5779 printable escape sequences.
5781 PH/41 $message_linecount is a new variable that contains the total number of
5782 lines in the message. Compare $body_linecount, which is the count for the
5785 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5786 bounce and delay warning messages, except in certain special cases, which
5789 (a) An SMTP error message from a remote host;
5790 (b) A message specified in a :fail: redirection;
5791 (c) A message specified in a "fail" command in a system filter;
5792 (d) A message specified in a FAIL return from the queryprogram router;
5793 (e) A message specified by the cannot_route_message router option.
5795 In these cases only, Exim does include the error details in bounce and
5796 warning messages. There are also a few cases where bland messages such
5797 as "unrouteable address" or "local delivery error" are given.
5799 PH/43 $value is now also set for the "else" part of a ${run expansion.
5801 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5802 being worked on, but at least Exim now implements the latest version to
5805 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5806 process to complete was treated as a delivery failure, a timeout while
5807 writing the message to the pipe was logged, but erroneously treated as a
5808 successful delivery. Such timeouts include transport filter timeouts. For
5809 consistency with the overall process timeout, these timeouts are now
5810 treated as errors, giving rise to delivery failures by default. However,
5811 there is now a new Boolean option for the pipe transport called
5812 timeout_defer, which, if set TRUE, converts the failures into defers for
5813 both kinds of timeout. A transport filter timeout is now identified in
5816 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5817 systems where "make" and "gmake" are different, calling "gmake" at top
5818 level broke things. I've arranged for the value of $(MAKE) to be passed
5819 from the Makefile to this script so that it can call the same version of
5823 A note about Exim versions 4.44 and 4.50
5824 ----------------------------------------
5826 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5827 changes of various kinds. As a consequence, a big documentation update was
5828 needed. This delayed the release for rather longer than seemed good, especially
5829 in the light of a couple of (minor) security issues. Therefore, the changes
5830 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5831 release. So 4.44 and 4.50 are in effect two different branches that both start
5834 I have left the 4.50 change log unchanged; it contains all the changes since
5835 4.43. The change log for 4.44 is below; many of its items are identical to
5836 those for 4.50. This seems to be the most sensible way to preserve the
5837 historical information.
5843 1. Minor wording change to the doc/README.SIEVE file.
5845 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5846 computation of the current number of files was incorrect.
5848 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5849 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5852 4. Give more explanation in the error message when the command for a transport
5853 filter fails to execute.
5855 5. There are several places where Exim runs a non-Exim command in a
5856 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5857 was being done only for the command run by the queryprogram router. It is
5858 now done for all such subprocesses. The other cases are: ${run, transport
5859 filters, and the commands run by the lmtp and pipe transports.
5861 6. Added CONFIGURE_GROUP build-time option.
5863 7. Some older OS have a limit of 256 on the maximum number of file
5864 descriptors. Exim was using setrlimit() to set 1000 as a large value
5865 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5866 systems. I've change it so that if it can't get 1000, it tries for 256.
5868 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5869 was an oversight, and furthermore, ever since the addition of extra
5870 controls (e.g. 4.43/32), the checks on when to allow different forms of
5871 "control" were broken. There should now be diagnostics for all cases when a
5872 control that does not make sense is encountered.
5874 9. Added the /retain_sender option to "control=submission".
5876 10. $recipients is now available in the predata ACL (oversight).
5878 11. Tidy the search cache before the fork to do a delivery from a message
5879 received from the command line. Otherwise the child will trigger a lookup
5880 failure and thereby defer the delivery if it tries to use (for example) a
5881 cached ldap connection that the parent has called unbind on.
5883 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5884 of $address_data from the recipient verification was clobbered by the
5885 sender verification.
5887 13. The value of address_data from a sender verification is now available in
5888 $sender_address_data in subsequent conditions in the ACL statement.
5890 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5892 15. Added a new option "connect=<time>" to callout options, to set a different
5895 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5896 was its contents. (It was OK if the option was not defined at all.)
5898 17. A "Completed" log line is now written for messages that are removed from
5899 the spool by the -Mrm option.
5901 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5902 information about exactly what failed.
5904 19. Added -dd to debug only the daemon process.
5906 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5907 handles timeouts, both on the server side and network timeouts. Renamed the
5908 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5910 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5911 It is now set to "smtps".
5913 22. $host_address is now set to the target address during the checking of
5914 ignore_target_hosts.
5916 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5917 being passed; this would have caused $sender_host_name to have been used if
5918 matching the list had actually called for a host name (not very likely,
5919 since this list is usually IP addresses). A host name is now passed as
5922 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5923 code that specifies a non-restarting handler (typically sigaction() in
5924 modern systems) in an attempt to fix a rare and obscure crash bug.
5926 25. Narrowed the window for a race in the daemon that could cause it to ignore
5927 SIGCHLD signals. This is not a major problem, because they are used only to
5928 wake it up if nothing else does.
5930 26. A malformed maildirsize file could cause Exim to calculate negative values
5931 for the mailbox size or file count. Odd effects could occur as a result.
5932 The maildirsize information is now recalculated if the size or filecount
5935 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5936 support for a long time. Removed HAVE_SYS_VFS_H.
5938 28. Installed the latest version of exipick from John Jetmore.
5940 29. In an address list, if the pattern was not a regular expression, an empty
5941 subject address (from a bounce message) matched only if the pattern was an
5942 empty string. Non-empty patterns were not even tested. This was the wrong
5943 because it is perfectly reasonable to use an empty address as part of a
5944 database query. An empty address is now tested by patterns that are
5945 lookups. However, all the other forms of pattern expect the subject to
5946 contain a local part and a domain, and therefore, for them, an empty
5947 address still always fails if the pattern is not itself empty.
5949 30. Exim went into a mad DNS loop when attempting to do a callout where the
5950 host was specified on an smtp transport, and looking it up yielded more
5951 than one IP address.
5953 31. Re-factored the code for checking spool and log partition space into a
5954 function that finds that data and another that does the check. The former
5955 is then used to implement four new variables: $spool_space, $log_space,
5956 $spool_inodes, and $log_inodes.
5958 32. The RFC2047 encoding function was originally intended for short strings
5959 such as real names; it was not keeping to the 75-character limit for
5960 encoded words that the RFC imposes. It now respects the limit, and
5961 generates multiple encoded words if necessary. To be on the safe side, I
5962 have increased the buffer size for the ${rfc2047: expansion operator from
5965 33. It is now permitted to omit both strings after an "if" condition; if the
5966 condition is true, the result is "true". As before, when the second string
5967 is omitted, a false condition yields an empty string. This makes it less
5968 cumbersome to write custom ACL and router conditions.
5970 34. Failure to deliver a bounce message always caused it to be frozen, even if
5971 there was an errors_to setting on the router. The errors_to setting is now
5974 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5975 canonical form (fully expanded) before being placed in
5976 $sender_host_address.
5978 36. The table in the code that translates DNS record types into text (T_A to
5979 "A" for instance) was missing entries for NS and CNAME. It is just possible
5980 that this could have caused confusion if both these types were looked up
5981 for the same domain, because the text type is used as part of Exim's
5982 per-process caching. But the chance of anyone hitting this buglet seems
5985 37. The dnsdb lookup has been extended in a number of ways.
5987 (1) There is a new type, "zns", which walks up the domain tree until it
5988 finds some nameserver records. It should be used with care.
5990 (2) There is a new type, "mxh", which is like "mx" except that it returns
5991 just the host names, not the priorities.
5993 (3) It is now possible to give a list of domains (or IP addresses) to be
5994 looked up. The behaviour when one of the lookups defers can be
5995 controlled by a keyword.
5997 (4) It is now possible to specify the separator character for use when
5998 multiple records are returned.
6000 38. The dnslists ACL condition has been extended: it is now possible to supply
6001 a list of IP addresses and/or domains to be looked up in a particular DNS
6004 39. Added log_selector=+queue_time_overall.
6006 40. When running the queue in the test harness, wait just a tad after forking a
6007 delivery process, to get repeatability of debugging output.
6009 41. Include certificate and key file names in error message when GnuTLS fails
6010 to set them up, because the GnuTLS error message doesn't include the name
6011 of the failing file when there is a problem reading it.
6013 42. Allow both -bf and -bF in the same test run.
6015 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6017 44. The "Exiscan patch" is now merged into the mainline Exim source.
6019 45. Sometimes the final signoff response after QUIT could fail to get
6020 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6021 < 0 before doing a fflush(). This bug looks as though it goes back to the
6022 introduction of TLS in release 3.20, but "sometimes" must have been rare
6023 because the tests only now provoked it.
6025 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6026 (this can affect the format of dates).
6028 47. exim_tidydb, when checking for the continued existence of a message for
6029 which it has found a message-specific retry record, was not finding
6030 messages that were in split spool directories. Consequently, it was
6031 deleting retry records that should have stayed in existence.
6033 48. Steve fixed some bugs in eximstats.
6035 49. The SPA authentication driver was not abandoning authentication and moving
6036 on to the next authenticator when an expansion was forced to fail,
6037 contradicting the general specification for all authenticators. Instead it
6038 was generating a temporary error. It now behaves as specified.
6040 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6041 (the order specifies the preference for clients). The order is now AES256,
6042 AES128, 3DES, ARCFOUR128.
6044 51. Small patch to Sieve code - explicitly set From: when generating an
6047 52. Exim crashed if a remote delivery caused a very long error message to be
6048 recorded - for instance if somebody sent an entire SpamAssassin report back
6049 as a large number of 550 error lines. This bug was coincidentally fixed by
6050 increasing the size of one of Exim's internal buffers (big_buffer) that
6051 happened as part of the Exiscan merge. However, to be on the safe side, I
6052 have made the code more robust (and fixed the comments that describe what
6055 53. Now that there can be additional text after "Completed" in log lines (if
6056 the queue_time_overall log selector is set), a one-byte patch to exigrep
6057 was needed to allow it to recognize "Completed" as not the last thing in
6060 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6061 patch that reportedly fixes this has been added. I am not expert enough to
6062 create a test for it. This is what the patch creator wrote:
6064 "I found a little strange behaviour of ldap code when working with
6065 Windows 2003 AD Domain, where users was placed in more than one
6066 Organization Units. When I tried to give exim partial DN, the exit code
6067 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6068 But simultaneously result of request was absolutely normal ldap result,
6069 so I produce this patch..."
6071 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6072 REFERENCE, so I have modified the code to exclude the patch when that macro
6075 55. Some experimental protocols are using DNS PTR records for new purposes. The
6076 keys for these records are domain names, not reversed IP addresses. The
6077 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6078 leaves it alone. Component reversal etc. now happens only for IP addresses.
6081 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6083 57. Double the size of the debug message buffer (to 2048) so that more of very
6084 long debug lines gets shown.
6086 58. The exicyclog utility now does better if the number of log files to keep
6087 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6089 59. Two changes related to the smtp_active_hostname option:
6091 (1) $smtp_active_hostname is now available as a variable.
6092 (2) The default for smtp_banner uses $smtp_active_hostname instead
6093 of $primary_hostname.
6095 60. The host_aton() function is supposed to be passed a string that is known
6096 to be a valid IP address. However, in the case of IPv6 addresses, it was
6097 not checking this. This is a hostage to fortune. Exim now panics and dies
6098 if the condition is not met. A case was found where this could be provoked
6099 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6100 components; fortuitously, this particular loophole had already been fixed
6101 by change 4.50/55 above.
6103 If there are any other similar loopholes, the new check in host_aton()
6104 itself should stop them being exploited. The report I received stated that
6105 data on the command line could provoke the exploit when Exim was running as
6106 exim, but did not say which command line option was involved. All I could
6107 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6108 running as the user.
