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.
5 Since Exim version 4.97
6 -----------------------
8 JH/s1 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
9 LF-only mode (as detected from the first header line). Previously we did
10 accept that in (normal) CRLF mode; this has been raised as a possible
11 attack scenario (under the name "smtp smuggling", CVE-2023-51766).
17 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
18 SMTP connection" log lines.
20 JH/02 Option default value updates:
21 - queue_fast_ramp (main) true (was false)
22 - remote_max_parallel (main) 4 (was 2)
24 JH/03 Cache static regex pattern compilations, for use by ACLs.
26 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
27 Make the rewrite never match and keep the logging. Trust the
28 admin to be using verify=header-syntax (to actually reject the message).
30 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
31 (under Linux) we watch the dir containing the final file; previously
32 it would be the dir with the first symlink. We still do not monitor
35 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
36 dn_expand() is happy to pass them through.
38 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
39 the file with the proof had an unchanged name, the new proof(s) were
40 loaded on top of the old ones (and nover used; the old ones were stapled).
42 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
43 more than one message arrived in a single connection a reference from
44 the earlier message could be re-used. Often a sigsegv resulted.
45 These variables were introduced in Exim 4.87.
46 Debug help from Graeme Fowler.
48 JH/09 Fix ${filter } for conditions that modify $value. Previously the
49 modified version would be used in construction the result, and a memory
52 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
53 Find and fix by Jasen Betts.
55 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
56 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
57 the systemwide configuration to override the Exim config.
59 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
60 API changes in libopendmarc.
62 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
63 pid 1, in the normal "background daemon" mode, having to drop process-
64 group leadership also lost track of needing to create listener sockets.
66 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
67 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
69 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
70 a capture group which obtained no text (eg. "(abc)*" matching zero
71 occurrences) could cause a segfault if the corresponding $<n> was
74 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
75 included a close-brace character (eg. it itself used an expansion) an
78 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
79 starting TLS. Previously it was after, meaning that attackers on such
80 ports had to be screened using the host_reject_connection main config
81 option. The new sequence aligns better with the STARTTLS behaviour, and
82 permits defences against crypto-processing load attacks, even though it
83 is strictly an incompatible change.
84 Also, avoid sending any SMTP fail response for either the connect ACL
85 or host_reject_connection, for TLS-on-connect ports.
87 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
88 Previously this was not permitted, but it makes reasonable sense.
89 While there, restore a restriction on using it from a connect ACL; given
90 the change JH/16 it could only return false (and before 4.91 was not
93 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
94 was exactly sized compared to the log buffer, a crash occurred with the
95 misleading message "bad memory reference; pool not found".
96 Found and traced by Jasen Betts.
98 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
99 dns_again_means_nonexist included an element causing a DNS lookup which
100 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
101 included (though probably not limited to) a process crash from stack
102 memory limit, or from excessive open files. Replace this with a paniclog
103 whine (as this is likely a configuration error), and returning
106 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
107 this always failed, probably leading to the usual downgrade to in-clear
110 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
111 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
112 to not checking that list for these lookups.
114 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
117 JH/23 Fix crash in string expansions. Previously, if an empty variable was
118 immediately followed by an expansion operator, a null-indirection read
119 was done, killing the process.
121 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
122 include an SMTP response string which is longer than that supported
123 by the delivering transport. Alleviate by wrapping such lines before
126 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
127 chars (RFC limit). Previously a limit of 12 items was made, which with
128 a not-impossible References: in the message being bounced could still
129 be over-large and get stopped in the transport.
131 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
132 close. Previously a bare socket close was done.
134 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
137 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
138 message being considered for delivery was already being handled by
139 another process, and having an SMTP connection already open, the function
140 to close it tried to use an uninitialized variable. This would afftect
141 high-volume sites more, especially when running mailing-list-style loads.
142 Pollution of logs was the major effect, as the other process delivered
143 the message. Found and partly investigated by Graeme Fowler.
145 JH/29 Change format of the internal ID used for message identification. The old
146 version only supported 31 bits for a PID element; the new 64 (on systems
147 which can use Base-62 encoding, which is all currently supported ones
148 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
149 and must use Base-36). The new ID is 23 characters rather than 16, and is
150 visible in various places - notably logs, message headers, and spool file
151 names. Various of the ancillary utilities also have to know the format.
152 As well as the expanded PID portion, the sub-second part of the time
153 recorded in the ID is expanded to support finer precision. Theoretically
154 this permits a receive rate from a single comms channel of better than the
156 The major timestamp part of the ID is not changed; at 6 characters it is
157 usable until about year 3700.
158 Updating from previously releases is fully supported: old-format spool
159 files are still usable, and the utilities support both formats. New
160 message will use the new format. The one hints-DB file type which uses
161 message-IDs (the transport wait- DB) will be discarded if an old-format ID
162 is seen; new ones will be built with only new-format IDs.
163 Optionally, a utility can be used to convert spool files from old to new,
164 but this is only an efficiency measure not a requirement for operation
165 Downgrading from new to old requires running a provided utility, having
166 first stopped all operations. This will convert any spool files from new
167 back to old (losing time-precision and PID information) and remove any
168 wait- hints databases.
170 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
171 we treated them as item separators when parsing for a list item, but they
172 need to be protected by the doublequotes. While there, add handling for
175 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
176 Found and fixed by Jasen Betts. No testcase for this as my usual text
177 editor insists on emitting only valid UTF-8.
179 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
180 a null-indirection SIGSEGV for the receive process.
182 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
183 -bh use. Internal checking would spot this and take a panic.
185 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
186 In 4.96 this would expand to empty.
188 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
189 certificate. Find and fix by Andreas Metzler.
191 JH/36 Add ARC info to DMARC hostory records.
193 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
194 or fakedefer. Previously the sender could discover that the message
195 had in fact been accepted.
197 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
198 sequences. Previously the input was not noted as being tainted; notably
199 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
200 bad coding of authenticators.
202 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
203 and ${tr...}. Found and diagnosed by Heiko Schlichting.
205 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
206 could be triggered by externally-supplied input. Found by Trend Micro.
209 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
210 be triggered by externally-controlled input. Found by Trend Micro.
213 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
214 be triggered by externally-controlled input. Found by Trend Micro.
217 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
218 Make the rewrite never match and keep the logging. Trust the
219 admin to be using verify=header-syntax (to actually reject the message).
221 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
224 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
229 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
230 after reception to before a subsequent reception. This should
231 mean slightly faster delivery, and also confirmation of reception
234 JH/02 Move from using the pcre library to pcre2. The former is no longer
235 being developed or supported (by the original developer).
237 JH/03 Constification work in the filters module required a major version
238 bump for the local-scan API. Specifically, the "headers_charset"
239 global which is visible via the API is now const and may therefore
240 not be modified by local-scan code.
242 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
243 sendfile() didi not account for the way the ClamAV driver code called it.
245 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
246 time check was being done for every character; replace that with one
249 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
250 sent was prefixed with a length byte.
252 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
253 RFC 5321. Previously Dovecot (at least) would log errors during
256 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
257 in 4.61 without a reason logged. Binaries will be bigger, which might
258 matter on diskspace-constrained systems, but debug is easier.
260 JH/09 Fix macro-definition during "-be" expansion testing. The move to
261 write-protected store for macros had not accounted for these runtime
262 additions; fix by removing this protection for "-be" mode.
264 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
265 handing out large-numbered file descriptors, violating the usual Unix
266 assumption (and required by Posix) that the lowest possible number will be
267 allocated by the kernel when a new one is needed. In the daemon, and any
268 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
269 are not useable for FD_SET() [and hence select()] and overwrite the stack.
270 Assorted crashes happen.
272 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
273 main-section options or in a connect ACL, the value from the first ever
274 connection was never replaced for subsequent connections. Found by
277 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
278 though only once PCRE2 was introduced: the memory accounting used under
279 debug offset allocations by an int, giving a hard trap in early startup.
280 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
282 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
283 with underbars is given. The write-protection of configuration introduced
284 in 4.95 trapped when normalisation was applied to an option not needing
287 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
289 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
290 failing to close the certificates directory, every hour or any time it
293 JH/16 Debugging initiated by an ACL control now continues through into routing
294 and transport processes. Previously debugging stopped any time Exim
295 re-execs, or for processing a queued message.
297 JH/17 The "expand" debug selector now gives more detail, specifically on the
298 result of expansion operators and items.
300 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
301 bad comparison between the option value and the name of the file to
302 be included was done, and a mismatch was wrongly identified.
303 4.88 to 4.95 are affected.
305 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
307 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
308 supplied as the db file-pair basename. Previously, if a directory
309 path was given, for example via the autoreply "once" option, the DB
310 file.pag and file.dir files would be created in that directory's
313 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
314 "taint" log_selector. These were previously deprecated.
316 JH/22 Fix static address-list lookups to properly return the matched item.
317 Previously only the domain part was returned.
319 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
320 the call into OpenSSL to send a TLS Close was being repeated; this
321 resulted in the library waiting for the peer's Close. If that was never
322 sent we waited forever. Fix by tracking send calls.
324 JH/24 The ${run} expansion item now expands its command string elements after
325 splitting. Previously it was before; the new ordering makes handling
326 zero-length arguments simpler. The old ordering can be obtained by
327 appending a new option "preexpand", after a comma, to the "run".
329 JH/25 Taint-check exec arguments for transport-initiated external processes.
330 Previously, tainted values could be used. This affects "pipe", "lmtp" and
331 "queryprogram" transport, transport-filter, and ETRN commands.
332 The ${run} expansion is also affected: in "preexpand" mode no part of
333 the command line may be tainted, in default mode the executable name
336 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
337 the the facility was not passed across execs, and only the first message
338 passed over a connection could use BDAT; any further ones using DATA.
340 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
341 uses $sending_ip_address and an interface is specified.
342 Previously any use of the local address in the EHLO name disabled
343 PIPECONNECT, the common case being to use the rDNS of it.
345 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
346 resumption. Previously verify failed because no certificate status is
347 passed on the wire for the restarted session. Fix by using the recorded
348 ocsp status of the stored session for the new connection.
350 JH/29 TLS resumption: the key for session lookup in the client now includes
351 more info that a server could potentially use in configuring a TLS
352 session, avoiding oferring mismatching sessions to such a server.
353 Previously only the server IP was used.
355 JH/30 Fix string_copyn() for limit greater than actual string length.
356 Previously the copied amount was the limit, which could result in a
357 overlapping memcpy for newly allocated destination soon after a
358 source string shorter than the limit. Found/investigated by KM.
360 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
361 close; it may be needed for a subsequent connection. This caused a
362 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
364 JH/32 Fix CHUNKING for a second message on a connection when the first was
365 rejected. Previously we did not reset the chunking-offered state, and
366 erroneously rejected the BDAT command. Investigation help from
369 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
370 an empty address. Previously the expansion returned an error.
372 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
373 proxy. Previously these were misparsed, leading to paniclog entries.
379 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
380 related applications. Previously an "H" was used where available info
381 says that "M" should be, so change to match.
383 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
384 as arguments, so an implementation trying to copy these into a local
385 buffer was taking a taint-enforcement trap. Fix by using dynamically
386 created buffers. Similar fix for radius expansion condition.
388 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
389 reasonable, eg. to count headers. Fix by using dynamically created
390 buffers rather than a local. Do similar fixes for ACL actions "dcc",
391 "log_reject_target", "malware" and "spam"; the arguments are expanded
392 so could be handling tainted values.
394 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
395 broken the (no-op) support for this sendmail command. Restore it
396 to doing nothing, silently, and returning good status.
398 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
399 record path was given (or the default used) without a leading directory
400 path, an error occurred on trying to open it. Use the transport's working
403 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
404 transport. Previously it was the name on the DNS A-record; use instead
405 the head of the CNAME chain leading there (if there is one). This seems
406 to align better with RFC 6125.
408 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
409 smtp_accept_max_per_host allocated resources which were not released
410 when the limit was exceeded. This eventually crashed the daemon. Fix
411 by adding a release action in that path.
413 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
414 expanded; previously using tainted values was rejected. Fix by using
415 dynamically-created buffers.
417 JH/09 Relax restrictions on ACL verify condition needing access to message
418 headers. Previously they were only permitted in data and non-smtp ACLs;
419 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
420 not_blind, header_sender and header_names_ascii verification.