6111 61. There was a buffer overflow vulnerability in the SPA authentication code
6112 (which came originally from the Samba project). I have added a test to the
6113 spa_base64_to_bits() function which I hope fixes it.
6116 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6117 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6119 63. The daemon start-up calls getloadavg() while still root for those OS that
6120 need the first call to be done as root, but it missed one case: when
6121 deliver_queue_load_max is set with deliver_drop_privilege. This is
6122 necessary for the benefit of the queue runner, because there is no re-exec
6123 when deliver_drop_privilege is set.
6125 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6126 This has been fixed.
6128 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6129 was in use, was not putting the data itself into the right store pool;
6130 consequently, it could be overwritten for a subsequent message in the same
6131 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6134 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6136 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6137 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6138 the function were treating the return as a boolean value, which happened to
6139 work because 0=false and not-0=true, but is not correct code.
6141 68. The host_aton() function was not handling scoped IPv6 addresses (those
6142 with, for example, "%eth0" on the end) correctly.
6144 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6145 negated items (that is, ~something) in unsigned ints. Some compilers
6146 apparently mutter when there is no cast.
6148 70. If an address verification called from an ACL failed, and did not produce a
6149 user-specific message (i.e. there was only a "system" message), nothing was
6150 put in $acl_verify_message. In this situation, it now puts the system
6153 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6154 session; change 4.31/43 added the unwanted input to the log line - except
6155 that it did not do this in the start of session case. It now does.
6157 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6158 This is wrong and can cause the other end to generate a synchronization
6159 error if it is another Exim or anything else that does the synchronization
6160 check. A QUIT command is no longer sent after a timeout.
6162 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6163 during host lookups.
6165 74. The defer_ok option of callout verification was not working if it was used
6166 when verifying addresses in header lines, that is, for this case:
6168 verify = header_sender/callout=defer_ok
6170 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6171 those file descriptors could be used for SMTP connections. If anything
6172 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6173 could be sent to the SMTP client, causing chaos. The daemon now opens
6174 stdin, stdout, and stderr to /dev/null when it puts itself into the
6177 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6178 log by default. The user can override this with suitable Perl magic.
6180 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6181 the log message when discard triggers, was not working for the DATA ACL or
6182 for the non-SMTP ACL.
6184 78. Error message wording change in sieve.c.
6186 79. If smtp_accept_max_per_host was set, the number of connections could be
6187 restricted to fewer than expected, because the daemon was trying to set up
6188 a new connection before checking whether the processes handling previous
6189 connections had finished. The check for completed processes is now done
6190 earlier. On busy systems, this bug wouldn't be noticed because something
6191 else would have woken the daemon, and it would have reaped the completed
6194 80. If a message was submitted locally by a user whose login name contained one
6195 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6196 It caused a spool format error. I have fixed the spool reading code. A
6197 related problem was that the "from" clause in the Received: line became
6198 illegal because of the space(s). It is now covered by ${quote_local_part.
6200 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6203 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6205 83. Patch from Sieve maintainer to fix unterminated string problem in
6206 "vacation" handling.
6208 84. Some minor changes to the Linux configuration files to help with other
6209 OS variants using glibc.
6211 85. One more patch for Sieve to update vacation handling to latest spec.
6214 ----------------------------------------------------
6215 See the note above about the 4.44 and 4.50 releases.
6216 ----------------------------------------------------
6222 1. Change 4.43/35 introduced a bug that caused file counts to be
6223 incorrectly computed when quota_filecount was set in an appendfile
6226 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6227 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6230 3. Give more explanation in the error message when the command for a transport
6231 filter fails to execute.
6233 4. There are several places where Exim runs a non-Exim command in a
6234 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6235 was being done only for the command run by the queryprogram router. It is
6236 now done for all such subprocesses. The other cases are: ${run, transport
6237 filters, and the commands run by the lmtp and pipe transports.
6239 5. Some older OS have a limit of 256 on the maximum number of file
6240 descriptors. Exim was using setrlimit() to set 1000 as a large value
6241 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6242 systems. I've change it so that if it can't get 1000, it tries for 256.
6244 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6245 was an oversight, and furthermore, ever since the addition of extra
6246 controls (e.g. 4.43/32), the checks on when to allow different forms of
6247 "control" were broken. There should now be diagnostics for all cases when a
6248 control that does not make sense is encountered.
6250 7. $recipients is now available in the predata ACL (oversight).
6252 8. Tidy the search cache before the fork to do a delivery from a message
6253 received from the command line. Otherwise the child will trigger a lookup
6254 failure and thereby defer the delivery if it tries to use (for example) a
6255 cached ldap connection that the parent has called unbind on.
6257 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6258 of $address_data from the recipient verification was clobbered by the
6259 sender verification.
6261 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6262 was its contents. (It was OK if the option was not defined at all.)
6264 11. A "Completed" log line is now written for messages that are removed from
6265 the spool by the -Mrm option.
6267 12. $host_address is now set to the target address during the checking of
6268 ignore_target_hosts.
6270 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6271 being passed; this would have caused $sender_host_name to have been used if
6272 matching the list had actually called for a host name (not very likely,
6273 since this list is usually IP addresses). A host name is now passed as
6276 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6277 code that specifies a non-restarting handler (typically sigaction() in
6278 modern systems) in an attempt to fix a rare and obscure crash bug.
6280 15. Narrowed the window for a race in the daemon that could cause it to ignore
6281 SIGCHLD signals. This is not a major problem, because they are used only to
6282 wake it up if nothing else does.
6284 16. A malformed maildirsize file could cause Exim to calculate negative values
6285 for the mailbox size or file count. Odd effects could occur as a result.
6286 The maildirsize information is now recalculated if the size or filecount
6289 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6290 support for a long time. Removed HAVE_SYS_VFS_H.
6292 18. Updated exipick to current release from John Jetmore.
6294 19. Allow an empty sender to be matched against a lookup in an address list.
6295 Previously the only cases considered were a regular expression, or an
6298 20. Exim went into a mad DNS lookup loop when doing a callout where the
6299 host was specified on the transport, if the DNS lookup yielded more than
6302 21. The RFC2047 encoding function was originally intended for short strings
6303 such as real names; it was not keeping to the 75-character limit for
6304 encoded words that the RFC imposes. It now respects the limit, and
6305 generates multiple encoded words if necessary. To be on the safe side, I
6306 have increased the buffer size for the ${rfc2047: expansion operator from
6309 22. Failure to deliver a bounce message always caused it to be frozen, even if
6310 there was an errors_to setting on the router. The errors_to setting is now
6313 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6314 canonical form (fully expanded) before being placed in
6315 $sender_host_address.
6317 24. Updated eximstats to version 1.33
6319 25. Include certificate and key file names in error message when GnuTLS fails
6320 to set them up, because the GnuTLS error message doesn't include the name
6321 of the failing file when there is a problem reading it.
6323 26. Expand error message when OpenSSL has problems setting up cert/key files.
6326 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6327 (this can affect the format of dates).
6329 28. exim_tidydb, when checking for the continued existence of a message for
6330 which it has found a message-specific retry record, was not finding
6331 messages that were in split spool directories. Consequently, it was
6332 deleting retry records that should have stayed in existence.
6334 29. eximstats updated to version 1.35
6335 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6336 1.35 - bugfix such that pie charts by volume are generated correctly
6338 30. The SPA authentication driver was not abandoning authentication and moving
6339 on to the next authenticator when an expansion was forced to fail,
6340 contradicting the general specification for all authenticators. Instead it
6341 was generating a temporary error. It now behaves as specified.
6343 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6344 (the order specifies the preference for clients). The order is now AES256,
6345 AES128, 3DES, ARCFOUR128.
6347 31. Small patch to Sieve code - explicitly set From: when generating an
6350 32. Exim crashed if a remote delivery caused a very long error message to be
6351 recorded - for instance if somebody sent an entire SpamAssassin report back
6352 as a large number of 550 error lines. This bug was coincidentally fixed by
6353 increasing the size of one of Exim's internal buffers (big_buffer) that
6354 happened as part of the Exiscan merge. However, to be on the safe side, I
6355 have made the code more robust (and fixed the comments that describe what
6358 33. Some experimental protocols are using DNS PTR records for new purposes. The
6359 keys for these records are domain names, not reversed IP addresses. The
6360 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6361 leaves it alone. Component reversal etc. now happens only for IP addresses.
6364 34. The host_aton() function is supposed to be passed a string that is known
6365 to be a valid IP address. However, in the case of IPv6 addresses, it was
6366 not checking this. This is a hostage to fortune. Exim now panics and dies
6367 if the condition is not met. A case was found where this could be provoked
6368 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6369 components; fortuitously, this particular loophole had already been fixed
6370 by change 4.50/55 or 4.44/33 above.
6372 If there are any other similar loopholes, the new check in host_aton()
6373 itself should stop them being exploited. The report I received stated that
6374 data on the command line could provoke the exploit when Exim was running as
6375 exim, but did not say which command line option was involved. All I could
6376 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6377 running as the user.
6380 35. There was a buffer overflow vulnerability in the SPA authentication code
6381 (which came originally from the Samba project). I have added a test to the
6382 spa_base64_to_bits() function which I hope fixes it.
6385 36. The daemon start-up calls getloadavg() while still root for those OS that
6386 need the first call to be done as root, but it missed one case: when
6387 deliver_queue_load_max is set with deliver_drop_privilege. This is
6388 necessary for the benefit of the queue runner, because there is no re-exec
6389 when deliver_drop_privilege is set.
6391 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6392 was in use, was not putting the data itself into the right store pool;
6393 consequently, it could be overwritten for a subsequent message in the same
6394 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6397 38. Sometimes the final signoff response after QUIT could fail to get
6398 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6399 < 0 before doing a fflush(). This bug looks as though it goes back to the
6400 introduction of TLS in release 3.20, but "sometimes" must have been rare
6401 because the tests only now provoked it.
6407 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6408 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6409 Now it returns an integer. A number of calls were still expecting a T/F
6410 return. Fortuitously, in all cases, the tests worked in OK situations,
6411 which is the norm. However, things would have gone wrong on any write
6412 failures on the smtp file descriptor. This function is used when sending
6413 messages over SMTP and also when doing verify callouts.
6415 2. When Exim is called to do synchronous delivery of a locally submitted
6416 message (the -odf or -odi options), it no longer closes stderr before doing
6419 3. Implemented the mua_wrapper option.
6421 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6423 5. Implemented the functions header_remove(), header_testname(),
6424 header_add_at_position(), and receive_remove_recipient(), and exported them
6427 6. If an ACL "warn" statement specified the addition of headers, Exim already
6428 inserted X-ACL-Warn: at the start if there was no header name. However, it
6429 was not making this test for the second and subsequent header lines if
6430 there were newlines in the string. This meant that an invalid header could
6431 be inserted if Exim was badly configured.
6433 7. Allow an ACL "warn" statement to add header lines at the start or after all
6434 the Received: headers, as well as at the end.
6436 8. Added the rcpt_4xx retry error code.
6438 9. Added postmaster_mailfrom=xxx to callout verification option.
6440 10. Added mailfrom=xxxx to the callout verification option, for verify=
6443 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6444 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6446 12. Inserted some casts to stop certain compilers warning when using pointer
6447 differences as field lengths or precisions in printf-type calls (mostly
6448 affecting debugging statements).
6450 13. Added optional readline() support for -be (dynamically loaded).
6452 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6453 same clock tick as a message's arrival, so that its received time was the
6454 same as the "first fail" time on the retry record, and that message
6455 remained on the queue past the ultimate address timeout, every queue runner
6456 would try a delivery (because it was past the ultimate address timeout) but
6457 after another failure, the ultimate address timeout, which should have then
6458 bounced the address, did not kick in. This was a "< instead of <=" error;
6459 in most cases the first failure would have been in the next clock tick
6460 after the received time, and all would be well.
6462 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6463 being recognized when the domain list was tested by the match_domain
6464 condition in an expansion string.
6466 16. Added the ${str2b64: operator.
6468 17. Exim was always calling setrlimit() to set a large limit for the number of
6469 processes, without checking whether the existing limit was already
6470 adequate. (It did check for the limit on file descriptors.) Furthermore,
6471 errors from getrlimit() and setrlimit() were being ignored. Now they are
6472 logged to the main and panic logs, but Exim does carry on, to try to do its
6473 job under whatever limits there are.
6475 18. Imported PCRE 5.0.
6477 19. Trivial typo in log message " temporarily refused connection" (the leading
6480 20. If the log selector return_path_on_delivery was set and an address was
6481 redirected to /dev/null, the delivery process crashed because it assumed
6482 that a return path would always be set for a "successful" delivery. In this
6483 case, the whole delivery is bypassed as an optimization, and therefore no
6486 21. Internal re-arrangement: the function for sending a challenge and reading
6487 a response while authentication was assuming a zero-terminated challenge
6488 string. It's now changed to take a pointer and a length, to allow for
6489 binary data in such strings.
6491 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6493 23. Exim was not respecting finduser_retries when seeking the login of the
6494 uid under which it was called; it was always trying 10 times. (The default
6495 setting of finduser_retries is zero.) Also, it was sleeping after the final
6496 failure, which is pointless.
6498 24. Implemented tls_on_connect_ports.
6500 25. Implemented acl_smtp_predata.
6502 26. If the domain in control=submission is set empty, Exim assumes that the
6503 authenticated id is a complete email address when it generates From: or
6504 Sender: header lines.
6506 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6507 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6508 chown and chgrp in /bin and hostname in /usr/bin.
6510 28. Exim was keeping the "process log" file open after each use, just as it
6511 does for the main log. This opens the possibility of it remaining open for
6512 long periods when the USR1 signal hits a daemon. Occasional processlog
6513 errors were reported, that could have been caused by this. Anyway, it seems
6514 much more sensible not to leave this file open at all, so that is what now
6517 29. The long-running daemon process does not normally write to the log once it
6518 has entered its main loop, and it closes the log before doing so. This is
6519 so that log files can straightforwardly be renamed and moved. However,
6520 there are a couple of unusual error situations where the daemon does write
6521 log entries, and I had neglected to close the log afterwards.
6523 30. The text of an SMTP error response that was received during a remote
6524 delivery was being truncated at 512 bytes. This is too short for some of
6525 the long messages that one sometimes sees. I've increased the limit to
6528 31. It is now possible to make retry rules that apply only when a message has a
6529 specific sender, in particular, an empty sender.
6531 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6532 it possible to be selective about when SMTP synchronization is enforced.
6534 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6536 32. Implemented hosts_connection_nolog.
6538 33. Added an ACL for QUIT.
6540 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6543 35. Added mailbox_size and mailbox_filecount to appendfile.
6545 36. Added control = no_multiline_responses to ACLs.
6547 37. There was a bug in the logic of the code that waits for the clock to tick
6548 in the case where the clock went backwards by a substantial amount such
6549 that the microsecond fraction of "now" was more than the microsecond
6550 fraction of "then" (but the whole seconds number was less).
6552 38. Added support for the libradius Radius client library this is found on
6553 FreeBSD (previously only the radiusclient library was supported).
6559 1. When certain lookups returned multiple values in the form name=value, the
6560 quoting of the values was not always being done properly. Specifically:
6561 (a) If the value started with a double quote, but contained no whitespace,
6563 (b) If the value contained whitespace other than a space character (i.e.
6564 tabs or newlines or carriage returns) it was not quoted.
6565 This fix has been applied to the mysql and pgsql lookups by writing a
6566 separate quoting function and calling it from the lookup code. The fix
6567 should probably also be applied to nisplus, ibase and oracle lookups, but
6568 since I cannot test any of those, I have not disturbed their existing code.
6570 2. A hit in the callout cache for a specific address caused a log line with no
6571 reason for rejecting RCPT. Now it says "Previous (cached) callout
6572 verification failure".
6574 3. There was an off-by-one bug in the queryprogram router. An over-long
6575 return line was truncated at 256 instead of 255 characters, thereby
6576 overflowing its buffer with the terminating zero. As well as fixing this, I
6577 have increased the buffer size to 1024 (and made a note to document this).
6579 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6580 when Exim is waiting for an SMTP response from a remote server, Exim
6581 restarts its select() call on the socket, thereby resetting its timeout.
6582 This is not a problem when such interrupts are rare. Somebody set up a cron
6583 job to run exiwhat every 2 minutes, which is less than the normal select()
6584 timeout (5 or 10 minutes). This meant that the select() timeout never
6585 kicked in because it was always reset. I have fixed this by comparing the
6586 time when an interrupt arrives with the time at the start of the first call
6587 to select(). If more time than the timeout has elapsed, the interrupt is
6588 treated as a timeout.
6590 5. Some internal re-factoring in preparation for the addition of Sieve
6591 extensions (by MH). In particular, the "personal" test is moved to a
6592 separate function, and given an option for scanning Cc: and Bcc: (which is
6593 not set for Exim filters).
6595 6. When Exim created an email address using the login of the caller as the
6596 local part (e.g. when creating a From: or Sender: header line), it was not
6597 quoting the local part when it contained special characters such as @.
6599 7. Installed new OpenBSD configuration files.
6601 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6602 try to make them clearer.
6604 9. Callout options, other than the timeout value, were being ignored when
6605 verifying sender addresses in header lines. For example, when using
6607 verify = header_sender/callout=no_cache
6609 the cache was (incorrectly) being used.
6611 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6612 only the Cygwin environment.
6614 11. When return_path_on_delivery was set as a log selector, if different remote
6615 addresses in the same message used different return paths and parallel
6616 remote delivery occurred, the wrong values would sometimes be logged.
6617 (Whenever a remote delivery process finished, the return path value from
6618 the most recently started remote delivery process was logged.)
6620 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6621 lines when AUTH and/or TLS are in use. This is the "received protocol"
6622 field. Exim used to use "asmtp" for authenticated SMTP, without any
6623 indication (in the protocol name) for TLS use. Now it follows the RFC and
6624 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6625 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6626 names appear in log lines as well as in Received: header lines.
6628 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6629 extensions, and comparison tests, and to fix some bugs.
6631 14. Changes to the "personal" filter test:
6633 (1) The test was buggy in that it was just doing the equivalent of
6634 "contains" tests on header lines. For example, if a user's address was
6635 anne@some.where, the "personal" test would incorrectly be true for
6637 To: susanne@some.where
6639 This test is now done by extracting each address from the header in turn,
6640 and checking the entire address. Other tests that are part of "personal"
6641 are now done using regular expressions (for example, to check local parts
6642 of addresses in From: header lines).
6644 (2) The list of non-personal local parts in From: addresses has been
6645 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6646 taken from the Sieve specification recommendations.
6648 (3) If the message contains any header line starting with "List-" it is
6649 treated as non-personal.
6651 (4) The test for "circular" in the Subject: header line has been removed
6652 because it now seems ill-conceived.
6654 15. Minor typos in src/EDITME comments corrected.
6656 16. Installed latest exipick from John Jetmore.
6658 17. If headers_add on a router specified a text string that was too long for
6659 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6660 of string_sprintf() is now avoided.
6662 18. $message_body_size was not set (it was always zero) when running the DATA
6663 ACL and the local_scan() function.
6665 19. For the "mail" command in an Exim filter, no default was being set for
6666 the once_repeat time, causing a random time value to be used if "once" was
6667 specified. (If the value happened to be <= 0, no repeat happened.) The
6668 default is now 0s, meaning "never repeat". The "vacation" command was OK
6669 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6670 (I found it when inspecting the code).
6672 20. There is now an overall timeout for performing a callout verification. It
6673 defaults to 4 times the callout timeout, which applies to individual SMTP
6674 commands during the callout. The overall timeout applies when there is more
6675 than one host that can be tried. The timeout is checked before trying the
6676 next host. This prevents very long delays if there are a large number of
6677 hosts and all are timing out (e.g. when the network connections are timing
6678 out). The value of the overall timeout can be changed by specifying an
6679 additional sub-option for "callout", called "maxwait". For example:
6681 verify = sender/callout=5s,maxwait=20s
6683 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6684 to the end before writing, but this should make it even safer).
6686 22. Exim was forgetting that it had advertised PIPELINING for the second and
6687 subsequent messages on an SMTP connection. It was also not resetting its
6688 memory on STARTTLS and an internal HELO.
6690 23. When Exim logs an SMTP synchronization error within a session, it now
6691 records whether PIPELINING has been advertised or not.
6693 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6694 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6695 rather than long int.
6697 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6699 26. Added the never_mail option to autoreply.
6705 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6706 crash if the getsockname() call failed; this can happen if a connection is
6707 closed very soon after it is established. The problem was simply in the
6708 order in which certain operations were done, causing Exim to try to write
6709 to the SMTP stream before it had set up the file descriptor. The bug has
6710 been fixed by making things happen in the correct order.
6716 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6717 before the connection was closed, thus losing the rejection response.
6719 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6720 some early Solaris releases, but causes trouble in current releases where
6721 socklen_t is defined.
6723 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6726 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6729 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6730 that releasing the top of it at the end releases what was used for sub-
6731 expansions (unless the block got too big). However, discard this block if
6732 the first thing is a variable or header, so that we can use its block when
6733 it is dynamic (useful for very large $message_headers, for example).
6735 6. Lookups now cache *every* query, not just the most recent. A new, separate
6736 store pool is used for this. It can be recovered when all lookup caches are
6737 flushed. Lookups now release memory at the end of their result strings.
6738 This has involved some general refactoring of the lookup sources.
6740 7. Some code has been added to the store_xxx() functions to reduce the amount
6741 of flapping under certain conditions.
6743 8. log_incoming_interface used to affect only the <= reception log lines. Now
6744 it causes the local interface and port to be added to several more SMTP log
6745 lines, for example "SMTP connection from", and rejection lines.
6747 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6749 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6751 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6752 could be overwritten at the end of the current message (or the start of a
6753 new message if it was set in a HELO ACL). The value is now preserved for
6754 the duration of the SMTP connection.
6756 12. If a transport had a headers_rewrite setting, and a matching header line
6757 contained an unqualified address, that address was qualified, even if it
6758 did not match any rewriting rules. The underlying bug was that the values
6759 of the flags that permit the existence of unqualified sender and recipient
6760 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6761 non-local messages, and by -bnq for local messages) were not being
6762 preserved with the message after it was received.
6764 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6765 "next input=" as part of the text comprising the host identity instead of
6766 the correct text. The code was using the same buffer for two different
6767 strings. However, depending on which order the printing function evaluated
6768 its arguments, the bug did not always show up. Under Linux, for example, my
6769 test suite worked just fine.
6771 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6772 doesn't work with some older versions of Perl. It has been changed to "my",
6773 which in any case is probably the better facility to use.
6775 15. A really picky compiler found some instances of statements for creating
6776 error messages that either had too many or two few arguments for the format
6779 16. The size of the buffer for calls to the DNS resolver has been increased
6780 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6781 for addresses that have a lot of PTR records. This alleviates a problem; it
6782 does not fully solve it.