422 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
423 Previously a macro used one argument twice; when called with the
424 argument as an expression having side-effects, incorrect operation
425 resulted. Use an inlineable function.
427 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
428 held open for a verify callout. Previously this wan not accounted for
429 and a corrupt onward SMTP conversation resulted.
431 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
432 Previously they were embedded naively in the constructed address; when
433 needed, strip the quoting and quote the entire local_part.
434 Also make the inbound_srs expansion condition handle quoting.
436 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
437 excluded, not matching the documentation.
439 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
440 was given for the sqlite_dbfile a trap resulted.
442 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
443 "name" argument resulted in a trap. There is no reason to disallow such;
444 this was a coding error.
446 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
447 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
448 spent suspended, ignoring the POSIX definition. Previously we assumed
449 it did and a constant offset from real time could be used as a correction.
450 Change to using the same clock source for the start-of-message and the
451 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
452 exists, just to get a clock slightly more aligned to reality.
454 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
455 RFC says it is optional some validators care. The missing char was not
456 intended but triggered by a line-wrap alignment. Discovery and fix by
457 Guillaume Outters, hacked on by JH.
459 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
460 name being quoted was tainted a trap would be taken. Fix by using
461 dynamically created buffers. The routine could have been called by a
462 rewrite with the "h" flag, by using the "-F" command-line option, or
463 by using a "name=" option on a control=submission ACL modifier.
465 JH/19 SPF: change the Authentication-Results expansion component to give
466 smtp.helo when the sender domain is empty. Previously it gave
469 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
470 not in 127.0.0.0/8 to help in spotting list domains taken over by a
471 domain-parking registrar.
473 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
474 Previously when a whitespace character was specified it was not inserted
475 after removing the newline.
477 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
478 the domain part of the recipient address. This overrides any tls_sni
479 option set, which was previously used.
481 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
484 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
485 is_tainted() had an off-by-one error in the overenthusiastic direction.
486 Find and fix by Gavan. Although NetBSD is not a supported platform for
487 4.94 this bug could affect other platforms.
489 PP/01 Fix default prime selection to be consistent.
490 One path used ike23 still, instead of exim.dev.20160529.3; now both
491 execution flows will use the same DH primes (currently
492 exim.dev.20160529.3).
494 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
495 option in smtp transport, to match the documentation. Previously
496 verification was not being done in some cases where it should have been.
498 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
499 than one server was defined and depending on the platform memory layout
500 details, an internal consistency trap could be hit while walking the list
503 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
504 files. The value is used by the authresults expansion item. Previously
505 if this was used in a router or transport, a crash could result.
507 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
508 have a sufficient incidence of EINTR returns from select that an
509 interaction with other operations done by the main daemon loop exposed
510 a bug in the error-handling. This was benign apart from the log
513 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
514 for consistency with delivered "=>" and failed "**" lines. While we're
515 there, handle PRX and TFO.
517 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
518 applied. This resulted, if any header-line rewrite rules were configured,
519 in a panic-log triggerable by sending a message with a long address in
520 a header. Fix by increasing the arbitrary limit to larger than a single
521 (dewrapped) 5322 header line maximum size.
523 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
524 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
525 RFC 6648 which deprecates X- options in protocols as a general practice.
526 Changeover between the implementations is handled by the mechanisms
529 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
530 a remote which uses callout/hold. Previously the local was queued.
532 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
535 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
536 log line. The asterisk is now consistently appended to the remote IP
537 (and port, if given), and will also be provided on defer and fail log
538 lines. Previously it could be placed on the local IP if that was being
539 logged, and was only provided on delivery lines.
541 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
543 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
544 authenticator client_send option. Previously the next char, after a pair
545 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
546 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
547 documentation. There is still no way to get a leading ^ immediately
548 after a NUL (ie. for the password of a PLAIN method authenticator.
550 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
551 files. For bad sizes read, delete the record and whine to paniclog.
553 JH/38 When logging an AUTH failure, as server, do not include sensitive
554 information. Previously, the credentials would be included if given
555 as part of the AUTH command line and an ACL denied authentication.
557 JH/39 Bug 2691: fix $local_part_data. When the matching list element
558 referred to a file, bad data was returned. This likely also affected
561 JH/40 The gsasl authenticator now supports caching of the salted password
562 generated by the client-side implementation. This required the addition
563 of a new variable: $auth4.
565 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
566 left undeleted; the attempt to re-create it then failed - resulting in
567 the usual "SIGHUP tp have daemon reload configuration" to not work.
568 This affected any platform not supporting "abstract" Unix-domain
569 sockets (i.e. not Linux).
571 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
572 recipients" response to RCPT in a later response, with a 250. The
573 previous coding assumed this would not happen, and under PIPELINING
574 would result in both lost and duplicate recipients for a message.
576 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
577 Previously the weighting was incorrectly applied. Similar fix for socks
578 proxies. Found and fixed by Heiko Schlichting.
580 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
581 not handle sub-lists included using the +namedlist syntax. While
582 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
583 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
586 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
587 Previously the permanent pool was used, so the sore could not be freed.
588 This meant a connection with many messages would use continually-growing
591 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
592 per-pool so as not to waste too much space. Previously a constant size
593 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
594 DOS attacks harder. The cost is wasted memory use in the larger blocks.
596 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
597 same space repeatedly, and a smaller process footprint.
599 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
600 for GnuTLS. Previously the name started "%s" which, while not a bug,
601 looked as if if might be one.
603 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
604 "continued connection") make the $proxy_* variables available. Previously
605 the information was not passed across the exec() call for subsequent
606 transport executions. This also mean that the log lines for the
607 messages can show the proxy information.
609 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
610 receive time. With modern systems the difference is significant.
611 The historical behaviour can be restored by disabling (a new) log_selector
612 "queue_time_exclusive".
614 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
615 content) were specifically tested for. Now, also cover expansions
616 resulting in ACL names and inline ACL content.
618 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
619 making it unusable in complex expressions.
621 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
622 fail with a taint-check on the spoolfile name, and leave the message
625 HS/01 Enforce absolute PID file path name.
627 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
629 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
631 PP/02 Bug 2643: Correct TLS DH constants.
632 A missing NUL termination in our code-generation tool had led to some
633 incorrect Diffie-Hellman constants in the Exim source.
634 Reported by kylon94, code-gen tool fix by Simon Arlott.
636 PP/03 Impose security length checks on various command-line options.
637 Fixes CVE-2020-SPRSS reported by Qualys.
639 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
640 better. Reported by Qualys.
642 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
643 providing a particularly obnoxious sender full name.
646 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
648 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
651 PP/08 Change default for recipients_max from unlimited to 50,000.
653 PP/09 Fix security issue with too many recipients on a message (to remove a
654 known security problem if someone does set recipients_max to unlimited,
655 or if local additions add to the recipient list).
656 Fixes CVE-2020-RCPTL reported by Qualys.
658 PP/10 Fix security issue in SMTP verb option parsing
659 Fixes CVE-2020-EXOPT reported by Qualys.
661 PP/11 Fix security issue in BDAT state confusion.
662 Ensure we reset known-good where we know we need to not be reading BDAT
663 data, as a general case fix, and move the places where we switch to BDAT
664 mode until after various protocol state checks.
665 Fixes CVE-2020-BDATA reported by Qualys.
667 HS/03 Die on "/../" in msglog file names
669 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
670 the Exim runtime user are allowed to create files.
672 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
675 QS/03 When reading the output from interpreted forward files we do not
676 pass the pipe between the parent and the interpreting process to
677 executed child processes (if any).
679 QS/04 Always die if requested from internal logging, even is logging is
682 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
683 the API; compilation noo longer completes with DMARC support included.
684 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
685 been reported on other platforms.
687 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
689 JH/56 Make the majority of info read from config files readonly, for defence-in-
690 depth against exploits. Suggestion by Qualys.
691 Not supported on Solaris 10.
693 JH/57 Fix control=fakreject for a custom message containing tainted data.
694 Previously this resulted in a log complaint, due to a re-expansion present
695 since fakereject was originally introduced.
697 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
698 modifier was given, a loop resulted.
700 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
701 pipelined SMTP command followed the BDAT LAST then it would be
702 incorrectly treated as part of the message body, causing a verification
705 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
706 require looser validation rules than those for 821-level addresses,
707 which only permit IP addresses.
713 JH/01 Avoid costly startup code when not strictly needed. This reduces time
714 for some exim process initialisations. It does mean that the logging
715 of TLS configuration problems is only done for the daemon startup.
717 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
719 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
720 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
723 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
724 DKIM signing is being done. Previously a transport_filter would always
725 disable CHUNKING, falling back to traditional DATA.
727 JH/05 Regard command-line recipients as tainted.
729 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
731 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
732 PAM library frees one of the arguments given to it, despite the
733 documentation. Therefore a plain malloc must be used.
735 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
736 on-stack buffers were used, resulting in a taint trap when DSN information
737 copied from a received message was written into the buffer.
739 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
740 the ordering of its ARC headers. This caused a crash.
742 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
743 a new record was being constructed with information from the peer, a trap
746 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
747 installation would get error messages from DMARC verify, when it hit the
748 nonexistent file indicated by the default. Distros wanting DMARC enabled
749 should both provide the file and set the option.
750 Also enforce no DMARC verification for command-line sourced messages.
752 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
753 could, depending on the platform, hang at the STARTTLS response.
755 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
756 message on a connection. Previously if one message had ARC headers and
757 the following one did not, a crash could result when adding an
758 Authentication-Results: header.
760 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
761 Exim main code and Exim-related utities. The introduction of taint
762 tracking also did many adjustments to string handling. Since then, eximon
763 frequently terminated with an assert failure.
765 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
766 check for 452 responses. This slightly helps the inefficieny of doing
767 a large alias-expansion into a recipient-limited target. The max_rcpt
768 transport option still applies (and at the current default, will override
769 the new feature). The check is done for either cause of synch, and forces
770 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
771 connection. The new facility is not tunable at this time.
773 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
774 library live data was being used, so the results became garbage. Make
775 copies while it is still usable.
777 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
778 on delivery deferred (==) and failed (**) lines (if a delivery was
779 attemtped). Previously it was only on completion (=>) lines.
781 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
782 for the expansion of the server_scram_iter and server_scram_salt options.
784 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
785 are now specifically given a NO_DATA response without hitting the system
786 resolver. The library goes on to do the now-standard TXT lookup.
787 Use of dnsdb lookups is not affected.
789 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
790 only retrieve the errormessage once. Previously two calls to dlerror()
791 were used, and the second one (for mainlog/paniclog) retrieved null
794 JH/20 Taint checking: disallow use of tainted data for
795 - the appendfile transport file and directory options
796 - the pipe transport command
797 - the autoreply transport file, log and once options
798 - file names used by the redirect router (including filter files)
800 - paths used by single-key lookups
801 Previously this was permitted.
803 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
804 adjusted the size of a major service buffer; this failed because the
805 buffer was in use at the time. Change to a compile-time increase in the
806 buffer size, when this authenticator is compiled into exim.
808 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
809 previous fast-mode was untenable in the face of glibs using mmap to
810 support larger malloc requests.
812 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
813 New values supported, if defined on system where compiled:
814 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
815 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
817 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
818 running a limited number of proceses in parallel, a benefit is gained. The
819 amount varies with the platform hardware and load. The use of the option
820 queue_run_in_order means we cannot do this, as ordering becomes
823 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
824 had introduced a string-copy (for ensuring NUL-termination) which was not
825 appropriate for that case, which can include embedded NUL bytes in the
826 block of data. Investigation showed the copy to actually be needless, the
827 data being length-specified.
829 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
830 done during a receiving connection, and both used TLS, global info was
831 used rather than per-connection info for tracking the state of data
832 queued for transmission. This could result in a connection hang.
834 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
835 Previously, when delivering serveral messages down a single connection
836 only the first would provide a SIZE. This was due to the size information
837 not being properly tracked.
839 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
840 TAI (at 37 seconds currently), pretend to be in UTC for time-related
841 expansion and logging. Previously, spurious values such as a future
842 minute could be seen.
844 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
845 it could crash from a null-deref. This could also affect the
846 ${addresses: } operator and ${readsock } item.
848 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
849 message following a mime one, the variable was not reset.
851 JH/30 When an pipelined-connect fails at the first response, assume incorrect
852 cached capability (perhaps the peer reneged?) and immediately retry in
855 JH/31 Fix spurious detection of timeout while writing to transport filter.
857 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
858 an attempt to copy the string was made before checking it.