6784 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6785 buffer now truncates the list and logs the incident, which is the same
6786 action as happens when Exim is looking up a host name and its aliases.
6787 Previously in this situation something unpredictable would happen;
6788 sometimes it was "internal error: store_reset failed".
6790 18. If a server dropped the connection unexpectedly when an Exim client was
6791 using GnuTLS and trying to read a response, the client delivery process
6792 crashed while trying to generate an error log message.
6794 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6795 string, for example:
6797 warn message = H1: something\nH2: something
6799 the text was added as a single header line from Exim's point of view
6800 though it ended up OK in the delivered message. However, searching for the
6801 second and subsequent header lines using $h_h2: did not work. This has been
6802 fixed. Similarly, if a system filter added multiple headers in this way,
6803 the routers could not see them.
6805 20. Expanded the error message when iplsearch is called with an invalid key to
6806 suggest using net-iplsearch in a host list.
6808 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6809 ACLs are no longer actually imposed (and a message to that effect is
6812 22. If a "gecos" field in a passwd entry contained escaped characters, in
6813 particular, if it contained a \" sequence, Exim got it wrong when building
6814 a From: or a Sender: header from that name. A second bug also caused
6815 incorrect handling when an unquoted " was present following a character
6816 that needed quoting.
6818 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6819 was not being matched caselessly.
6821 24. Arranged for all hyphens in the exim.8 source to be escaped with
6824 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6825 properly. Recipient callout cache records were still being keyed to include
6826 the sender, even when use_sender was set false. This led to far more
6827 callouts that were necessary. The sender is no longer included in the key
6828 when use_sender is false.
6830 26. Added "control = submission" modifier to ACLs.
6832 27. Added the ${base62d: operator to decode base 62 numbers.
6834 28. dnsdb lookups can now access SRV records.
6836 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6837 the configuration file.
6839 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6840 (-v) mode. This makes the output for a verbose queue run more intelligible.
6842 31. Added a use_postmaster feature to recipient callouts.
6844 32. Added the $body_zerocount variable, containing the number of binary zero
6845 bytes in the message body.
6847 33. The time of last modification of the "new" subdirectory is now used as the
6848 "mailbox time last read" when there is a quota error for a maildir
6851 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6853 35. Added +ignore_unknown as a special item in host lists.
6855 36. Code for decoding IPv6 addresses in host lists is now included, even if
6856 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6857 address was recognized as an IP address, but was then not correctly decoded
6858 into binary, causing unexpected and incorrect effects when compared with
6865 1. Very minor rewording of debugging text in manualroute to say "list of
6866 hosts" instead of "hostlist".
6868 2. If verify=header_syntax was set, and a header line with an unqualified
6869 address (no domain) and a large number of spaces between the end of the
6870 name and the colon was received, the reception process suffered a buffer
6871 overflow, and (when I tested it) crashed. This was caused by some obsolete
6872 code that should have been removed. The fix is to remove it!
6874 3. When running in the test harness, delay a bit after writing a bounce
6875 message to get a bit more predictability in the log output.
6877 4. Added a call to search_tidyup() just before forking a reception process. In
6878 theory, someone could use a lookup in the expansion of smtp_accept_max_
6879 per_host which, without the tidyup, could leave open a database connection.
6881 5. Added the variables $recipient_data and $sender_data which get set from a
6882 lookup success in an ACL "recipients" or "senders" condition, or a router
6883 "senders" option, similar to $domain_data and $local_part_data.
6885 6. Moved the writing of debug_print from before to after the "senders" test
6888 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6889 problems for message scanning, either using a data ACL, or using
6890 local_scan() because the Received: header was not generated till after they
6891 were called (in order to set the time as the time of reception completion).
6892 I have revised the way this works. The header is now generated after the
6893 body is received, but before the ACL or local_scan() are called. After they
6894 are run, the timestamp in the header is updated.
6900 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6901 before starting a queue runner without re-exec. This happened only when
6902 deliver_drop_privilege was set or when the Exim user was set to root. The
6903 effect of the bug was that timeouts during subsequent deliveries caused
6904 crashes instead of being properly handled. The handler is now left at its
6905 default (and expected) setting.
6907 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6908 message, again when deliver_drop_privilege is set or Exim is run as root.
6909 The bug described in (1) was not present in this case, but the tidying up
6910 of the other signals was missing. I have made the two cases consistent.
6912 3. The ignore_target_hosts setting on a manualroute router was being ignored
6913 for hosts that were looked up using the /MX notation.
6915 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6918 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6919 operated on the sender address. After changing the $sender_address to <>
6920 for the sender address verify, Exim was re-instated it as the original
6921 (before rewriting) address, but remembering that it had rewritten it, so it
6922 wasn't rewriting it again. This bug also had the effect of breaking the
6923 sender address verification caching when the sender address was rewritten.
6925 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6926 This has been changed so that if the ip literal address matches
6927 ignore_target_hosts, the router declines.
6929 7. Added expansion conditions match_domain, match_address, and match_local_
6930 part (NOT match_host).
6932 8. The placeholder for the Received: header didn't have a length field set.
6934 9. Added code to Exim itself and to exim_lock to test for a specific race
6935 condition that could lead to file corruption when using MBX delivery. The
6936 issue is with the lockfile that is created in /tmp. If this file is removed
6937 after a process has opened it but before that process has acquired a lock,
6938 there is the potential for a second process to recreate the file and also
6939 acquire a lock. This could lead to two Exim processes writing to the file
6940 at the same time. The added code performs the same test as UW imapd; it
6941 checks after acquiring the lock that its file descriptor still refers to
6942 the same named file.
6944 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6945 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6948 11. Added the smtp_active_hostname option. If used, this will typically be made
6949 to depend on the incoming interface address. Because $interface_address is
6950 not set up until the daemon has forked a reception process, error responses
6951 that can happen earlier (such as "too many connections") no longer contain
6954 12. If an expansion in a condition on a "warn" statement fails because a lookup
6955 defers, the "warn" statement is abandoned, and the next ACL statement is
6956 processed. Previously this caused the whole ACL to be aborted.
6958 13. Added the iplsearch lookup type.
6960 14. Added ident_timeout as a log selector.
6962 15. Added tls_certificate_verified as a log selector.
6964 16. Added a global option tls_require_ciphers (compare the smtp transport
6965 option of the same name). This controls incoming TLS connections.
6967 17. I finally figured out how to make tls_require_ciphers do a similar thing
6968 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6969 before starting the TLS session.
6971 18. Tabs are now shown as \t in -bP output.
6973 19. If the log selector return_path_on_delivery was set, Exim crashed when
6974 bouncing a message because it had too many Received: header lines.
6976 20. If two routers both had headers_remove settings, and the first one included
6977 a superfluous trailing colon, the final name in the first list and the
6978 first name in the second list were incorrectly joined into one item (with a
6979 colon in the middle).
6985 1. Added -C and -D options to the exinext utility, mainly to make it easier
6986 to include in the automated testing, but these could be helpful when
6987 multiple configurations are in use.
6989 2. The exinext utility was not formatting the output nicely when there was
6990 an alternate port involved in the retry record key, nor when there was a
6991 message id as well (for retries that were specific to a specific message
6992 and a specific host). It was also confused by IPv6 addresses, because of
6993 the additional colons they contain. I have fixed the IPv4 problem, and
6994 patched it up to do a reasonable job for IPv6.
6996 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6997 delivery, the log line now contains "pipelined" if PIPELINING was used.
6999 4. An SMTP transport process used to panic and die if the bind() call to set
7000 an explicit outgoing interface failed. This has been changed; it is now
7001 treated in the same way as a connect() failure.
7003 5. A reference to $sender_host_name in the part of a conditional expansion
7004 that was being skipped was still causing a DNS lookup. This no longer
7007 6. The def: expansion condition was not recognizing references to header lines
7008 that used bh_ and bheader_.
7010 7. Added the _cache feature to named lists.
7012 8. The code for checking quota_filecount in the appendfile transport was
7013 allowing one more file than it should have been.
7015 9. For compatibility with Sendmail, the command line option
7023 and sets the incoming protocol and host name (for trusted callers). The
7024 host name and its colon can be omitted when only the protocol is to be set.
7025 Note the Exim already has two private options, -pd and -ps, that refer to
7026 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7027 "s", but I don't think that's a major issue.
7029 10. A number of refactoring changes to the code, none of which should affect
7032 (a) The number of logging options was getting close to filling up the
7033 32-bit word that was used as a bit map. I have split them into two classes:
7034 those that are passed in the argument to log_write(), and those that are
7035 only ever tested independently outside of that function. These are now in
7036 separate 32-bit words, so there is plenty of room for expansion again.
7037 There is no change in the user interface or the logging behaviour.
7039 (b) When building, for example, log lines, the code previously used a
7040 macro that called string_cat() twice, in order to add two strings. This is
7041 not really sufficiently general. Furthermore, there was one instance where
7042 it was actually wrong because one of the argument was used twice, and in
7043 one call a function was used. (As it happened, calling the function twice
7044 did not affect the overall behaviour.) The macro has been replaced by a
7045 function that can join an arbitrary number of extra strings onto a growing
7048 (c) The code for expansion conditions now uses a table and a binary chop
7049 instead of a serial search (which was left over from when there were very
7050 few conditions). Also, it now recognizes conditions like "pam" even when
7051 the relevant support is not compiled in: a suitably worded error message is
7052 given if an attempt is made to use such a condition.
7054 11. Added ${time_interval:xxxxx}.
7056 12. A bug was causing one of the ddress fields not to be passed back correctly
7057 from remote delivery subprocesses. The field in question was not being
7058 subsequently used, so this caused to problems in practice.
7060 13. Added new log selectors queue_time and deliver_time.
7062 14. Might have fixed a bug in maildirsizefile handling that threw up
7063 "unexpected character" debug warnings, and recalculated the data
7064 unnecessarily. In any case, I expanded the warning message to give more
7067 15. Added the message "Restricted characters in address" to the statements in
7068 the default ACL that block characters like @ and % in local parts.
7070 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7071 Three changes have been made:
7073 (a) There was a serious bug; a negative response to MAIL caused the whole
7074 recipient domain to be cached as invalid, thereby blocking all messages
7075 to all local parts at the same domain, from all senders. This bug has
7076 been fixed. The domain is no longer cached after a negative response to
7077 MAIL if the sender used is not empty.
7079 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7082 (c) A new callout option, "use_sender" has been added for people who want
7083 the modified behaviour.
7089 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7092 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7093 indeed breaks things for older releases.
7095 3. Added additional logging to the case where there is a problem reading data
7096 from a filter that is running in a subprocess using a pipe, in order to
7097 try to track down a specific problem.
7099 4. Testing facility fudge: when running in the test harness and attempting
7100 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7101 getting "No route to host". Convert this to a timeout.
7103 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7106 6. Some OS don't have socklen_t but use size_t instead. This affects the
7107 fifth argument of getsockopt() amongst other things. This is now
7108 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7109 can be set for individual OS. I have set it for SunOS5, OSF1, and
7110 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7111 some earlier ones do not.
7113 7. Change 4.30/15 was not doing the test caselessly.
7115 8. The standard form for an IPv6 address literal was being rejected by address
7116 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7117 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7118 this, as well as the form without the "IPv6" on the front (but only when
7119 address literals are enabled, of course).
7121 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7123 10. Exim crashed if a message with an empty sender address specified by -f
7124 encountered a router with an errors_to setting. This could be provoked only
7125 by a command such as
7129 where an empty string was supplied; "<>" did not hit this bug.
7131 11. Installed PCRE release 4.5.
7133 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7134 remained set. It is now erased.
7136 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7137 times from message ids (which are base 36 rather than the normal 62).
7139 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7140 were being counted as actual protocol errors, and logged if the log
7141 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7142 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7143 and DATA following a set of rejected RCPTs do not count as protocol errors.
7144 In other words, Exim assumes they were pipelined, though this may not
7145 actually be the case. Of course, in all cases the client gets an
7146 appropriate error code.
7148 15. If a lookup fails in an ACL condition, a message about the failure may
7149 be available; it is used if testing the ACL cannot continue, because most
7150 such messages specify what the cause of the deferral is. However, some
7151 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7152 that caused an old message to be retained and used if a later statement
7153 caused a defer, replacing the real cause of the deferral.
7155 16. If an IP address had so many PTR records that the DNS lookup buffer
7156 was not large enough to hold them, Exim could crash while trying to process
7157 the truncated data. It now detects and logs this case.
7159 17. Further to 4.21/58, another change has been made: if (and only if) the
7160 first line of a message (the first header line) ends with CRLF, a bare LF
7161 in a subsequent header line has a space inserted after it, so as not to
7162 terminate the header.
7164 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7165 unnecessarily, used atoi() instead of strtol(), and didn't check the
7166 termination when getting file sizes from file names by regex.
7168 19. Completely re-implemented the support for maildirsize files, in the light
7169 of a number of problems with the previous contributed implementation
7170 (4.30/29). In particular:
7172 . If the quota is zero, the maildirsize file is maintained, but no quota is
7175 . If the maildir directory does not exist, it is created before any attempt
7176 to write a maildirsize file.
7178 . The quota value in the file is just a cache; if the quota is changed in
7179 the transport, the new value overrides.
7181 . A regular expression is available for excluding directories from the
7184 20. The autoreply transport checks the characters in options that define the
7185 message's headers; it allows continued headers, but it was checking with
7186 isspace() after an embedded newline instead of explicitly looking for a
7189 21. If all the "regular" hosts to which an address was routed had passed their
7190 expiry times, and had not reached their retry times, the address was
7191 bounced, even if fallback hosts were defined. Now Exim should go on to try
7194 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7195 equivalent code in the SMTP transport. Some hosts send humungous responses
7196 to HELO/EHLO, more than 1024 it seems.
7198 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7199 gives compiler warnings in some environments. I've now done it "properly",
7202 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7203 (because of problems with the built-in one) was declared to return uschar *
7204 instead of char *, causing compiler failure.
7206 25. Fixed a file descriptor leak when processing alias/forward files.
7208 26. Fixed a minor format string issue in dbfn.c.
7210 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7212 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7213 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7216 29. When a sender address is verified, it is cached, to save repeating the test
7217 when there is more than one recipient in a message. However, when the
7218 verification involves a callout, it is possible for different callout
7219 options to be set for different recipients. It is too complicated to keep
7220 track of this in the cache, so now Exim always runs a verification when a
7221 callout is required, relying on the callout cache for the optimization.
7222 The overhead is duplication of the address routing, but this should not be
7225 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7226 to be verified with callout=postmaster, and the main callout worked but the
7227 postmaster check failed, the verification correctly failed. However, if a
7228 subsequent RCPT command asked for sender verification *without* the
7229 postmaster check, incorrect caching caused this verification also to fail,
7232 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7233 it was not caching the DNS options (qualify_single, search_parents) that
7234 were used when the lookup failed. A subsequent lookup with different
7235 options therefore always gave the same answer, though there were cases
7236 where it should not have. (Example: a "domains = !$mx_any" option on a
7237 dnslookup router: the "domains" option is always processed without any
7238 widening, but the router might have qualify_single set.) Now Exim uses the
7239 cached value only when the same options are set.
7241 32. Added John Jetmore's "exipick" utility to the distribution.
7243 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7244 than a timeout (e.g. a certificate is required, and is not provided), an
7245 Exim server now closes the connection immediately. Previously it waited for
7246 the client to close - but if the client is SSL, it seems that they each
7247 wait for each other, leading to a delay before one of them times out.
7249 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7250 maintained 0.8.x compatibility because I don't think many are using it, and
7251 it is clearly obsolete.
7253 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7256 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7257 client certificate was expired. A simple patch fixes this, though I don't
7258 understand the full logic of why the verify callback is called multiple
7261 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7262 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7263 which causes problems with some clients (such as the Certicom SSL Plus
7264 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7265 disables the coutermeasure allowing Eudora to connect."
7267 38. Exim was not checking that a write() to a log file succeeded. This could
7268 lead to Bad Things if a log got too big, in particular if it hit a file
7269 size limit. Exim now panics and dies if it cannot write to a log file, just
7270 as it does if it cannot open a log file.
7272 39. Modified OS/Makefile-Linux so that it now contains
7274 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7276 The two -D definitions ensure that Exim is compiled with large file
7277 support, which makes it possible to handle log files that are bigger than
7280 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7281 instance) a domain was checked against a named list that involved a lookup,
7282 causing $domain_data to be set, then another domain was checked against the
7283 same list, then the first domain was re-checked, the value of $domain_data
7284 after the final check could be wrong. In particular, if the second check
7285 failed, it could be set empty. This bug probably also applied to
7288 41. The strip_trailing_dot option was not being applied to the address given
7289 with the -f command-line option.
7291 42. The code for reading a message's header from the spool was incrementing
7292 $received_count, but never initializing it. This meant that the value was
7293 incorrect (doubled) while delivering a message in the same process in which
7294 it was received. In the most common configuration of Exim, this never
7295 happens - a fresh exec is done - but it can happen when
7296 deliver_drop_privilege is set.
7298 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7299 it now includes up to 150 characters of the unexpected data in the log
7302 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7303 and building data strings. The size of both of these buffers was 10 000
7304 bytes - far larger than anybody would *ever* want, thought I. Needless to
7305 say, somebody hit the limit. I have increased the maximum line length to
7306 20 000 and the maximum data length of concatenated lines to 100 000. I have
7307 also fixed two bugs, because there was no checking on these buffers. Tsk,
7308 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7309 buffer is too small.
7311 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7312 lsearch lookups. Now it does.
7314 46. When parsing a route_list item in a manualroute router, a fixed-length
7315 buffer was used for the list of hosts. I made this 1024 bytes long,
7316 thinking that nobody would ever have a list of hosts that long. Wrong.
7317 Somebody had a whole pile of complicated expansion conditions, and the
7318 string was silently truncated, leading to an expansion error. It turns out
7319 that it is easier to change to an unlimited length (owing to other changes
7320 that have happened since this code was originally written) than to build
7321 structure for giving a limitation error. The length of the item that
7322 expands into the list of hosts is now unlimited.
7324 47. The lsearch lookup could not handle data where the length of text line was
7325 more than 4095 characters. Such lines were truncated, leading to shortened
7326 data being returned. It should now handle lines of any length.
7328 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7329 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7332 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7333 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7334 no longer happens in comment lines. A list of replacements is now placed
7335 at the head of all of the source files, except those whose only change is
7336 to replace PERL_COMMAND in the very first #! line.
7338 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7339 messages on the queue, with a bottom-up merge sort, using code contributed
7340 by Michael Haardt. This should make operations like -bp somewhat faster on
7341 large queues. It won't affect queue runners, except when queue_run_in_order
7344 51. Installed eximstats 1.31 in the distribution.
7346 52. Added support for SRV lookups to the dnslookup router.
7348 53. If an ACL referred to $message_body or $message_body_end, the value was not
7349 reset for any messages that followed in the same SMTP session.
7351 54. The store-handling optimization for building very long strings was not
7352 differentiating between the different store pools. I don't think this
7353 actually made any difference in practice, but I've tidied it.
7355 55. While running the routers to verify a sender address, $sender_address
7356 was still set to the sender address. This is wrong, because when routing to
7357 send a bounce to the sender, it would be empty. Therefore, I have changed
7358 it so that, while verifying a sender address, $sender_address is set to <>.
7359 (There is no change to what happens when verifying a recipient address.)
7361 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7362 target A or AAAA records (if not already returned) without resetting the
7363 qualify_single or search_parents options of the DNS resolver. These are
7364 inappropriate in this case because the targets of MX and SRV records must
7365 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7366 target that, when qualified, matched something in the local domain. These
7367 two options are now turned off when doing these lookups.
7369 57. It seems that at least some releases of Reiserfs (which does not have the
7370 concept of a fixed number of inodes) returns zero and not -1 for the
7371 number of available inodes. This interacted badly with check_spool_inodes,
7372 which assumed that -1 was the "no such thing" setting. What I have done is
7373 to check that the total number of inodes is greater than zero before doing
7374 the test of how many are available.
7376 58. When a "warn" ACL statement has a log_message modifier, the message is
7377 remembered, and not repeated. This is to avoid a lot of repetition when a
7378 message has many recipients that cause the same warning to be written.
7379 However, Exim was preserving the list of already written lines for an
7380 entire SMTP session, which doesn't seem right. The memory is now reset if a
7381 new message is started.
7383 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7384 debugging output unless log_rewrite was also set.
7386 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7387 of (int)(handle) when we know that handle contains (void *)(-1).
7389 61. The Exim daemon panic-logs an error return when it closes the incoming
7390 connection. However "connection reset by peer" seems to be common, and
7391 isn't really an error worthy of noting specially, so that particular error
7394 62. When Exim is trying to find all the local interfaces, it used to panic and
7395 die if the ioctl to get the interface flags failed. However, it seems that
7396 on at least one OS (Solaris 9) it is possible to have an interface that is
7397 included in the list of interfaces, but for which you get a failure error
7398 for this call. This happens when the interface is not "plumbed" into a
7399 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7400 failure of the "get flags" call assumes that the interface is down.
7402 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7403 makes life easier for people who are doing arithmetic on fields extracted
7404 from dates, where you often get leading zeros that should not be
7405 interpreted as octal.
7407 64. Added qualify_domain to the redirect router, to override the global
7410 65. If a pathologically long header line contained very many addresses (the
7411 report of this problem mentioned 10 000) and each of them was rewritten,
7412 Exim could use up a very large amount of memory. (It kept on making new
7413 copies of the header line as it rewrote, and never released the old ones.)
7414 At the expense of a bit more processing, the header rewriting function has
7415 been changed so that it no longer eats memory in this way.
7417 66. The generation of the Received: header has been moved from the time that a
7418 message starts to be received, to the time that it finishes. The timestamp
7419 in the Received: header should now be very close to that of the <= log
7420 line. There are two side-effects of this change:
7422 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7423 logged header lines no longer include the local Received: line, because
7424 it has not yet been created. The same applies to a copy of the message
7425 that is returned to a non-SMTP sender when a message is rejected.
7427 (b) When a filter file is tested using -bf, no additional Received: header
7428 is added to the test message. After some thought, I decided that this
7431 This change does not affect the value of $received_for. It is still set
7432 after address rewriting, but before local_scan() is called.
7434 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7436 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7437 gave an unhelpful panic error message, and a defer error. I have managed to
7438 change this behaviour so that it now rejects any supplied certificate,
7439 which seems right, as the list of acceptable certificates is empty.
7441 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7442 gave an unhelpful defer error. I have not managed to make this reject any
7443 supplied certificates, but the error message it gives is "no certificate
7444 supplied", which is not helpful.