860 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
861 taint of the lookup key was maintained; we now regard the presence in the
862 filesystem as sufficient validation.
864 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
865 argument is supplied.
867 JH/35 The dsearch lookup now requires that the directory is an absolute path.
868 Previously this was not checked, and nonempty relative paths made an
869 access under Exim's current working directory.
871 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
872 Previously no event was raised.
874 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
875 parameter supplied by the sender MAIL FROM command. Previously it was
876 ignored, and only the check_spool_space option value for the required
879 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
880 the size of the signing public-key. Previously it was instead giving
881 the size of the signature hash.
883 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
884 the default. See the (new) dkim_verify_min_keysizes option.
886 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
887 and an ACL use a lookup for checking either the local_part or domain,
888 stale data could be accessed. Ensure that variable references are
889 dropped between messages.
891 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
892 by the client was not checked as pointing within response data before
893 being used. A malicious client could thus cause an out-of-bounds read and
894 possibly gain authentication. Fix by adding the check.
896 JH/42 Internationalisation: change the default for downconversion in the smtp
897 transport to be "if needed". Previously it was "as previously set" for
898 the message, which usually meant "if needed" for message-submission but
899 "no" for everything else. However, MTAs have been seen using SMTPUTF8
900 even when the envelope addresses did not need it, resulting in forwarding
901 failures to non-supporting MTAs. A downconvert in such cases will be
902 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
903 transport. The change does mean that addresses needing conversion will
904 be converted when previously a delivery failure would occur.
906 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
907 response was received it would be used unchecked in a fail-DSN, violating
908 standards on line-length limits. Truncate if needed.
910 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
911 doesn't work. (Additionally add a new main config option to configure the
918 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
919 side, to decode a TLS 1.3 packet capture.
921 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
922 Previously the default library behaviour applied, sending two, each in
925 JH/03 Debug output for ACL now gives the config file name and line number for
928 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
930 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
932 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
933 buffer overrun for (non-chunking) other transports.
935 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
936 TLS1.3, means that a server rejecting a client certificate is not visible
937 to the client until the first read of encrypted data (typically the
938 response to EHLO). Add detection for that case and treat it as a failed
939 TLS connection attempt, so that the normal retry-in-clear can work (if
940 suitably configured).
942 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
943 and/or domain. Found and fixed by Jason Betts.
945 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
946 configuration). If a CNAME target was not a wellformed name pattern, a
949 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
950 the OS reports them interleaved with other addresses.
952 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
953 used both for input and for a verify callout, both encrypted, SMTP
954 responses being sent by the server could be lost. This resulted in
955 dropped connections and sometimes bounces generated by a peer sending
958 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
959 string. Previously it was possible to cause undefined behaviour in a
960 library routine (usually a crash). Found by "zerons".
962 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
965 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
966 API was removed, so update to use the newer ones.
968 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
969 any timeout set, is taking a long time. Previously we would hang on to a
970 rotated logfile "forever" if the input was arriving with long gaps
971 (a previous attempt to fix addressed lack, for a long time, of initial
974 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
975 shared (NFS) environment. The length of the tempfile name is now
976 4 + 16 ("hdr.$message_exim_id") which might break on file
977 systems which restrict the file name length to lower values.
980 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
981 shared (NFS) environment.
983 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
984 did for all versions <4.90). Notably -M, -m, --invert, -I may be
987 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
988 on some platforms for bit 31.
990 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
991 to changes apparently associated with TLS1.3 handling some of the APIs
992 previously used were either nonfunctional or inappropriate. Strings
993 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
994 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
995 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
996 This affects log line X= elements, the $tls_{in,out}_cipher variables,
997 and the use of specific cipher names in the encrypted= ACL condition.
999 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1001 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1002 verification result was not updated unless hosts_require_ocsp applied.
1004 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1005 queue_list_requires_admin set to false, non-admin users were denied the
1008 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1009 directory-of-certs mode. Previously they were advertised despite the
1012 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1013 A single TCP connection by a client will now hold a TLS connection open
1014 for multiple message deliveries, by default. Previously the default was to
1017 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1018 default. If built with the facility, DANE will be used. The facility
1019 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1021 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1022 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1023 must be defined and you must still, unless you define DISABLE_TLS, manage
1024 the the include-dir and library-file requirements that go with that
1025 choice. Non-TLS builds are still supported.
1027 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1030 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1031 default. If the platform supports and has the facility enabled, it will
1032 be requested on all coneections.
1034 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1035 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1037 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1039 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1040 success-DSN messages. Previously the From: header was always the default
1041 one for these; the option was ignored.
1043 JH/28 Fix the timeout on smtp response to apply to the whole response.
1044 Previously it was reset for every read, so a teergrubing peer sending
1045 single bytes within the time limit could extend the connection for a
1046 long time. Credit to Qualsys Security Advisory Team for the discovery.
1048 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1049 delivery address, which leaked information of the results of local
1050 forwarding. Change to the original envelope recipient address, per
1053 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1054 requested. Previously not bounce was generated and a log entry of
1055 error ignored was made.
1057 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1059 JH/32 Introduce a general tainting mechanism for values read from the input
1060 channel, and values derived from them. Refuse to expand any tainted
1061 values, to catch one form of exploit.
1063 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1064 was unused and the unexpanded text used for the test. Found and
1065 fixed by Ruben Jenster.
1067 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1068 an attempt to use a TLS library read routine dereffed a nul pointer,
1071 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1072 them in our smtp response.
1074 JH/36 Have the general router option retry_use_local_part default to true when
1075 any of the restrictive preconditions are set (to anything). Previously it
1076 was only for check_local user. The change removes one item of manual
1077 configuration which is required for proper retries when a remote router
1078 handles a subset of addresses for a domain.
1080 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1081 link count into consideration.
1083 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1084 caused the extension of big_buffer, the following lines were ignored.
1086 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1087 accordance with RFC 2308. Previously there was no expiry, so a longlived
1088 receive process (eg. due to ACL delays) versus a short SOA value could
1091 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1093 JH/39 Promote DMARC support to mainline.
1095 JH/40 Bug 2452: Add a References: header to DSNs.
1097 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1098 parameters. The relevant library call is documented as "Deprecated: This
1099 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1100 3.6.0, DH parameters are negotiated following RFC7919."
1102 HS/06 Change the default of dnssec_request_domains to "*"
1104 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1105 carried on and emitted a BDAT command, even when PIPELINING was not
1108 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1109 buffer was used for the filename, resulting in a trap when tainted
1110 arguments (eg. $domain) were used.
1112 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1113 recommended to avoid a possible server-load attack. The feature can be
1114 re-enabled via the openssl_options main cofiguration option.
1116 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1117 for version 4.90 - adding a "more data" boolean to the arguments.
1118 Bumped the ABI version number also, this having been missed previously;
1119 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1120 respect of usage of smtp_printf() by either local_scan code or libraries
1121 accessed via the ${dlfunc } expansion item. Both will need coding
1122 adjustment for any calls to smtp_printf() to match the new function
1123 signature; a FALSE value for the new argument is always safe.
1125 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1126 the file-offset (which the Linux syscall does, and exim expects); this
1127 resulted in an indefinite loop.
1129 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1130 to do ARC verification. The Authentication-Results: header line added
1131 by the configuration then had no ARC item.
1133 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1134 an incorrect mode was used for reading the result, resulting in it being
1141 JH/01 Remove code calling the customisable local_scan function, unless a new
1142 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1144 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1145 non-signal-safe functions being used.
1147 JH/03 Bug 2269: When presented with a received message having a stupidly large
1148 number of DKIM-Signature headers, disable DKIM verification to avoid
1149 a resource-consumption attack. The limit is set at twenty.
1151 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1152 report of oldest_pass in ${authres } in consequence, and separate out
1153 some descriptions of reasons for verification fail.
1155 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1156 files in the spool were present and unlocked. A queue-runner could spot
1157 them, resulting in a duplicate delivery. Fix that by doing the unlock
1158 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1159 add more error-checking on spoolfile handling while that code is being
1162 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1163 No known attacks, no CVE, this is defensive hardening.
1165 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1166 a queue-runner could start a delivery while other operations were ongoing.
1167 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1168 Found and investigated by Tim Stewart. Fix by using the open message data
1169 file handle rather than opening another, and not locally closing it (which
1170 releases a lock) for that case, while creating the temporary .eml format
1171 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1173 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1174 $sender_verify_failure/$recipient_verify_failure to "random".
1176 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1179 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1180 Previously this would segfault.
1182 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1185 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1186 like zero, since the resolver should be doing this for us, But we need one
1187 as a CNAME but no MX presence gets the CNAME returned; we need to check
1188 that doesn't point to an MX to declare it "no MX returned" rather than
1189 "error, loop". A new main option is added so the older capability of
1190 following some limited number of chain links is maintained.
1192 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1194 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1195 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1196 now-deprecated earlier definitions used only specified the range up to TLS
1197 1.2 (in the older-version library docs).
1199 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1201 JH/15 Rework TLS client-side context management. Stop using a global, and
1202 explicitly pass a context around. This enables future use of TLS for
1203 connections to service-daemons (eg. malware scanning) while a client smtp
1204 connection is using TLS; with cutthrough connections this is quite likely.
1206 JH/16 Fix ARC verification to do AS checks in reverse order.
1208 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1210 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1211 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1212 Previously the "utf8" would be re-prepended for every additional message.
1214 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1215 Previously thery were accepted, resulting in issues when attempting to
1216 forward messages to a non-supporting MTA.
1218 PP/02 Let -n work with printing macros too, not just options.
1220 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1221 one parent address was copied, and bogus data was used at delivery-logging
1222 time. Either a crash (after delivery) or bogus log data could result.
1223 Discovery and analysis by Tim Stewart.
1225 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1226 Previously if the string ended mid-character, we did not insert the
1227 promised '?' replacement.
1229 PP/04 Documentation: current string operators work on bytes, not codepoints.
1231 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1232 should pack well giving a smaller memory footprint so better caching and
1233 therefore performance. Group the declarations where this can't be done so
1234 that the byte-sized flag variables are not interspersed among pointer
1235 variables, giving a better chance of good packing by the compiler.
1237 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1238 non-null, to avoid issues with sites running BATV. Previously reports were
1239 sent with an empty envelope sender so looked like bounces.
1241 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1242 The ignore_error flag wasn't being returned from the filter subprocess so
1243 was not set for later routers. Investigation and fix by Matthias Kurz.
1245 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1246 and a msg:complete for the whole, when a message is manually removed using
1247 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1249 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1250 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1251 hope that is portable enough.
1253 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1254 requiring a known-CA anchor certificate; make it now rely entirely on the
1255 TLSA as an anchor. Checking the name on the leaf cert against the name
1256 on the A-record for the host is still done for TA (but not for EE mode).
1258 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1259 included in delivery lines for non-proxied connections, when compiled with
1260 SUPPORT_SOCKS and running with proxy logging enabled.
1262 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1263 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1264 move the existing event to fire before the normal logging of message
1265 failure so that custom logging is bracketed by normal logging.
1267 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1268 msg:fail:internal event. Developement by Matthias Kurz.
1270 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1271 far too small for todays use of crypto signatures stored there. Go all
1272 the way to the max DNS message size of 64kB, even though this might be
1273 overmuch for IOT constrained device use.
1275 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1276 copy a string over itself. The library routine is documented as not
1277 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1279 JH/32 For main options check_spool_space and check_inode_space, where the
1280 platform supports 64b integers, support more than the previous 2^31 kB
1281 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1282 the previous G, M, k.
1284 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1285 $authenticated_fail_id variable on authentication failure. Previously
1288 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1289 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1290 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1291 GNUTLS_SEC_PARAM_MEDIUM.
1293 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1294 side. Previously we would continue as if no SNI had been received.
1296 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1297 "<" character, which should be a list-separator specification, we walked
1298 off past the nul-terimation.
1300 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1301 causes) even when the retry time is not yet met. Previously they were
1302 not, meaning that when (say) an account was over-quota and temp-rejecting,
1303 and multiple senders' messages were queued, only one sender would get
1304 notified on each configured delay_warning cycle.
1306 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1308 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1309 cipher-suites, an error can be left on the stack even for a succeeding
1310 accept; this results in impossible error messages when a later operation
1313 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1314 return error codes indicating retry. Under TLS1.3 this becomes required.
1316 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1317 it only wrote the new authenticators, resulting in a lack of tracking of
1318 peer changes of ESMTP extensions until the next cache flush.
1320 JH/41 Fix the loop reading a message header line to check for integer overflow,
1321 and more-often against header_maxsize. Previously a crafted message could
1322 induce a crash of the recive process; now the message is cleanly rejected.
1324 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1325 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1331 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1332 When redis_servers is set to a list of > 1 element, and the Redis servers
1333 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1334 case of MOVED into a DEFER case instead, thus moving the query onto the
1335 next server in the list. For a cluster of N elements, all N servers must
1336 be defined in redis_servers.
1338 GF/02 Catch and remove uninitialized value warning in exiqsumm
1339 Check for existence of @ARGV before looking at $ARGV[0]
1341 JH/01 Replace the store_release() internal interface with store_newblock(),
1342 which internalises the check required to safely use the old one, plus
1343 the allocate and data copy operations duplicated in both (!) of the
1344 extant use locations.
1346 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1347 modifier. This matches the restriction on the commandline.
1349 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1350 Previously only the last row was returned.
1352 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1353 we assumed that tags in the header were well-formed, and parsed the
1354 element content after inspecting only the first char of the tag.
1355 Assumptions at that stage could crash the receive process on malformed
1358 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1359 While running the DKIM ACL we operate on the Permanent memory pool so that
1360 variables created with "set" persist to the DATA ACL. Also (at any time)
1361 DNS lookups that fail create cache records using the Permanent pool. But
1362 expansions release any allocations made on the current pool - so a dnsdb
1363 lookup expansion done in the DKIM ACL releases the memory used for the
1364 DNS negative-cache, and bad things result. Solution is to switch to the
1365 Main pool for expansions.
1366 While we're in that code, add checks on the DNS cache during store_reset,
1367 active in the testsuite.
1368 Problem spotted, and debugging aided, by Wolfgang Breyha.
1370 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1371 When none of the hosts presented to a transport match an already-open
1372 connection, close it and proceed with the list. Previously we would
1373 queue the message. Spotted by Lena with Yahoo, probably involving
1376 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1377 Previously a spurious "250 OK id=" response was appended to the proper
1380 JH/08 The "support for" informational output now, which built with Content
1381 Scanning support, has a line for the malware scanner interfaces compiled
1382 in. Interface can be individually included or not at build time.
1384 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1385 by the template makefile "src/EDITME". The "STREAM" support for an older
1386 ClamAV interface method is removed.
1388 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1389 rows affected is given instead).
1391 JH/11 The runtime Berkeley DB library version is now additionally output by
1392 "exim -d -bV". Previously only the compile-time version was shown.
1394 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1395 SMTP connection. Previously, when one had more recipients than the
1396 first, an abortive onward connection was made. Move to full support for
1397 multiple onward connections in sequence, handling cutthrough connection
1398 for all multi-message initiating connections.
1400 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1401 routers. Previously, a multi-recipient message would fail to match the
1402 onward-connection opened for the first recipient, and cause its closure.
1404 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1405 a timeout on read on a GnuTLS initiating connection, resulting in the
1406 initiating connection being dropped. This mattered most when the callout
1407 was marked defer_ok. Fix to keep the two timeout-detection methods
1410 JH/15 Relax results from ACL control request to enable cutthrough, in
1411 unsupported situations, from error to silently (except under debug)
1412 ignoring. This covers use with PRDR, frozen messages, queue-only and
1415 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1417 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1418 metadata, resulting in a crash in free().
1420 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1421 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1422 Broken also in d185889f4, with init system revamp.
1424 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1425 Previously we abruptly closed the connection after reading a malware-
1426 found indication; now we go on to read the "scan ok" response line,
1429 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1430 ACL. Previously, a crash would result.
1432 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1433 macro text after a replacement (previously it was only once per line) and
1434 by skipping builtin macros when searching for an uppercase lead character.
1436 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1437 for the build is renamed.
1439 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1440 was allocated for every new TLS startup, meaning one per message. Fix
1441 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1443 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1444 reported the original. Fix to report (as far as possible) the ACL
1445 result replacing the original.
1447 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1448 OpenSSL. Certificate information is loaded for every new TLS startup,
1449 and the resources needed to be freed.
1451 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1453 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1456 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1457 DATA response info to the (existing) per-recipient response info for
1458 the "C=" log element. It can have useful tracking info from the
1459 destination system. Patch from Simon Arlott.
1461 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1462 length value. Previously this would segfault.
1464 HS/02 Support Avast multiline protoocol, this allows passing flags to
1465 newer versions of the scanner.
1467 JH/28 Ensure that variables possibly set during message acceptance are marked
1468 dead before release of memory in the daemon loop. This stops complaints
1469 about them when the debug_store option is enabled. Discovered specifically
1470 for sender_rate_period, but applies to a whole set of variables.
1471 Do the same for the queue-runner and queue-list loops, for variables set
1472 from spool message files. Do the same for the SMTP per-message loop, for
1473 certain variables indirectly set in ACL operations.
1475 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1476 as a multi-recipient message from a mailinglist manager). The coding had
1477 an arbitrary cutoff number of characters while checking for more input;
1478 enforced by writing a NUL into the buffer. This corrupted long / fast
1479 input. The problem was exposed more widely when more pipelineing of SMTP
1480 responses was introduced, and one Exim system was feeding another.
1481 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1482 receiving system, and refused recipients seen by the sending system
1483 (propating to people being dropped from mailing lists).
1484 Discovered and pinpointed by David Carter.
1486 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1487 replaced by the ${authresults } expansion.
1489 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1491 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1492 allows proper process termination in container environments.
1494 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1495 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1497 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1498 and "err_perm", deprecated since 4.83 when the RFC-defined words
1499 "temperror" and "permerror" were introduced.
1501 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1502 transport-filters or DKIM-signing. The restriction was lost in the
1503 consolidation of verify-callout and delivery SMTP handling.
1504 Extend the restriction to also cover ARC-signing.
1506 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1507 in defer=pass mode supply a 450 to the initiator. Previously the message
1510 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1511 tls_require_ciphers is used as before.
1513 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1514 "pass_unscanned". Only tmpfails from the scanner are written to
1515 the paniclog, as they may require admin intervention (permission
1516 denied, license issues). Other scanner errors (like decompression
1517 bombs) do not cause a paniclog entry.
1519 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1520 Previously it was possible to log spurious information in receive log
1523 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1524 triggered odd behaviour from Outlook Express clients.
1526 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1529 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1530 since the IETF WG has not yet settled on that versus the original
1531 "bare" representation.
1533 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1534 Previously the millisecond value corrupted the output.
1535 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1536 corrupted the output.
1542 JH/01 Rework error string handling in TLS interface so that the caller in
1543 more cases is responsible for logging. This permits library-sourced
1544 string to be attached to addresses during delivery, and collapses
1545 pairs of long lines into single ones.
1547 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1548 during configuration. Wildcards are allowed and expanded.
1550 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1551 This permits better logging.
1553 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1554 do not close it down and have the child transport start it up again on
1555 the passed-on TCP connection. Instead, proxy the child (and any
1556 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1557 affected: the continued delivery log lines do not have any DNSSEC, TLS
1558 Certificate or OCSP information. TLS cipher information is still logged.
1560 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1561 identical IP addresses on different listening ports. Will also affect
1564 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1565 add noisy ifdef guards to special-case this sillyness.
1566 Patch from Bernd Kuhls.
1568 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1569 than 255 are no longer allowed.
1571 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1572 Disable the session-cache too, which might reduce our load. Since we
1573 currrectly use a new context for every connection, both as server and
1574 client, there is no benefit for these.
1575 GnuTLS appears to not support tickets server-side by default (we don't
1576 call gnutls_session_ticket_enable_server()) but client side is enabled
1577 by default on recent versions (3.1.3 +) unless the PFS priority string
1580 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1581 <https://reproducible-builds.org/specs/source-date-epoch/>.
1583 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1584 the check for any unsuccessful recipients did not notice the limit, and
1585 erroneously found still-pending ones.
1587 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1588 MSG_MORE. Only in-clear (not on TLS connections).
1590 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1591 transport-filter is involved we can buffer the headers in memory for
1592 creating the signature, and read the spool data file once for the
1593 signature and again for transmission.
1595 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1596 4.77 as the kernel support then wasn't solid, having issues in 64bit
1597 mode. Now, it's been long enough. Add support for FreeBSD also.
1599 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1600 case where the routing stage had gathered several addresses to send to
1601 a host before calling the transport for the first, we previously failed
1602 to close down TLS in the old transport process before passing the TCP
1603 connection to the new process. The new one sent a STARTTLS command
1604 which naturally failed, giving a failed delivery and bloating the retry
1605 database. Investigation and fix prototype from Wolfgang Breyha.
1607 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1608 false-negatives in the check that the sender had not preempted a response
1609 or prompt from Exim (running as a server), due to that code's lack of
1610 awareness of the SMTP input buffering.
1612 PP/04 Add commandline_checks_require_admin option.
1613 Exim drops privileges sanely, various checks such as -be aren't a
1614 security problem, as long as you trust local users with access to their
1615 own account. When invoked by services which pass untrusted data to
1616 Exim, this might be an issue. Set this option in main configuration
1617 AND make fixes to the calling application, such as using `--` to stop
1620 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1621 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1622 responses to those, into a single TLS record each way (this usually means
1623 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1626 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1627 affects you only if you're dancing at the edge of the param size limits.
1628 If you are, and this message makes sense to you, then: raise the
1629 configured limit or use OpenSSL 1.1. Nothing we can do for older
1632 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1633 cmdline element to get the documented default one. Previously it was
1636 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1637 are unsuable for non-message-specific reasons - eg. connection timeout,
1638 banner-time rejection.
1640 JH/16 Fix logging of delivery remote port, when specified by router, under
1643 PP/06 Repair manualroute's ability to take options in any order, even if one
1644 is the name of a transport.
1647 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1649 JH/17 Change the list-building routines interface to use the expanding-string
1650 triplet model, for better allocation and copying behaviour.
1652 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1653 Previously it was constructed the first time a possibly-matching string
1654 was met in the configuration file input during startup; now it is done
1657 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1658 compatible one, to avoid the (poorly documented) possibility of a config
1659 file in the working directory redirecting the DB files, possibly correpting
1660 some existing file. CVE-2017-10140 assigned for BDB.
1662 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1663 cache-hot. Previously, although the result was properly cached, the
1664 initial verify call returned a defer.
1666 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1667 the main verify for receipient in uncached-mode.
1669 JH/22 Retire historical build files to an "unsupported" subdir. These are
1670 defined as "ones for which we have no current evidence of testing".
1672 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1673 if present. Previously it was ignored.
1675 JH/24 Start using specified-initialisers in C structure init coding. This is
1676 a C99 feature (it's 2017, so now considered safe).
1678 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1679 if was a fixed-sized field and bitmask ops via macros; it is now more
1682 PP/07 GitHub PR 56: Apply MariaDB build fix.
1683 Patch provided by Jaroslav Å karvada.
1685 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1686 during Coverity cleanups [4.87 JH/47]
1687 Diagnosis and fix provided by Michael Fischer v. Mollard.
1689 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1690 the right size to place the terminating semicolon on its own folded
1691 line, the header hash was calculated to an incorrect value thanks to
1692 the (relaxed) space the fold became.
1694 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1695 and confused the parent.
1697 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1698 which could crash as a result. This could lead to undeliverable messages.
1700 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1703 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1704 matters on fast-turnover and PID-randomising systems, which were getting
1705 out-of-order delivery.
1707 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1708 a possibly-overlapping copy. The symptom was that "Remote host closed
1709 connection in response to HELO" was logged instead of the actual 4xx
1712 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1713 Previously only that bufferd was discarded, resulting in SYMTP command
1716 JH/32 DKIM: when a message has multiple signatures matching an identity given
1717 in dkim_verify_signers, run the dkim acl once for each. Previously only
1718 one run was done. Bug 2189.
1720 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1721 "panic the current process" to "deliberately defer". The panic log is
1722 still written with the problem list name; the mail and reject logs now
1723 get a temp-reject line for the message that was being handled, saying
1724 something like "domains check lookup or other defer". The SMTP 451
1725 message is still "Temporary local problem".