7446 70. exigrep's output now also includes lines that are not associated with any
7447 message, but which match the given pattern. Implemented by a patch from
7448 Martin Sluka, which also tidied up the Perl a bit.
7450 71. Recipient callout verification, like sender verification, was using <> in
7451 the MAIL FROM command. This isn't really the right thing, since the actual
7452 sender may affect whether the remote host accepts the recipient or not. I
7453 have changed it to use the actual sender in the callout; this means that
7454 the cache record is now keyed on a recipient/sender pair, not just the
7455 recipient address. There doesn't seem to be a real danger of callout loops,
7456 since a callout by the remote host to check the sender would use <>.
7457 [SEE ABOVE: changed after hitting problems.]
7459 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7460 temporary errors. However, in the case of such a code being given after
7461 the end of a data transmission (i.e. after ".") Exim was failing to write
7462 a retry record for the message. (Yes, there was some broken host that was
7463 actually sending 8xx at this point.)
7465 73. An unknown lookup type in a host list could cause Exim to panic-die when
7466 the list was checked. (An example that provoked this was putting <; in the
7467 middle of a list instead of at the start.) If this happened during a DATA
7468 ACL check, a -D file could be left lying around. This kind of configuration
7469 error no longer causes Exim to die; instead it causes a defer error. The
7470 incident is still logged to the main and panic logs.
7472 74. Buglet left over from Exim 3 conversion. The message "too many messages
7473 in one connection" was written to the rejectlog but not the mainlog, except
7474 when address rewriting (yes!) was being logged.
7476 75. Added write_rejectlog option.
7478 76. When a system filter was run not as root (that is, when system_filter_user
7479 was set), the values of the $n variables were not being returned to the
7480 main process; thus, they were not subsequently available in the $sn
7483 77. Added +return_path_on_delivery log selector.
7485 78. A connection timeout was being treated differently from recipients deferred
7486 when testing hosts_max_try with a message that was older than the host's
7487 retry timeout. (The host should not be counted, thus allowing all hosts to
7488 be tried at least once before bouncing.) This may have been the cause of an
7489 occasionally reported bug whereby a message would remain on the queue
7490 longer than the retry timeout, but would be bounced if a delivery was
7491 forced. I say "may" because I never totally pinned down the problem;
7492 setting up timeout/retry tests is difficult. See also the next item.
7494 79. The ultimate address timeout was not being applied to errors that involved
7495 a combination of host plus message (for example, a timeout on a MAIL
7496 command). When an address resolved to a number of possible hosts, and they
7497 were not all tried for each delivery (e.g. because of hosts_max_try), a
7498 message could remain on the queue longer than the retry timeout.
7500 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7503 81. Fixed an obscure SMTP outgoing bug which required at least the following
7504 conditions: (a) there was another message waiting for the same server;
7505 (b) the server returned 5xx to all RCPT commands in the first message so
7506 that the message was not completed; (c) the server dropped the connection
7507 or gave a negative response to the RSET that Exim sends to abort the
7508 transaction. The observed case was a dropped connection after DATA that had
7509 been sent in pipelining mode. That is, the server had advertised PIPELINING
7510 but was not implementing it correctly. The effect of the bug was incorrect
7511 behaviour, such as trying another host, and this could lead to a crash.
7517 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7518 and daemon.c were passed as pointers to ints; they should have been
7519 pointers to socklen_t variables (which are typically unsigned ints).
7521 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7524 3. Fixed a really odd bug that affected only the testing scheme; patching a
7525 certain fixed string in the binary changed the value of another string that
7526 happened to be identical to the end of the original first string.
7528 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7529 name", it returns that address as the IP address. On some operating
7530 systems (e.g. Solaris), it also passes back the IP address string as the
7531 "host name". However, on others (e.g. Linux), it passes back an empty
7532 string. Exim wasn't checking for this, and was changing the host name to an
7533 empty string, assuming it had been canonicalized.
7535 5. Although rare, it is permitted to have more than one PTR record for a given
7536 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7537 all the names associated with an address, because they do in Solaris.
7538 However, it seems that they do not in Linux for data that comes from the
7539 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7540 I found this out when I moved to a new Linux workstation and tried to run
7541 the Exim test suite.
7543 To get round this problem I have changed the code so that it now does its
7544 own call to the DNS to look up PTR records when searching for a host name.
7545 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7546 addresses that are only in /etc/hosts are still found.
7548 This behaviour is, however, controlled by an option called host_lookup_
7549 order, which defaults to "bydns:byaddr". If people want to use the other
7550 order, or indeed, just use one or the other means of lookup, they can
7551 specify it in this variable.
7553 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7554 some operating systems, this comes back from gethostbyaddr() as an empty
7555 string, and this is what Exim used to test for. However, it seems that in
7556 other systems, "." is yielded. Exim now tests for this case too.
7558 7. The values of check_spool_space and check_log_space are now held internally
7559 as a number of kilobytes instead of an absolute number of bytes. If a
7560 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7561 kilobyte. This means that much larger values can be stored.
7563 8. Exim monitor: an attempt to get the action menu when not actually pointing
7564 at a message produces an empty menu entitled "No message selected". This
7565 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7566 no entries in it ("Shell widget menu has zero width and/or height"). So I
7567 have added a single, blank menu entry in this case.
7569 9. Added ${quote_local_part.
7571 10. MIME decoding is now applied to the contents of Subject: header lines when
7574 11. Now that a reference to $sender_host_address automatically causes a reverse
7575 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7576 host lookup before query-style lookups in lists that might use this
7577 variable. This has therefore been abolished, and the "net-" prefix is no
7578 longer necessary for query-style lookups.
7580 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7581 have been a typo for LFLAGS, so it has been changed.
7583 13. The install script calls Exim with "-C /dev/null" in order to find the
7584 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7585 to be output. However, since Exim outputs its version number before the
7586 error, it didn't break the script. It just looked ugly. I fixed this by
7587 always allowing "-C /dev/null" if the caller is root.
7589 14. Ignore overlarge ACL variable number when reading spool file - insurance
7590 against a later release with more variables having written the file.
7592 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7593 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7594 this, as well as the form without the "IPv6" on the front.
7596 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7597 OS/Makefile-Darwin file.
7599 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7600 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7602 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7603 variable. However, some OS use ints for inodes. I've added cast to long int
7604 to get rid of the compiler warning.
7606 19. I had forgotten to lock out "/../" in configuration file names when
7607 ALT_CONFIG_PREFIX was set.
7609 20. Routers used for verification do not need to specify transports. However,
7610 if such a router generated a host list, and callout was configured, Exim
7611 crashed, because it could not find a port number from the (non-existent)
7612 transport. It now assumes port 25 in this circumstance.
7614 21. Added the -t option to exigrep.
7616 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7617 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7618 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7619 them is set, LOOKUP_LSEARCH is forced.
7621 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7623 24. Added sender and host information to the "rejected by local_scan()" log
7624 line; previously there was no indication of these.
7626 25. Added .include_if_exists.
7628 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7629 Linux. It turns out that not all file systems support this. Apparently some
7630 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7631 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7632 not supported on the file descriptor, is now ignored when Exim is trying to
7633 sync a directory. This applies only to Linux.
7635 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7637 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7638 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7641 29. Added support for maildirsize files from supplied patch (modified a bit).
7643 30. The use of :fail: followed by an empty string could lead Exim to respond to
7644 sender verification failures with (e.g.):
7646 550 Verification failed for <xxx>
7647 550 Sender verify failed
7649 where the first response line was missing the '-' that indicates it is not
7650 the final line of the response.
7652 31. The loop for finding the name of the user that called Exim had a hardwired
7653 limit of 10; it now uses the value of finduser_retries, which is used for
7654 all other user lookups.
7656 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7659 33. Exim was neglecting to zero errno before one call of strtol() when
7660 expanding a string and expecting an integer value. On some systems this
7661 resulted in spurious "integer overflow" errors. Also, it was casting the
7662 result into an int without checking.
7664 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7665 did not work. The code looks as if it has *never* worked, though it appears
7666 to have been documented since at least release 1.62. I have made it work.
7668 35. The "timeout_DNS" error in retry rules, also documented since at least
7669 1.62, also never worked. As it isn't clear exactly what this means, and
7670 clearly it isn't a major issue, I have abolished the feature by treating it
7671 as "timeout", and writing a warning to the main and panic logs.
7673 36. The display of retry rules for -brt wasn't always showing the error code
7676 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7677 timeout_connect_A, timeout_connect_MX.
7679 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7680 to the empty sender.
7682 39. The daemon was not analysing the content of -oX till after it had closed
7683 stderr and disconnected from the controlling terminal. This meant that any
7684 syntax errors were only noted on the panic log, and the return code from
7685 the command was 0. By re-arranging the code a little, I've made the
7686 decoding happen first, so such errors now appear on stderr, and the return
7687 code is 1. However, the actual setting up of the sockets still happens in
7688 the disconnected process, so errors there are still only recorded on the
7691 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7692 connections (as happens on some IP stacks) was logged at start up time as
7693 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7694 it from "IPv6 and IPv4", which means that two separate sockets are being
7697 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7698 says whether AF_INET or AF_INET6 was passed as an argument.
7700 42. Exiwhat output was messed up when time zones were included in log
7703 43. Exiwhat now gives more information about the daemon's listening ports,
7704 and whether -tls-on-connect was used.
7706 44. The "port" option of the smtp transport is now expanded.
7708 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7709 silently ignored. Now an error message is written to the main and panic
7712 46. There's a new ACL modifier called "logwrite" which writes to a log file
7713 as soon as it is encountered.
7715 47. Added $local_user_uid and $local_user_gid at routing time.
7717 48. Exim crashed when trying to verify a sender address that was being
7720 49. Exim was recognizing only a space character after ".include". It now also
7721 recognizes a tab character.
7723 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7724 extracting the relevant information from the specification. The man page no
7725 longer contains scrambled data for the -d option, and I've added a section
7726 at the front about calling Exim under different names.
7728 51. Added "extra_headers" argument to the "mail" command in filter files.
7730 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7733 53. Installed eximstats 1.29.
7735 54. Added transport_filter_timeout as a generic transport option.
7737 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7738 Cc: header lines. This was required by RFC 822, but it not required by RFC
7741 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7742 incoming messages that did not have them. Now it does so only if the
7743 message originates locally, that is, if there is no associated remote host
7744 address. When Resent- header lines are present, this applies to the Resent-
7745 lines rather than the non-Resent- lines.
7747 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7748 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7750 58. Messages for configuration errors now include the name of the main
7751 configuration file - useful now that there may be more than one file in a
7752 list (.included file names were always shown).
7754 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7755 for those rare installations that do not start the daemon as root or run it
7756 setuid root. I've cut out the call to initgroups() if the daemon is not
7759 60. The Exim user and group can now be bound into the binary as text strings
7760 that are looked up at the start of Exim's processing.
7762 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7764 62. Added $mailstore_basename variable.
7766 63. Installed patch to sieve.c from Michael Haardt.
7768 64. When Exim failed to open the panic log after failing to open the main log,
7769 the original message it was trying to log was written to stderr and debug
7770 output, but if they were not available (the usual case in production), it
7771 was lost. Now it is written to syslog before the two lines that record the
7772 failures to open the logs.
7774 65. Users' Exim filters run in subprocesses under the user's uid. It is
7775 possible for a "deliver" command or an alias in a "personal" command to
7776 provoke an address rewrite. If logging of address rewriting is configured,
7777 this fails because the process is not running as root or exim. There may be
7778 a better way of dealing with this, but for the moment (because 4.30 needs
7779 to be released), I have disabled address rewrite logging when running a
7780 filter in a non-root, non-exim process.
7786 1. The buildconfig auxiliary program wasn't quoting the value set for
7787 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7788 not defined. This bug was masked in 4.22 by the effect that was fixed in
7791 2. Some messages that were rejected after a message id was allocated were
7792 shown as "incomplete" by exigrep. It no longer does this for messages that
7793 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7795 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7796 have allow_domain_literals set, the ID did not get logged in the <= line.