1727 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1728 A crafted sequence of BDAT commands could result in in-use memory beeing
1729 freed. CVE-2017-16943.
1731 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1732 from SMTP input. Previously it was always done; now only done for DATA
1733 and not BDAT commands. CVE-2017-16944.
1735 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1736 to the message (such as an overlong header line). Previously this was
1737 not done and we did not exit BDAT mode. Followon from the previous item
1738 though a different problem.
1744 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1745 than -2003 did; needs libidn2 in addition to libidn.
1747 JH/02 The path option on a pipe transport is now expanded before use.
1749 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1750 Patch provided by "Björn", documentation fix added too.
1752 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1753 missing a wire-to-host endian conversion.
1755 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1756 close after a BDAT command line could be taken as a following command,
1757 giving a synch failure. Fix by only checking for synch immediately
1758 before acknowledging the chunk.
1760 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1761 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1762 macro. Patches provided by Josh Soref.
1764 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1765 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1766 by IANA as a service extension. However, John Klensin suggests that we
1769 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1770 the dkim code may be unix-mode line endings rather than smtp wire-format
1771 CRLF, so prepend a CR to any bare LF.
1773 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1774 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1776 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1777 the first were themselves being wrongly included in the feed into dkim
1778 processing; with most chunk sizes in use this resulted in an incorrect
1779 body hash calculated value.
1781 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1782 DKIM signature block, for verification. Although advised against by
1783 standards it is specifically not ruled illegal.
1785 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1787 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1788 missing a body hash (the bh= tag).
1790 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1791 It seems that HAProxy sends the Proxy Protocol information in clear and
1792 only then does a TLS startup, so do the same.
1794 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1795 TCP connections (such as for Spamd) unless the daemon successfully set
1796 Fast Open mode on its listening sockets. This fixes breakage seen on
1797 too-old kernels or those not configured for Fast Open, at the cost of
1798 requiring both directions being enabled for TFO, and TFO never being used
1799 by non-daemon-related Exim processes.
1801 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1802 endings, at least on the first header line. Try to canonify any that get
1803 past that check, despite the cost.
1805 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1806 now limited to an arbitrary five deep, while parsing addresses with the
1807 strip_excess_angle_brackets option enabled.
1809 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1810 instead leave the unprompted TLS handshake in socket buffer for the
1811 TLS library to consume.
1813 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1815 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1817 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1818 region in one operation, for speed, and this leaves assigned pointers
1819 dangling. Add checks run only under the testsuite which checks all
1820 variables at a store-reset and panics on a dangling pointer; add code
1821 explicitly nulling out all the variables discovered. Fixes one known
1822 bug: a transport crash, where a dangling pointer for $sending_ip_address
1823 originally assigned in a verify callout, is re-used.
1825 PP/06 Drop '.' from @INC in various Perl scripts.
1827 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1829 PP/08 Reduce a number of compilation warnings under clang; building with
1830 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1831 should be warning-free.
1833 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1835 HS/01 Fix portability problems introduced by PP/08 for platforms where
1836 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1838 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1839 chunk. This allows us to accept broken chunked messages. We need a more
1840 general solution here.
1842 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1843 already-broken messages in the queue.
1845 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1847 JH/19 Fix reference counting bug in routing-generated-address tracking.
1853 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1854 supports it and a size is available (ie. the sending peer gave us one).
1856 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1857 years of being deprecated). The replacements are the ACLs
1858 acl_smtp_mime and acl_not_smtp_mime.
1860 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1861 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1862 or even an in-clear connection were permitted. Now, if the host lookup
1863 was dnssec and dane was requested then the host is only used if the
1864 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1865 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1866 if one fails this test.
1867 This means that a poorly-configured remote DNS will make it incommunicado;
1868 but it protects against a DNS-interception attack on it.
1870 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1871 non-noisy when a race steals the message being considered.
1873 JH/05 If main configuration option tls_certificate is unset, generate a
1874 self-signed certificate for inbound TLS connections.
1876 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1877 in rewrites and routers.
1879 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1880 and logged a warning sing 4.83; now they are a configuration file error.
1882 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1883 (lacking @domain). Apply the same qualification processing as RCPT.
1885 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1887 JH/10 Support ${sha256:} applied to a string (as well as the previous
1890 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1891 a cutthrough deliver is pending, as we always want to make a connection.
1892 This also avoids re-routing the message when later placing the cutthrough
1893 connection after a verify cache hit.
1894 Do not update it with the verify result either.
1896 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1897 when routing results in more than one destination address.
1899 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1900 signing (which inhibits the cutthrough capability). Previously only
1901 the presence of an option was tested; now an expansion evaluating as
1902 empty is permissible (obviously it should depend only on data available
1903 when the cutthrough connection is made).
1905 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1906 the relevant preceding SMTP command did not note the pipelining mode.
1908 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1909 Previously they were not counted.
1911 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1912 as one having no matching records. Previously we deferred the message
1913 that needed the lookup.
1915 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1916 distinguished as "(=".
1918 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1919 for missing MX records. Previously it only worked for missing A records.
1921 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1923 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1924 after the data-go-ahead and data-ack. Patch from Jason Betts.
1926 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1927 even for a "none" policy. Patch from Tony Meyer.
1929 JH/22 Fix continued use of a connection for further deliveries. If a port was
1930 specified by a router, it must also match for the delivery to be
1933 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1934 When one of the recipients of a message was unsuitable for the connection
1935 (has no matching addresses), we lost track of needing to mark it
1936 deferred. As a result mail would be lost.
1938 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1940 JH/25 Decoding ACL controls is now done using a binary search; the source code
1941 takes up less space and should be simpler to maintain. Merge the ACL
1942 condition decode tables also, with similar effect.
1944 JH/26 Fix problem with one_time used on a redirect router which returned the
1945 parent address unchanged. A retry would see the parent address marked as
1946 delivered, so not attempt the (identical) child. As a result mail would
1949 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1950 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1951 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1954 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1955 Main config options check_{log,spool}_{inodes,space} are now
1956 100 inodes, 10MB unless set otherwise in the configuration.
1958 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1959 Previously it only applied to the main-section connection policy
1962 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1964 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1965 by me. Added RFC7919 DH primes as an alternative.
1967 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1968 are not in the system include path.
1970 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1971 GnuTLS, when a session startup failed (eg because the client disconnected)
1972 Exim did stdio operations after fclose. This was exposed by a recent
1973 change which nulled out the file handle after the fclose.
1975 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1976 signed directly by the cert-signing cert, rather than an intermediate
1977 OCSP-signing cert. This is the model used by LetsEncrypt.
1979 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1981 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1982 an incoming connection.
1984 HS/02 Bug 1802: Do not half-close the connection after sending a request
1987 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1988 fallback to "prime256v1".
1990 JH/34 SECURITY: Use proper copy of DATA command in error message.
1991 Could leak key material. Remotely exploitable. CVE-2016-9963.
1997 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1998 and 3.4.4 - once the server is enabled to respond to an OCSP request
1999 it does even when not requested, resulting in a stapling non-aware
2000 client dropping the TLS connection.
2002 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2003 support variable-length bit vectors. No functional change.
2005 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2006 The I= interface field on outgoing lines is now after the H= remote
2007 host field, same as incoming lines. There is a separate
2008 outgoing_interface log selector which allows you to disable the
2011 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2012 If not running log_selector +smtp_connection the mainlog would be held
2013 open indefinitely after a "too many connections" event, including to a
2014 deleted file after a log rotate. Leave the per net connection logging
2015 leaving it open for efficiency as that will be quickly detected by the
2016 check on the next write.
2018 HS/01 Bug 1671: Fix post transport crash.
2019 Processing the wait-<transport> messages could crash the delivery
2020 process if the message IDs didn't exist for some reason. When
2021 using 'split_spool_directory=yes' the construction of the spool
2022 file name failed already, exposing the same netto behaviour.
2024 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2025 mime_regex ACL conditions.
2027 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2028 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2029 to HELO, local diagnostic string.
2031 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2032 log line to debug. Even when configured with a tls authenticator many
2033 client connections are expected to not authenticate in this way, so
2034 an authenticate fail is not an error.
2036 HS/02 Add the Exim version string to the process info. This way exiwhat
2037 gives some more detail about the running daemon.
2039 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2040 matter for fast-change records such as DNSBLs.
2042 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2043 retry record, even if constant. There may be multiple transports with
2044 different interface settings and the retry behaviour needs to be kept
2047 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2049 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2051 JH/10 Bug 840: fix log_defer_output option of pipe transport
2053 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2054 hosts_randomize. This matters a lot when combined with mua_wrapper.
2056 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2057 ${quote_pgsql:<string>} operator.
2059 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2061 JH/14 Change header file name for freeradius-client. Relevant if compiling
2062 with Radius support; from the Gentoo tree and checked under Fedora.
2064 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2066 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2067 option transport_filter, meaning no filtering.
2069 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2071 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2072 defaults to "*" (all hosts). The variable is now available when not built
2073 with TLS, default unset, mainly to enable keeping the testsuite sane.
2074 If a server certificate is not supplied (via tls_certificate) an error is
2075 logged, and clients will find TLS connections fail on startup. Presumably
2076 they will retry in-clear.
2077 Packagers of Exim are strongly encouraged to create a server certificate
2078 at installation time.
2080 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2081 with the $config_file variable.
2083 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2084 in transport context, after the attempt, and per-recipient. The latter type
2085 is per host attempted. The event data is the error message, and the errno
2086 information encodes the lookup type (A vs. MX) used for the (first) host,
2087 and the trailing two digits of the smtp 4xx response.
2089 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2090 to write to mainlog (or rejectlog, paniclog) in the window between file
2091 creation and permissions/ownership being changed. Particularly affects
2092 installations where exicyclog is run as root, rather than exim user;
2093 result is that the running daemon panics and dies.
2095 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2097 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2098 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2099 "pri" and "weight". Note that the previous implicit priority given by the
2100 list order is no longer honoured.
2102 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2103 for DKIM processing.
2105 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2106 by defining SUPPORT_SOCKS.
2108 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2109 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2110 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2111 variables are renamed to proxy_{local,external}_{address,port}.
2113 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2114 for a build by defining SUPPORT_I18N
2116 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2117 of the query string, and make ${quote_redis:} do that quoting.
2119 JH/29 Move Events support from Experimental to mainline, enabled by default
2120 and removable for a build by defining DISABLE_EVENT.
2122 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2124 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2125 cached by the daemon.
2127 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2128 by defining LOOKUP_REDIS. The libhiredis library is required.
2130 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2131 keys are given for lookup.
2133 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2134 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2135 only supported when built with TLS support. The PolarSSL SHA routines
2136 are still used when the TLS library is too old for convenient support.
2138 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2139 openssl_options), for security. OpenSSL forces this from version 1.1.0
2140 server-side so match that on older versions.
2142 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2143 allocation for $value could be released as the expansion processing
2144 concluded, but leaving the global pointer active for it.
2146 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2147 and to use the domains and local_parts ACL conditions.
2149 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2150 incorrectly not doubled on cutthrough transmission, hence seen as a
2151 body-termination at the receiving system - resulting in truncated mails.
2152 Commonly the sender saw a TCP-level error, and retransmitted the message
2153 via the normal store-and-forward channel. This could result in duplicates
2154 received - but deduplicating mailstores were liable to retain only the
2155 initial truncated version.
2157 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2159 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2161 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2162 we're in there, support oversigning also; bug 1309.
2164 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2166 HS/04 Add support for keep_environment and add_environment options.
2168 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2169 either intentional arithmetic overflow during PRNG, or testing config-
2172 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2173 delivery resulted in actual delivery. Cancel cutthrough before DATA
2176 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2177 rejects a recipient: pass the reject to the originator.
2179 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2180 Many were false-positives and ignorable, but it's worth fixing the
2183 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2184 for the new environment-manipulation done at startup. Move the routines
2185 from being local to tls.c to being global via the os.c file.
2187 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2188 an extract embedded as result-arg for a map, the first arg for extract
2189 is unavailable so we cannot tell if this is a numbered or keyed
2190 extraction. Accept either.
2196 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2199 JH/02 The smtp transport option "multi_domain" is now expanded.
2201 JH/03 The smtp transport now requests PRDR by default, if the server offers
2204 JH/04 Certificate name checking on server certificates, when exim is a client,
2205 is now done by default. The transport option tls_verify_cert_hostnames
2206 can be used to disable this per-host. The build option
2207 EXPERIMENTAL_CERTNAMES is withdrawn.
2209 JH/05 The value of the tls_verify_certificates smtp transport and main options
2210 default to the word "system" to access the system default CA bundle.
2211 For GnuTLS, only version 3.0.20 or later.
2213 JH/06 Verification of the server certificate for a TLS connection is now tried
2214 (but not required) by default. The verification status is now logged by
2215 default, for both outbound TLS and client-certificate supplying inbound
2218 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2221 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2222 Status Notification (bounce) messages are now MIME format per RFC 3464.
2223 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2224 under the control of the dsn_advertise_hosts option, and routers may
2225 have a dsn_lasthop option.
2227 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2228 default, modifiable by a malware= option. The list separator for
2229 the options can now be changed in the usual way. Bug 68.
2231 JH/10 The smtp_receive_timeout main option is now expanded before use.
2233 JH/11 The incoming_interface log option now also enables logging of the
2234 local interface on delivery outgoing connections.
2236 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2237 if the interface and destination host and port all match.
2239 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2242 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2243 Patch from Andrew Lewis.
2245 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2246 now supports optional time-restrictions, weighting, and priority
2247 modifiers per server. Patch originally by <rommer@active.by>.
2249 JH/16 The spamd_address main option now supports a mixed list of local
2250 and remote servers. Remote servers can be IPv6 addresses, and
2251 specify a port-range.
2253 JH/17 Bug 68: The spamd_address main option now supports an optional
2254 timeout value per server.
2256 JH/18 Bug 1581: Router and transport options headers_add/remove can
2257 now have the list separator specified.
2259 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2262 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2265 JH/21 Support for the A6 type of dns record is withdrawn.
2267 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2268 rather than the verbs used.
2270 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2271 from 255 to 1024 chars.
2273 JH/24 Verification callouts now attempt to use TLS by default.
2275 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2276 are generic router options now. The defaults didn't change.
2278 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2279 Original patch from Alexander Shikoff, worked over by JH.
2281 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2282 files if ZCAT_COMMAND is not executable.
2284 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2286 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2288 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2289 Normally benign, it bites when the pair was led to by a CNAME;
2290 modern usage is to not canonicalize the domain to a CNAME target
2291 (and we were inconsistent anyway for A-only vs AAAA+A).
2293 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2295 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2296 when evaluating $sender_host_dnssec.
2298 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2299 $sender_helo_dnssec variable.
2301 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2303 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2305 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2307 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2308 documented as working, but never had. Support all but $spam_report.
2310 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2311 added for tls authenticator.
2313 HS/03 Add perl_taintmode main config option
2319 TL/01 When running the test suite, the README says that variables such as
2320 no_msglog_check are global and can be placed anywhere in a specific
2321 test's script, however it was observed that placement needed to be near
2322 the beginning for it to behave that way. Changed the runtest perl
2323 script to read through the entire script once to detect and set these
2324 variables, reset to the beginning of the script, and then run through
2325 the script parsing/test process like normal.
2327 TL/02 The BSD's have an arc4random API. One of the functions to induce
2328 adding randomness was arc4random_stir(), but it has been removed in
2329 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2330 function when detected.
2332 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2333 cause callback expansion.
2335 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2336 syntax errors in an expansion can be treated as a string instead of
2337 logging or causing an error, due to the internal use of bool_lax
2338 instead of bool when processing it.
2340 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2341 server certificates when making smtp deliveries.
2343 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2345 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2347 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2349 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2350 Merged patch from Sebastian Wiedenroth.
2352 JH/05 Fix results-pipe from transport process. Several recipients, combined
2353 with certificate use, exposed issues where response data items split
2354 over buffer boundaries were not parsed properly. This eventually
2355 resulted in duplicates being sent. This issue only became common enough
2356 to notice due to the introduction of connection certificate information,
2357 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2359 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2360 size buffer was used, resulting in syntax errors when an expansion
2363 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2364 version 3.3.6 or later.
2366 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2367 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2368 both become "event_action", the variables become $event_name, $event_data
2369 and $event_defer_errno. There is a new variable $verify_mode, usable in
2370 routers, transports and related events. The tls:cert event is now also
2371 raised for inbound connections, if the main configuration event_action
2374 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2375 early OCSP support, but no stapling (appears to be less than 1.0.0).
2377 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2378 server certificate names available under the smtp transport option
2379 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2382 JH/10 Time-related extraction expansions from certificates now use the main
2383 option "timezone" setting for output formatting, and are consistent
2384 between OpenSSL and GnuTLS compilations. Bug 1541.
2386 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2387 encoded parameter in the incoming message. Bug 1558.
2389 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2390 include certificate info, eximon was claiming there were spoolfile
2393 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2395 JH/14 Log delivery-related information more consistently, using the sequence
2396 "H=<name> [<ip>]" wherever possible.
2398 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2399 are problematic for Debian distribution, omit them from the release
2402 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2404 JH/16 Fix string representation of time values on 64bit time_t architectures.
2407 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2408 output list separator was used.
2413 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2414 checkers that were complaining about end of non-void function with no
2417 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2418 This was a regression introduced in 4.83 by another bugfix.
2420 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2422 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2423 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2429 TF/01 Correctly close the server side of TLS when forking for delivery.
2431 When a message was received over SMTP with TLS, Exim failed to clear up
2432 the incoming connection properly after forking off the child process to
2433 deliver the message. In some situations the subsequent outgoing
2434 delivery connection happened to have the same fd number as the incoming
2435 connection previously had. Exim would try to use TLS and fail, logging
2436 a "Bad file descriptor" error.
2438 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2439 utilities have not been installed.
2441 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2442 temporary space as the ACL may create new global variables.
2444 TL/01 LDAP support uses per connection or global context settings, depending
2445 upon the detected version of the libraries at build time.
2447 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2448 to extract and use the src ip:port in logging and expansions as if it
2449 were a direct connection from the outside internet. PPv2 support was
2450 updated based on HAProxy spec change in May 2014.
2452 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2454 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2455 Properly escape header and check for NULL return.
2457 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2460 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2462 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2463 characters in header names, implemented as a verify condition.
2464 Contributed by Michael Fischer v. Mollard.
2466 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2467 results permerror and temperror. Previous values are deprecated but
2468 still accepted. In a future release, err_perm and err_temp will be
2469 completely removed, which will be a backward incompatibility if the
2470 ACL tests for either of these two old results. Patch contributed by
2471 user bes-internal on the mailing list.
2473 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2475 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2476 selectors, in both main and reject logs.
2478 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2481 JH/07 Add malware type "sock" for talking to simple daemon.
2483 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2485 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2486 routers/transports under cutthrough routing.
2488 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2489 numbers. Touch up "bool" conditional to keep the same definition.
2491 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2493 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2495 JH/12 Expand items in router/transport headers_add or headers_remove lists
2496 individually rather than the list as a whole. Bug 1452.
2498 Required for reasonable handling of multiple headers_ options when
2499 they may be empty; requires that headers_remove items with embedded
2500 colons must have them doubled (or the list-separator changed).
2502 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2503 view the policy declared in the DMARC record. Currently, $dmarc_status
2504 is a combined value of both the record presence and the result of the
2507 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2509 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2510 dnslookup router and the smtp transport (applying to the forward
2513 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2514 of ldap servers used for a specific lookup. Patch provided by Heiko
2517 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2518 New variable $lookup_dnssec_authenticated for observability.
2520 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2521 Patch submitted by Lars Timman.
2523 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2525 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2526 Requires trusted mode and valid format message id, aborts otherwise.
2527 Patch contributed by Heiko Schlichting.
2529 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2530 certextract with support for various fields. Bug 1358.
2532 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2533 is requested by default, modifiable by smtp transport option
2536 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2537 operate on certificate variables to give certificate fingerprints
2538 Also new ${sha256:cert_variable}.
2540 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2542 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2543 Christian Aistleitner.
2545 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2547 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2548 file. Patch from Wolfgang Breyha.
2550 JH/25 Expand the coverage of the delivery $host and $host_address to
2551 client authenticators run in verify callout. Bug 1476.
2553 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2554 align with daemon_smtp_ports. Bug 72.
2556 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2557 support and error reporting did not work properly.
2559 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2560 and is readable. Patch from Andrew Colin Kissa.
2562 TL/14 Enhance documentation of ${run expansion and how it parses the
2563 commandline after expansion, particularly in the case when an
2564 unquoted variable expansion results in an empty value.
2566 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2568 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2569 by renaming away from it. Addresses GH issue 15, reported by
2572 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2574 TL/15 SECURITY: prevent double expansion in math comparison functions
2575 (can expand unsanitized data). Not remotely exploitable.
2582 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2584 PP/02 Make -n do something, by making it not do something.
2585 When combined with -bP, the name of an option is not output.
2587 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2590 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2591 $sender_host_name and config options to manage this, and basic check
2594 PP/05 DSCP support for outbound connections and control modifier for inbound.
2596 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2597 (Only plugin which currently uses this is kerberos4, which nobody should
2598 be using, but we should make it available and other future plugins might
2599 conceivably use it, even though it would break NAT; stuff *should* be
2600 using channel bindings instead).
2602 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2603 name; added for Sendmail compatibility; requires admin caller.
2604 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2605 just ignore it); requires trusted caller.
2606 Also parse but ignore: -Ac -Am -X<logfile>
2609 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2611 TL/02 Add +smtp_confirmation as a default logging option.
2613 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2614 Patch by Magnus Holmgren from 2007-02-20.
2616 TL/04 Bugzilla 1281 - Spec typo.
2617 Bugzilla 1283 - Spec typo.
2618 Bugzilla 1290 - Spec grammar fixes.
2620 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2622 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2624 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2625 mailing list by by Dmitry Isaikin.
2627 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2629 JH/02 Support "G" suffix to numbers in ${if comparisons.
2631 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2633 NM/01 Bugzilla 1197 - Spec typo
2634 Bugzilla 1196 - Spec examples corrections
2636 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2638 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2639 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2642 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2643 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2645 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2646 "acl {{name}{arg}...}", and optional args on acl condition
2649 JH/05 Permit multiple router/transport headers_add/remove lines.
2651 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2653 JH/07 Avoid using a waiting database for a single-message-only transport.
2654 Performance patch from Paul Fisher. Bugzilla 1262.
2656 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2659 JH/09 Add $headers_added variable, with content from use of ACL modifier
2660 add_header (but not yet added to the message). Bugzilla 199.
2662 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2663 Pulled from Bugzilla 817 by Wolfgang Breyha.
2665 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2667 (nb: this is the same fix as in Exim 4.80.1)
2669 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2672 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2675 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2677 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2678 advertise SMTP AUTH mechanism to us, instead of a generic
2679 protocol violation error. Also, make Exim more robust to bad
2680 data from the Dovecot auth socket.
2682 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2684 When a queue runner is handling a message, Exim first routes the
2685 recipient addresses, during which it prunes them based on the retry
2686 hints database. After that it attempts to deliver the message to
2687 any remaining recipients. It then updates the hints database using
2690 So if a recipient address works intermittently, it can get repeatedly
2691 deferred at routing time. The retry hints record remains fresh so the
2692 address never reaches the final cutoff time.
2694 This is a fairly common occurrence when a user is bumping up against
2695 their storage quota. Exim had some logic in its local delivery code
2696 to deal with this. However it did not apply to per-recipient defers
2697 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2699 This change adds a proper retry rule check during routing so that the
2700 final cutoff time is checked against the message's age. We only do
2701 this check if there is an address retry record and there is not a
2702 domain retry record; this implies that previous attempts to handle
2703 the address had the retry_use_local_parts option turned on. We use
2704 this as an approximation for the destination being like a local
2705 delivery, as in LMTP.
2707 I suspect this new check makes the old local delivery cutoff check
2708 redundant, but I have not verified this so I left the code in place.
2710 TF/02 Correct gecos expansion when From: is a prefix of the username.
2712 Test 0254 submits a message to Exim with the header
2716 When I ran the test suite under the user fanf2, Exim expanded
2717 the header to contain my full name, whereas it should have added
2718 a Resent-Sender: header. It erroneously treats any prefix of the
2719 username as equal to the username.
2721 This change corrects that bug.
2723 GF/01 DCC debug and logging tidyup
2724 Error conditions log to paniclog rather than rejectlog.
2725 Debug lines prefixed by "DCC: " to remove any ambiguity.
2727 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2729 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2730 Bug spotted by Jeremy Harris; was flawed since initial commit.