7797 Domain literals are now always recognized in Message-ID: header lines.
7799 4. The first argument for a ${extract expansion item is the key name or field
7800 number. Leading and trailing spaces in this item were not being ignored,
7801 causing some misleading effects.
7803 5. When deliver_drop_privilege was set, single queue runner processes started
7804 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7805 same command in the process it spins off) were not dropping privilege.
7807 6. When the daemon running as "exim" started a queue runner, it always
7808 re-executed Exim in the spun-off process. This is a waste of effort when
7809 deliver_drop_privilege is set. The new process now just calls the
7810 queue-runner function directly.
7816 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7819 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7820 ignored. Though the use of -oP was forcing the writing of a pid file, it
7821 was always written to the default place.
7823 3. If the message "no IP address found for host xxxx" is generated during
7824 incoming verification, it is now followed by identification of the incoming
7825 connection (so you can more easily find what provoked it).
7827 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7829 5. Added some features to "harden" Exim a bit more against certain attacks:
7831 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7832 be put in Local/Makefile. This is like the never_users runtime option,
7833 but it cannot be overridden. The default setting is "root".
7835 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7836 prefix string with which any file named in a -C command line option
7839 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7840 is retained for -C and -D only if the caller of Exim is root. Without
7841 it, the exim user may also use -C and -D and retain privilege.
7843 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7844 command line option is disabled.
7846 6. Macro names set by the -D option must start with an upper case letter, just
7847 like macro names defined in the configuration file.
7849 7. Added "dereference=" facility to LDAP.
7851 8. Two instances of the typo "uknown" in the source files are fixed.
7853 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7854 the Configure-Makefile script screwed up while processing it.
7856 10. Incorporated PCRE 4.4.
7858 11. The SMTP synchronization check was not operating right at the start of an
7859 SMTP session. For example, it could not catch a HELO sent before the client
7860 waited for the greeting. There is now a check for outstanding input at the
7861 point when the greeting is written. Because of the duplex, asynchronous
7862 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7863 way, but not yet received, when the check is performed.
7865 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7866 on TCP/IP sockets, because this apparently causes some broken clients to
7869 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7870 unchanged) from the Cygwin maintainer.
7872 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7873 was supported instead of testing for mbx. Effectively a typo.
7875 15. The spa authenticator server code was not checking that the input it
7876 received was valid base64.
7878 16. The debug output line for the "set" modifier in ACLs was not showing the
7879 name of the variable that was being set.
7881 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7883 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7884 lookup on its own. Something else was needed to trigger the lookup. For
7885 example, a match in host_lookup or the need for a host name in a host list.
7886 Now, if $sender_host_name is referenced and the host name has not yet been
7887 looked up, a lookup is performed. If the lookup fails, the variable remains
7888 empty, and $host_lookup_failed is set to "1".
7890 19. Added "eqi" as a case-independent comparison operator.
7892 20. The saslauthd authentication condition could segfault if neither service
7893 nor realm was specified.
7895 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7896 error message that was logged was misleading, and incoming SMTP
7897 connections were dropped. The message is now more accurate, and temporary
7898 errors are given to SMTP connections.
7900 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7901 (except RSET) with a 421 error, until QUIT is received. However, it was
7902 failing to send a response to QUIT.
7904 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7905 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7906 ACL is now run after the helo_try_verify_hosts code.
7908 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7909 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7910 case-independent, so other case variants are also recognized). Apparently
7911 some systems use these upper case variants.
7913 25. If more than two messages were waiting for the same host, and a transport
7914 filter was specified for the transport, Exim sent two messages over the
7915 same TCP/IP connection, and then failed with "socket operation on non-
7916 socket" when it tried to send the third.
7918 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7920 27. The extern definition of crypt16() in expand.c was not being excluded when
7921 the OS had its own crypt16() function.
7923 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7924 as a preferred synonym for return_size_limit, both as an option and as an
7927 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7929 30. Changed the default configuration ACL to relax the local part checking rule
7930 for addresses that are not in any local domains. For these addresses,
7931 slashes and pipe symbols are allowed within local parts, but the sequence
7932 /../ is explicitly forbidden.
7934 31. SPA server authentication was not clearing the challenge buffer before
7937 32. log_message in a "warn" ACL statement was writing to the reject log as
7938 well as to the main log, which contradicts the documentation and doesn't
7939 seem right (because no rejection is happening). So I have stopped it.
7941 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7942 However, I am unable to do any testing of this.
7944 34. Fixed an infelicity in the appendfile transport. When checking directories
7945 for a mailbox, to see if any needed to be created, it was accidentally
7946 using path names with one or more superfluous leading slashes; tracing
7947 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7949 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7950 contents are added to the log line that is written for every discarded
7951 recipient. (Previously a log_message setting was ignored.)
7953 36. The ${quote: operator now quotes the string if it is empty.
7955 37. The install script runs exim in order to find its version number. If for
7956 some reason other than non-existence or emptiness, which it checks, it
7957 could not run './exim', it was installing it with an empty version number,
7958 i.e. as "exim-". This error state is now caught, and the installation is
7961 38. An argument was missing from the function that creates an error message
7962 when Exim fails to connect to the socket for saslauthd authentication.
7963 This could cause Exim to crash, or give a corrupted message.
7965 39. Added isip, isip4, and isip6 to ${if conditions.
7967 40. The ACL variables $acl_xx are now saved with the message, and can be
7968 accessed later in routers, transports, and filters.
7970 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7971 strings in the file are not string-expanded.
7973 42. If a MAIL command specified a SIZE value that was too large to fit into an
7974 int variable, the check against message_size_limit failed. Such values are
7975 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7976 day this will have to be increased, but I don't think I want to be around
7977 when emails are that large.
7984 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7985 iconv() is not standard in FreeBSD.
7987 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7988 IPv6 enabled. The observed symptom was a segmentation fault on return from
7989 the function os_common_find_running_interfaces() in src/os.c.
7991 3. In the check_special_case() function in daemon.c I had used "errno" as an
7992 argument name, which causes warnings on some systems. This was basically a
7993 typo, since it was named "eno" in the comments!
7995 4. The code that waits for the clock to tick (at a resolution of some fraction
7996 of a second) so as to ensure message-id uniqueness was always waiting for
7997 at least one whole tick, when it could have waited for less. [This is
7998 almost certainly not relevant at current processor speeds, where it is
7999 unlikely to ever wait at all. But we try to future-proof.]
8001 5. The function that sleeps for a time interval that includes fractions of a
8002 second contained a race. It did not block SIGALRM between setting the
8003 timer, and suspending (a couple of lines later). If the interval was short
8004 and the sigsuspend() was delayed until after it had expired, the suspension
8005 never ended. On busy systems this could lead to processes getting stuck for
8008 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8009 process, before it forks any delivery processes. The open lookup caching
8010 mechanism meant that the open file or database connection was passed into
8011 the delivery process. The problem was that delivery processes always tidy
8012 up cached lookup data. This could cause a problem for the next delivery
8013 process started by the queue runner, because the external queue runner
8014 process does not know about the closure. So the next delivery process
8015 still has data in the lookup cache. In the case of a file lookup, there was
8016 no problem because closing a file descriptor in a subprocess doesn't affect
8017 the parent. However, if the lookup was caching a connection to a database,
8018 the connection was closed, and the second delivery process was likely to
8019 see errors such as "PGSQL: query failed: server closed the connection
8020 unexpectedly". The problem has been fixed by closing all cached lookups
8021 in a queue runner before running a delivery process.
8023 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8024 seem to have the "const" qualifier which it has on other OS. I've
8027 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8028 *of the same type* (client or server) with the same public name that an
8029 error should be diagnosed.
8031 9. When Exim looked up a host name for an IP address, but failed to find the
8032 original IP address when looking up the host name (a safety check), it
8033 output the message "<ip address> does not match any IP for NULL", which was
8034 confusing, to say the least. The bug was that the host name should have
8035 appeared instead of "NULL".
8037 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8038 user that is built into the binary, and the -C or -D options is used, root
8039 privilege is dropped before the configuration file is read. In addition,
8040 logging is switched to stderr instead of the normal log files. If the
8041 configuration then re-defines the Exim user, the unprivileged environment
8042 is probably not what is expected, so Exim logs a panic warning message (but
8045 However, if deliver_drop_privilege is set, the unprivileged state may well
8046 be exactly what is intended, so the warning has been cut out in that case,
8047 and Exim is allowed to try to write to its normal log files.
8053 1. smtp_return_error_details was not giving details for temporary sender
8054 or receiver verification errors.
8056 2. Diagnose a configuration error if two authenticators have the same public
8059 3. Exim used not to create the message log file for a message until the first
8060 delivery attempt. This could be confusing when incoming messages were held
8061 for policy or load reasons. The message log file is now created at the time
8062 the message is received, and an initial "Received" line is written to it.
8064 4. The automatically generated man page for command line options had a minor
8065 bug that caused no ill effects; however, a more serious problem was that
8066 the procedure for building the man page automatically didn't always
8067 operate. Consequently, release 4.20 contains an out-of-date version. This
8068 shouldn't happen again.
8070 5. When building Exim with embedded Perl support, the script that builds the
8071 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8072 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8074 6. The freeze_tell option was not being used for messages that were frozen on
8075 arrival, either by an ACL or by local_scan().
8077 7. Added the smtp_incomplete_transaction log selector.
8079 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8080 was accepting AUTH without a new EHLO.
8082 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8083 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8086 10. Logging of TCP/IP connections (when configured) now happens in the main
8087 daemon process instead of the child process, so that the TCP/IP connection
8088 count is more accurate (but it can never be perfect).
8090 11. The use of "drop" in a nested ACL was not being handled correctly in the
8091 outer ACL. Now, if condition failure induced by the nested "drop" causes
8092 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8093 or "require"), the connection is dropped.
8095 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8096 that yield "discard" can now be used with an "accept" or a "discard" verb,
8097 but an error is generated for any others (because I can't see a useful way
8098 to define what should happen).
8100 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8101 are now processed in the same way as lines in the Exim configuration file.
8102 In particular, continuation lines are supported.
8104 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8106 15. Added -ti meaning -t -i.
8108 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8109 domains, and warn by logging if others are found.
8111 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8112 returned from GIFCONF when Exim is trying to find the list of interfaces on
8113 a host. The code in os.c has been modified to copy each ifreq to an aligned
8114 structure in all cases.
8116 Also, in some cases, the returned ifreq's were being copied to a 'struct
8117 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8118 means the last couple of bytes of an IPv6 address could be chopped if the
8119 ifreq contained only a normal sockaddr (14 bytes storage).
8121 18. Named domain lists were not supported in the hosts_treat_as_local option.
8122 An entry such as +xxxx was not recognized, and was treated as a literal
8125 19. Ensure that header lines added by a DATA ACL are included in the reject log
8126 if the ACL subsequently rejects the message.
8128 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8129 MD5 (which is deprecated).
8131 21. When testing a filter file using -bf, Exim was writing a message when it
8132 took the sender from a "From " line in the message, but it was not doing so
8133 when it took $return_path from a Return-Path: header line. It now does.
8135 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8136 with a valid header line field name (a series of printing characters
8137 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8139 23. Changed "disc" in the source to "disk" to conform to the documentation and
8140 the book and for uniformity.
8142 24. Ignore Sendmail's -Ooption=value command line item.