2731 Would have resulted in OCSP responses post-SNI triggering an Exim
2732 NULL dereference and crash.
2734 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2736 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2737 Bug detection, analysis and fix by Samuel Thibault.
2738 Bugzilla 1331, Debian bug #698092.
2740 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2742 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2743 Server implementation by Todd Lyons, client by JH.
2744 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2745 config variable "prdr_enable" controls whether the server
2746 advertises the facility. If the client requests PRDR a new
2747 acl_data_smtp_prdr ACL is called once for each recipient, after
2748 the body content is received and before the acl_smtp_data ACL.
2749 The client is controlled by both of: a hosts_try_prdr option
2750 on the smtp transport, and the server advertisement.
2751 Default client logging of deliveries and rejections involving
2752 PRDR are flagged with the string "PRDR".
2754 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2755 fclose(). Diagnosis by Todd Lyons.
2757 PP/17 Update configure.default to handle IPv6 localhost better.
2758 Patch by Alain Williams (plus minor tweaks).
2761 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2762 This is now consistent with GnuTLS, and is now documented: the
2763 previous undocumented portable approach to treating the option as
2764 unset was to force an expansion failure. That still works, and
2765 an empty string is now equivalent.
2767 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2768 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2769 not performing validation itself.
2771 PP/20 Added force_command boolean option to pipe transport.
2772 Patch from Nick Koston, of cPanel Inc.
2774 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2777 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2779 PP/21 Fix eximon continuous updating with timestamped log-files.
2780 Broken in a format-string cleanup in 4.80, missed when I repaired the
2781 other false fix of the same issue.
2782 Report and fix from Heiko Schlichting.
2785 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2786 Report from Prashanth Katuri.
2788 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2789 It's SecureTransport, so affects any MacOS clients which use the
2790 system-integrated TLS libraries, including email clients.
2792 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2793 using a MIME ACL for non-SMTP local injection.
2794 Report and assistance in diagnosis by Warren Baker.
2796 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2798 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2800 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2801 last id that failed so it may be referenced in subsequent ACL's.
2803 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2806 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2807 ldap library initialization, allowing self-signed CA's to be
2808 used. Also properly sets require_cert option later in code by
2809 using NULL (global ldap config) instead of ldap handle (per
2810 session). Bug diagnosis and testing by alxgomz.
2812 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2813 the src/util/ subdirectory.
2815 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2816 renamed to Transport Post Delivery Action by Jeremy Harris, as
2819 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2820 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2821 redis_servers = needs to be configured which will be used by the redis
2822 lookup. Patch from Warren Baker, of The Packet Hub.
2824 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2826 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2827 hostname or reverse DNS when processing a host list. Used suggestions
2828 from multiple comments on this bug.
2830 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2832 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2833 Missed a few lines, added it to make the runtest require no keyboard
2836 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2837 contains upper case chars. Make router use caseful_local_part.
2839 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2840 support when GnuTLS has been built with p11-kit.
2846 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2848 This, or similar/improved, will also be change PP/11 of 4.82.
2854 PP/01 Handle short writes when writing local log-files.
2855 In practice, only affects FreeBSD (8 onwards).
2856 Bugzilla 1053, with thanks to Dmitry Isaikin.
2858 NM/01 Bugzilla 949 - Documentation tweak
2860 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2863 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2865 PP/02 Implemented gsasl authenticator.
2867 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2869 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2870 `pkg-config foo` for cflags/libs.
2872 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2873 with rest of GSASL and with heimdal_gssapi.
2875 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2876 `pkg-config foo` for cflags/libs for the TLS implementation.
2878 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2879 properties get this fed in as external SSF. A number of robustness
2880 and debugging improvements to the cyrus_sasl authenticator.
2882 PP/08 cyrus_sasl server now expands the server_realm option.
2884 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2885 Patch by Jeremy Harris.
2887 PP/10 Added dbmjz lookup type.
2889 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2891 PP/12 MAIL args handles TAB as well as SP, for better interop with
2892 non-compliant senders.
2893 Analysis and variant patch by Todd Lyons.
2895 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2896 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2897 Patch from Dirk Mueller <dmueller@suse.com>
2899 PP/13 tls_peerdn now print-escaped for spool files.
2900 Observed some $tls_peerdn in wild which contained \n, which resulted
2901 in spool file corruption.
2903 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2904 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2905 or write after TLS renegotiation, which otherwise led to messages
2908 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2909 as a tracking header (ie: a signed header comes before the signature).
2910 Patch from Wolfgang Breyha.
2912 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2913 comma-sep list; embedded commas doubled.
2915 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2917 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2919 Report and patch from Dmitry Banschikov.
2921 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2922 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2923 failures. We appear to now support TLS1.1+ with Exim.
2925 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2926 lets Exim select keys and certificates based upon TLS SNI from client.
2927 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2928 before an outbound SMTP session. New log_selector, +tls_sni.
2930 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2931 NULL dereference. Report and patch from Alun Jones.
2933 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2934 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2935 Not seeing resolver debug output on NetBSD, but suspect this is a
2936 resolver implementation change.
2938 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2939 Left warnings. Added "eximon gdb" invocation mode.
2941 PP/21 Defaulting "accept_8bitmime" to true, not false.
2943 PP/22 Added -bw for inetd wait mode support.
2945 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2946 locate the relevant includes and libraries. Made this the default.
2948 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2949 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2951 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2952 This may cause build issues on older platforms.
2954 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2955 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2956 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2957 Added SNI support via GnuTLS too.
2958 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2960 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2962 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2963 Applied second patch from Janne, implementing suggestion to default
2964 multiple-strings-in-record handling to match SPF spec.
2966 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2968 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2969 read-only, out of scope).
2970 Patch from Wolfgang Breyha, report from Stuart Northfield.
2972 PP/29 Fix three issues highlighted by clang analyser static analysis.
2973 Only crash-plausible issue would require the Cambridge-specific
2974 iplookup router and a misconfiguration.
2975 Report from Marcin Mirosław.
2977 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2979 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2980 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2981 As part of this, removing so much warning spew let me fix some minor
2982 real issues in debug logging.
2984 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2985 assignment on my part. Fixed.
2987 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2988 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2989 Janne Snabb (who went above and beyond: thank you).
2991 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2992 string otherwise requires a connection and a bunch more work and it's
2993 relatively easy to get wrong. Should also expose TLS library linkage
2996 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2997 64-bit ${eval} (JH/03).
2999 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3000 GNU libc to support some of the 64-bit stuff, should not lead to
3001 conflicts. Defined before os.h is pulled in, so if a given platform
3002 needs to override this, it can.
3004 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3005 protection layer was required, which is not implemented.
3006 Bugzilla 1254, patch from Wolfgang Breyha.
3008 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3009 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3010 tls_dhparam take prime identifiers. Also unbreak combination of
3011 OpenSSL+DH_params+TLSSNI.
3013 PP/39 Disable SSLv2 by default in OpenSSL support.
3019 PP/01 Solaris build fix for Oracle's LDAP libraries.
3020 Bugzilla 1109, patch from Stephen Usher.
3022 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3024 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3027 TF/02 Fix a couple more cases where we did not log the error message
3028 when unlink() failed. See also change 4.74-TF/03.
3030 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3031 lock up or crash if it happened to be inside a call to libc when it
3032 got a SIGUSR1 from exiwhat.
3034 The SIGUSR1 handler appends the current process status to the process
3035 log which is later printed by exiwhat. It used to use the general
3036 purpose logging code to do this, but several functions it calls are
3037 not safe for signals.
3039 The new output code in the SIGUSR1 handler is specific to the process
3040 log, and simple enough that it's easy to inspect for signal safety.
3041 Removing some special cases also simplifies the general logging code.
3042 Removing the spurious timestamps from the process log simplifies
3045 TF/04 Improved ratelimit ACL condition.
3047 The /noupdate option has been deprecated in favour of /readonly which
3048 has clearer semantics. The /leaky, /strict, and /readonly update modes
3049 are mutually exclusive. The update mode is no longer included in the
3050 database key; it just determines when the database is updated. (This
3051 means that when you upgrade Exim will forget old rate measurements.)
3053 Exim now checks that the per_* options are used with an update mode that
3054 makes sense for the current ACL. For example, when Exim is processing a
3055 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3056 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3057 must specify per_mail/readonly. If you omit the update mode it defaults to
3058 /leaky where that makes sense (as before) or /readonly where required.
3060 The /noupdate option is now undocumented but still supported for
3061 backwards compatibility. It is equivalent to /readonly except that in
3062 ACLs where /readonly is required you may specify /leaky/noupdate or
3063 /strict/noupdate which are treated the same as /readonly.
3065 A useful new feature is the /count= option. This is a generalization
3066 of the per_byte option, so that you can measure the throughput of other
3067 aggregate values. For example, the per_byte option is now equivalent
3068 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3070 The per_rcpt option has been generalized using the /count= mechanism
3071 (though it's more complicated than the per_byte equivalence). When it is
3072 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3073 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3074 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3075 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3076 Note that using per_rcpt with a non-readonly update mode in more than
3077 one ACL will cause the recipients to be double-counted. (The per_mail
3078 and per_byte options don't have this problem.)
3080 The handling of very low rates has changed slightly. If the computed rate
3081 is less than the event's count (usually one) then this event is the first
3082 after a long gap. In this case the rate is set to the same as this event's
3083 count, so that the first message of a spam run is counted properly.
3085 The major new feature is a mechanism for counting the rate of unique
3086 events. The new per_addr option counts the number of different
3087 recipients that someone has sent messages to in the last time period. It
3088 behaves like per_rcpt if all the recipient addresses are different, but
3089 duplicate recipient addresses do not increase the measured rate. Like
3090 the /count= option this is a general mechanism, so the per_addr option
3091 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3092 example, measure the rate that a client uses different sender addresses
3093 with the options per_mail/unique=$sender_address. There are further
3094 details in the main documentation.
3096 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3098 TF/06 Removed a few PCRE remnants.
3100 TF/07 Automatically extract Exim's version number from tags in the git
3101 repository when doing development or release builds.
3103 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3104 Bugzilla 879. Patch from Paul Fisher.
3106 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3107 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3110 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3112 PP/05 Variable $av_failed, true if the AV scanner deferred.
3113 Bugzilla 1078. Patch from John Horne.
3115 PP/06 Stop make process more reliably on build failure.
3116 Bugzilla 1087. Patch from Heiko Schlittermann.
3118 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3119 Bugzilla 1089. Patch from Heiko Schlittermann.
3121 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3122 Bugzilla 1131. Patch from Holger Weiß.
3124 PP/09 Handle IPv6 addresses with SPF.
3125 Bugzilla 860. Patch from Wolfgang Breyha.
3127 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3129 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3132 PP/11 match_* no longer expand right-hand-side by default.
3133 New compile-time build option, EXPAND_LISTMATCH_RHS.
3134 New expansion conditions, "inlist", "inlisti".
3136 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3138 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3140 PP/14 fix log_write() format string regression from TF/03.
3141 Bugzilla 1152. Patch from Dmitry Isaikin.
3147 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3149 PP/02 Harmonised TLS library version reporting; only show if debugging.
3150 Layout now matches that introduced for other libraries in 4.74 PP/03.
3152 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3154 PP/04 New "dns_use_edns0" global option.
3156 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3159 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3160 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3162 TK/01 Updated PolarSSL code to 0.14.2.
3163 Bugzilla 1097. Patch from Andreas Metzler.
3165 PP/07 Catch divide-by-zero in ${eval:...}.
3166 Fixes bugzilla 1102.
3168 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3171 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3172 format-string attack -- SECURITY: remote arbitrary code execution.
3174 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3175 time unintentionally subject to list matching rules, letting the header
3176 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3177 arbitrary string expansion). This allowed for information disclosure.
3179 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3180 INT_MIN/-1 -- value coerced to INT_MAX.
3186 NM/01 Workaround for PCRE version dependency in version reporting
3189 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3190 This fixes portability to compilers other than gcc, notably
3191 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3193 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3194 makefiles for portability to HP-UX and POSIX correctness.
3196 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3197 Also via indented variable definition in the Makefile.
3198 (Debugging by Oliver Heesakkers).
3200 PP/02 Restore caching of spamd results with expanded spamd_address.
3201 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3203 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3204 Improves build reliability. Fix from: Frank Elsner
3206 NM/02 Fix wide character breakage in the rfc2047 coding
3207 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3209 NM/03 Allow underscore in dnslist lookups
3210 Fixes bug 1026. Patch from Graeme Fowler
3212 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3213 Code patches from Adam Ciarcinski of NetBSD.
3215 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3218 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3219 is logged, to avoid truncation. Patch from John Horne.
3221 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3222 Patch from Jakob Hirsch.
3224 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3225 SQL string expansion failure details.
3226 Patch from Andrey Oktyabrski.
3228 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3229 Patch from Simon Arlott.
3231 PP/09 New lookups functionality failed to compile on old gcc which rejects
3232 extern declarations in function scope.
3233 Patch from Oliver Fleischmann
3235 PP/10 Use sig_atomic_t for flags set from signal handlers.
3236 Check getgroups() return and improve debugging.
3237 Fixed developed for diagnosis in bug 927 (which turned out to be
3240 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3241 Patch from Mark Zealey.
3243 PP/12 Bugzilla 1056: Improved spamd server selection.
3244 Patch from Mark Zealey.
3246 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3247 Based on patch from Heiko Schlittermann.
3249 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3250 Patch from Uwe Doering, sign-off by Michael Haardt.
3252 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3253 variable declaration deep within a block. Bug and patch from
3256 PP/15 lookups-Makefile IRIX compatibility coercion.
3258 PP/16 Make DISABLE_DKIM build knob functional.
3260 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3261 Patch by Simon Arlott
3263 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3264 variable argument macros. Our copy now differs from upstream.
3270 TF/01 Failure to get a lock on a hints database can have serious
3271 consequences so log it to the panic log.
3273 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3274 controlled using the smtp_confirmation log selector.
3276 TF/03 Include the error message when we fail to unlink a spool file.
3278 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3279 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3280 for maintaining out-of-tree patches for some time.
3282 PP/01 Bugzilla 139: Documentation and portability issues.
3283 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3284 Handle per-OS dynamic-module compilation flags.
3286 PP/02 Let /dev/null have normal permissions.
3287 The 4.73 fixes were a little too stringent and complained about the
3288 permissions on /dev/null. Exempt it from some checks.
3289 Reported by Andreas M. Kirchwitz.
3291 PP/03 Report version information for many libraries, including
3292 Exim version information for dynamically loaded libraries. Created
3293 version.h, now support a version extension string for distributors
3294 who patch heavily. Dynamic module ABI change.
3296 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3297 privilege escalation vulnerability whereby the Exim run-time user
3298 can cause root to append content of the attacker's choosing to
3301 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3304 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3305 If dropping privileges for untrusted macros, we disabled normal logging
3306 on the basis that it would fail; for the Exim run-time user, this is not
3307 the case, and it resulted in successful deliveries going unlogged.
3308 Fixed. Reported by Andreas Metzler.
3314 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3315 only prepend for the Resent-* case. Fixes regression introduced in
3316 Exim 4.70 by NM/22 for Bugzilla 607.
3318 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3319 increasing numbers of administrators be bitten by this.
3321 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3323 PP/03 Bugzilla 994: added openssl_options main configuration option.
3325 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3327 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3329 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3331 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3332 without a peer certificate, leading to a segfault because of an
3333 assumption that peers always have certificates. Be a little more
3334 paranoid. Problem reported by Martin Tscholak.
3336 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3337 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3338 NB: ClamAV planning to remove STREAM in "middle of 2010".
3339 CL also introduces -bmalware, various -d+acl logging additions and
3340 more caution in buffer sizes.
3342 PP/09 Implemented reverse_ip expansion operator.
3344 PP/10 Bugzilla 937: provide a "debug" ACL control.
3346 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3348 PP/12 Bugzilla 973: Implement --version.
3350 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3352 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3354 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3356 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3357 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3358 ignore trailing whitespace.
3360 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3362 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3365 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3366 Notification from Dr Andrew Aitchison.
3368 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3369 ExtendedDetectionInfo response format.
3370 Notification from John Horne.
3372 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3375 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3376 XSL and documented dependency on system catalogs, with examples of how
3379 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3382 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3383 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3384 configuration file which is writeable by the Exim user or group.
3386 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3387 of configuration files to cover files specified with the -C option if
3388 they are going to be used with root privileges, not just the default
3391 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3392 option (effectively making it always true).
3394 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3395 files to be used while preserving root privileges.
3397 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3398 that rogue child processes cannot use them.
3400 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3401 run-time user, instead of root.
3403 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3404 Exim run-time user without dropping privileges.
3406 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3407 result string, instead of calling string_vformat() twice with the same
3410 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3411 for other users. Others should always drop root privileges if they use
3412 -C on the command line, even for a whitelisted configure file.
3414 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3416 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3422 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3423 $data_path, and $header_path variables; fixed documentation bugs and
3426 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3427 exipick to access non-standard spools, including the "frozen" queue
3430 NM/01 Bugzilla 965: Support mysql stored procedures.
3431 Patch from Alain Williams
3433 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3435 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3436 Patch from Andreas Metzler
3438 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3439 Patch from Kirill Miazine
3441 NM/05 Bugzilla 671: Added umask to procmail example.
3443 JJ/03 installed exipick 20100323.0, fixing doc bug
3445 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3446 directory. Notification and patch from Dan Rosenberg.
3448 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3450 TK/02 Improve log output when DKIM signing operation fails.
3452 MH/01 Treat the transport option dkim_domain as a colon separated
3453 list, not as a single string, and sign the message with each element,
3454 omitting multiple occurences of the same signer.
3456 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3457 Bugzilla 985, 986. Patch by Simon Arlott
3459 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3460 Patch by Simon Arlott
3462 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3463 MBX locking. Notification from Dan Rosenberg.
3469 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3471 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3473 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3475 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3477 NM/04 Bugzilla 915: Flag broken perl installation during build.
3483 TK/01 Added patch by Johannes Berg that expands the main option
3484 "spamd_address" if it starts with a dollar sign.
3486 TK/02 Write list of recipients to X-Envelope-Sender header when building
3487 the mbox-format spool file for content scanning (suggested by Jakob
3490 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3491 (http://www.dcc-servers.net/) support via dccifd. Activated by
3492 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3494 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3495 by Mark Daniel Reidel <mr@df.eu>.
3497 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3498 When building exim an external PCRE library is now needed -
3499 PCRE is a system library on the majority of modern systems.
3500 See entry on PCRE_LIBS in EDITME file.
3502 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3503 conversation. Added nologin parameter to request.
3504 Patch contributed by Kirill Miazine.
3506 TF/01 Do not log submission mode rewrites if they do not change the address.
3508 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3510 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3511 log files in place. Contributed by Roberto Lima.
3513 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3515 TF/03 Bugzilla 615: When checking the local_parts router precondition
3516 after a local_part_suffix or local_part_prefix option, Exim now
3517 does not use the address's named list lookup cache, since this
3518 contains cached lookups for the whole local part.
3520 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3521 Robert Millan. Documentation is in experimental-spec.txt.
3523 TF/04 Bugzilla 668: Fix parallel build (make -j).
3525 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3527 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3528 Patch provided by Jan Srzednicki.
3530 TF/05 Leading white space used to be stripped from $spam_report which
3531 wrecked the formatting. Now it is preserved.
3533 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3534 that they are available at delivery time.
3536 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3538 TF/08 TLS error reporting now respects the incoming_interface and
3539 incoming_port log selectors.
3541 TF/09 Produce a more useful error message if an SMTP transport's hosts
3542 setting expands to an empty string.
3544 NM/06 Bugzilla 744: EXPN did not work under TLS.
3545 Patch provided by Phil Pennock.
3547 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3548 Patch provided by Richard Godbee.
3550 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3551 acl_smtp_notquit, added index entry.
3553 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3554 Patch provided by Eugene Bujak.
3556 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3557 accept(). Patch provided by Maxim Dounin.
3559 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3560 Patch provided by Phil Pennock.
3562 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3564 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3565 Patch provided by Brad "anomie" Jorsch.
3567 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3568 Patch provided by Dean Brooks.
3570 TK/05 Add native DKIM support (does not depend on external libraries).
3572 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3573 Patch provided by Graeme Fowler.
3575 NM/16 Bugzilla 851: Documentation example syntax fix.
3577 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3579 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3582 NM/19 Bugzilla 745: TLS version reporting.
3583 Patch provided by Phil Pennock.
3585 NM/20 Bugzilla 167: bool: condition support.
3586 Patch provided by Phil Pennock.
3588 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3589 clients. Patch provided by Phil Pennock.
3591 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3592 Patch provided by Brad "anomie" Jorsch.
3594 NM/23 Bugzilla 687: Fix misparses in eximstats.
3595 Patch provided by Heiko Schlittermann.
3597 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3598 Patch provided by Heiko Schlittermann.
3600 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3601 plus update to original patch.
3603 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3605 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3606 Patch provided by David Brownlee.
3608 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3610 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3612 NM/30 Bugzilla 888: TLS documentation bugfixes.
3614 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3616 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3617 Unlike the original bugzilla I have changed all shell scripts in src tree.
3619 NM/33 Bugzilla 898: Transport filter timeout fix.
3620 Patch by Todd Rinaldo.
3622 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3623 Patch by Serge Demonchaux.
3625 NM/35 Bugzilla 39: Base64 decode bug fixes.
3626 Patch by Jakob Hirsch.
3628 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3630 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3632 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3634 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3640 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3641 ALT-N's libdkim that I have put here:
3642 http://duncanthrax.net/exim-experimental/
3644 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3645 were called 'true' and it seems that C99 defines that as a reserved
3646 keyword to be used with 'bool' variable types. That means you could
3647 not include C99-style headers which use bools without triggering
3648 build errors in sieve.c.
3650 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3651 as mailq or other aliases. Changed the --help handling significantly
3652 to do whats expected. exim_usage() emits usage/help information.
3654 SC/01 Added the -bylocaldomain option to eximstats.
3656 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3658 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3660 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3666 PH/01 Another patch from the Sieve maintainer.
3668 PH/02 When an IPv6 address is converted to a string for single-key lookup
3669 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3670 dots are used instead of colons so that keys in lsearch files need not
3671 contain colons. This was done some time before quoting was made available
3672 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3673 (notated using the quote facility) so as to distinguish them from IPv4
3674 keys. This meant that lookups for IP addresses in host lists did not work
3675 for iplsearch lookups.
3677 This has been fixed by arranging for IPv6 addresses to be expressed with
3678 colons if the lookup type is iplsearch. This is not incompatible, because
3679 previously such lookups could never work.
3681 The situation is now rather anomalous, since one *can* have colons in
3682 ordinary lsearch keys. However, making the change in all cases is
3683 incompatible and would probably break a number of configurations.
3685 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3688 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3689 conversion specification without a maximum field width, thereby enabling
3690 a rogue spamd server to cause a buffer overflow. While nobody in their
3691 right mind would setup Exim to query an untrusted spamd server, an
3692 attacker that gains access to a server running spamd could potentially
3693 exploit this vulnerability to run arbitrary code as the Exim user.
3695 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3696 $primary_hostname instead of what libspf2 thinks the hosts name is.
3698 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3699 a directory entry by the name of the lookup key. Previously, if a
3700 symlink pointed to a non-existing file or a file in a directory that
3701 Exim lacked permissions to read, a lookup for a key matching that
3702 symlink would fail. Now it is enough that a matching directory entry
3703 exists, symlink or not. (Bugzilla 503.)
3705 PH/03 The body_linecount and body_zerocount variables are now exported in the
3708 PH/04 Added the $dnslist_matched variable.
3710 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3711 This means they are set thereafter only if the connection becomes
3714 PH/06 Added the client_condition to authenticators so that some can be skipped
3715 by clients under certain conditions.
3717 PH/07 The error message for a badly-placed control=no_multiline_responses left
3718 "_responses" off the end of the name.
3720 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3722 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3723 (without spaces) instead of just copying the configuration text.
3725 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3727 PH/11 Added $max_received_linelength.
3729 PH/12 Added +ignore_defer and +include_defer to host lists.
3731 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3732 way in which PCRE > 7.0 is built.
3734 PH/14 Implemented queue_only_load_latch.
3736 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3737 MAIL command. The effect was to mangle the value on 64-bit systems.
3739 PH/16 Another patch from the Sieve maintainer.
3741 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