8144 25. When execve() failed while trying to run a command in a pipe transport,
8145 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8146 could be confused with a return value of 69 from the command itself. This
8147 has been changed to 127, the value the shell returns if it is asked to run
8148 a non-existent command. The wording for the related log line suggests a
8149 non-existent command as the problem.
8151 26. If received_header_text expands to an empty string, do not add a Received:
8152 header line to the message. (Well, it adds a token one on the spool, but
8153 marks it "old" so that it doesn't get used or transmitted.)
8155 27. Installed eximstats 1.28 (addition of -nt option).
8157 28. There was no check for failure on the call to getsockname() in the daemon
8158 code. This can fail if there is a shortage of resources on the system, with
8159 ENOMEM, for example. A temporary error is now given on failure.
8161 29. Contrary to the C standard, it seems that in some environments, the
8162 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8163 program. Exim now does this explicitly; it affects the formatting of
8164 timestamps using strftime().
8166 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8167 complaints. I've now initialized all the variables, to avoid this.
8169 32. Header lines added by a system filter were not being "seen" during
8170 transport-time rewrites.
8172 33. The info_callback() function passed to OpenSSL is set up with type void
8173 (*)(SSL *, int, int), as described somewhere. However, when calling the
8174 function (actually a macro) that sets it up, the type void(*)() is
8175 expected. I've put in a cast to prevent warnings from picky compilers.
8177 34. If a DNS black list lookup found a CNAME record, but there were no A
8178 records associated with the domain it pointed at, Exim crashed.
8180 35. If a DNS black list lookup returned more than one A record, Exim ignored
8181 all but the first. It now scans all returned addresses if a particular IP
8182 value is being sought. In this situation, the contents of the
8183 $dnslist_value variable are a list of all the addresses, separated by a
8186 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8187 to accept a host name and all its aliases if the forward lookup for any of
8188 them yielded the IP address of the incoming connection. Now it accepts only
8189 those names whose forward lookup yields the correct IP address. Any other
8190 names are discarded. This closes a loophole whereby a rogue DNS
8191 administrator could create reverse DNS records to break through a
8192 wildcarded host restriction in an ACL.
8194 37. If a user filter or a system filter that ran in a subprocess used any of
8195 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8196 the wrong values were passed to the pipe command ($thisaddress had the
8197 value of $0, $0 had the value of $1, etc). This bug was introduced by
8198 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8200 38. Improved the line breaking for long SMTP error messages from ACLs.
8201 Previously, if there was no break point between 40 and 75 characters, Exim
8202 left the rest of the message alone. Two changes have been made: (a) I've
8203 reduced the minimum length to 35 characters; (b) if it can't find a break
8204 point between 35 and 75 characters, it looks ahead and uses the first one
8205 that it finds. This may give the occasional overlong line, but at least the
8206 remaining text gets split now.
8208 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8209 file descriptors might be low, and that setting 1000 would always raise it.
8210 It turns out that in some environments, the limit is already over 1000 and
8211 that lowering it causes trouble. So now Exim takes care not to decrease it.
8213 40. When delivering a message, the value of $return_path is set to $sender_
8214 address at the start of routing (routers may change the value). By an
8215 oversight, this default was not being set up when an address was tested by
8216 -bt or -bv, which affected the outcome if any router or filter referred to
8219 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8220 log only when "message" or "log_message" are set. However, if one of the
8221 conditions was an address verification, or a call to a nested ACL, the
8222 messages generated by the underlying test were being passed through. This
8223 no longer happens. The underlying message is available in $acl_verify_
8224 message for both "message" and "log_message" expansions, so it can be
8225 passed through if needed.
8227 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8228 new expansion $bh_ to give the encoded byte string without charset
8229 translation. Translation happens only if iconv() is available; HAVE_ICONV
8230 indicates this at build time. HEADERS_CHARSET gives the charset to
8231 translate to; headers_charset can change it in the configuration, and
8232 "headers charset" can change it in an individual filter file.
8234 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8235 that creates RFC 2047 encoded "words" labels them as that charset instead
8236 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8237 expansion operator; (ii) when Exim creates a From: line for a local
8238 message; (iii) when a header line is rewritten to include a "phrase" part.
8240 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8241 buggy, causing it to skip the first lines of messages whose message ID
8242 ended in 'D'. This would not have bitten before Exim release 4.14, because
8243 message IDs were unlikely to end in 'D' before then. The effect was to have
8244 incorrect size information for certain domains.
8246 45. #include "config.h" was missing at the start of the crypt16.c module. This
8247 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8250 46. If there was a timeout during a "random" callout check, Exim treated it as
8251 a failure of the random address, and carried on sending RSET and the real
8252 address. If the delay was just some slowness somewhere, the response to the
8253 original RCPT would be taken as a response to RSET and so on, causing
8254 mayhem of various kinds.
8256 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8257 when I implemented it. It didn't allow for the fact that some option values
8258 may legitimately be negative (e.g. size_addition), and it didn't even do
8259 the right test for positive values.
8261 48. Domain names in DNS records are case-independent. Exim always looks them up
8262 in lower case. Some resolvers return domain names in exactly the case they
8263 appear in the zone file, that is, they may contain uppercase letters. Not
8264 all resolvers do this - some return always lower case. Exim was treating a
8265 change of case by a resolver as a change of domain, similar to a widening
8266 of a domain abbreviation. This triggered its re-routing code and so it was
8267 trying to route what was effectively the same domain again. This normally
8268 caused routing to fail (because the router wouldn't handle the domain
8269 twice). Now Exim checks for this case specially, and just changes the
8270 casing of the domain that it ultimately uses when it transmits the message
8273 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8276 50. If a filter generated a file delivery with a non-absolute name (possible if
8277 no home directory exists for the router), the forbid_file option was not
8280 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8281 the existing equality matching.
8283 52. Exim was using ints instead of ino_t variables in some places where it was
8284 dealing with inode numbers.
8286 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8287 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8288 called TMPDIR, and if it finds it is different, it changes its value.
8290 54. The smtp_printf() function is now made available to local_scan() so
8291 additional output lines can be written before returning. There is also an
8292 smtp_fflush() function to enable the detection of a dropped connection.
8293 The variables smtp_input and smtp_batched_input are exported to
8296 55. Changed the default runtime configuration: the message "Unknown user"
8297 has been removed from the ACL, and instead placed on the localuser router,
8298 using the cannot_route_message feature. This means that any verification
8299 failures that generate their own messages won't get overridden. Similarly,
8300 the "Unrouteable address" message that was in the ACL for unverifiable
8301 relay addresses has also been removed.
8303 56. Added hosts_avoid_esmtp to the smtp transport.
8305 57. The exicyclog script was not checking for the esoteric option
8306 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8307 will work only if exicyclog is run under the appropriate euid.
8309 58. Following a discussion on the list, the rules by which Exim recognises line
8310 endings on incoming messages have been changed. The -dropcr and drop_cr
8311 options are now no-ops, retained only for backwards compatibility. The
8312 following line terminators are recognized: LF CRLF CR. However, special
8313 processing applies to CR:
8315 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8316 nor a local message in the state where . is a terminator.
8318 (ii) If a bare CR is encountered in a header line, an extra space is added
8319 after the line terminator so as not to end the header. The reasoning
8320 behind this is that bare CRs in header lines are most likely either
8321 to be mistakes, or people trying to play silly games.
8323 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8324 was being incorrectly given as 18 bytes larger than it should have been.
8325 This is a VOB (very old bug).
8327 60. This may never have affected anything current, but just in case it has:
8328 When the local host is found other than at the start of a list of hosts,
8329 the local host, those with the same MX, and any that follow, are discarded.
8330 When the list in question was part of a longer list of hosts, the following
8331 hosts (not currently being processed) were also being discarded. This no
8332 longer happens. I'm not sure if this situation could ever has previously
8335 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8338 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8339 header. This is something that is recommended in a new Internet Draft, and
8340 is something that is documented as being done by Sendmail. There are two
8341 possible values. For messages generated by the autoreply transport, Exim
8344 Auto-Submitted: auto-replied
8346 whereas for all other generated messages (e.g. bounces) it adds
8348 Auto-Submitted: auto-generated
8350 63. The "personal" condition in filters now includes a test for the
8351 Auto-Submitted: header. If it contains the string "auto-" the message it
8352 not considered personal.
8354 64. Added rcpt_include_affixes as a generic transport option.
8356 65. Added queue_only_override (default true).
8358 66. Added the syslog_duplication option.
8360 67. If what should have been the first header line of a message consisted of
8361 a space followed by a colon, Exim was mis-interpreting it as a header line.
8362 It isn't of course - it is syntactically invalid and should therefore be
8363 treated as the start of the message body. The misbehaviour could have
8364 caused a number of strange effects, including loss of data in subsequent
8365 header lines, and spool format errors.
8367 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8368 client host had authenticated. This control can now be exercised by an ACL
8369 for more flexibility.
8371 69. By default, callouts do not happen when testing with -bh. There is now a
8372 variant, -bhc, which does actually run the callout code, including
8373 consulting and updating the callout cache.
8375 70. Added support for saslauthd authentication, courtesy of Alexander
8378 71. If statvfs() failed on the spool or log directories while checking their
8379 size for availability, Exim confusingly gave the error "space shortage".
8380 Furthermore, in debugging mode it crashed with a floating point exception.
8381 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8382 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8383 serious problem, Exim now writes to the main and panic logs when this
8384 happens, with details of the failure. It then refuses to accept the
8385 incoming message, giving the message "spool directory problem" or "log
8386 directory problem" with a 421 code for SMTP messages.
8388 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8389 0, 1, and 2 exist, because some OS complain for execs without them (see
8390 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8391 descriptors. However, the code omitted to check that the open succeeded,
8392 causing mysterious errors if for some reason the permissions on /dev/null
8393 got screwed. Now Exim writes a message to the main and panic logs, and
8394 bombs out if it can't open /dev/null.
8396 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8397 interact so that it is all more flexible. It is supposed to remain
8398 backwards compatible. Also added extra_local_interfaces.
8400 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8401 to bomb out with an assertion failure - to the client this appears as a
8402 connection drop. This problem occurs in the part of the code that was taken
8403 from the Samba project. Fortunately, the assertion is in a very simple
8404 function, so I have fixed this by reproducing the function inline in the
8405 one place where it is called, and arranging for authentication to fail
8406 instead of killing the process with assert().
8408 75. The SPA client code was not working when the server requested OEM rather
8409 than Unicode encoding.
8411 76. Added code to make require_files with a specific uid setting more usable in
8412 the case where statting the file as root fails - usually a non-root-mounted
8413 NFS file system. When this happens and the failure is EACCES, Exim now
8414 forks a subprocess and does the per-uid checking as the relevant uid.
8416 77. Added process_log_path.
8418 78. If log_file_path was not explicitly set, a setting of check_log_space or
8419 check_log_inodes was ignored.
8421 79. If a space check for the spool or log partitions fails, the incident is now
8422 logged. Of course, in the latter case the data may get lost...
8424 80. Added the %p formatting code to string_format() so that it can be used to
8425 print addresses in debug_print(). Adjusted all the address printing in the
8426 debugging in store.c to use %p rather than %d.
8428 81. There was a concern that a line of code in smtp_in.c could overflow a
8429 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8430 initially expressed, the concern was not well-founded, because trailing
8431 spaces are removed early. However, if the trailing spaces were followed by
8432 a NULL, they did not get removed, so the overflow was possible. Two fixes
8435 (a) I re-wrote the offending code in a cleaner fashion.
8436 (b) If an incoming SMTP command contains a NULL character, it is rejected
8439 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8440 runs initgroups(), so that if the Exim user is in any additional groups,
8441 they will be used during message reception.
8447 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob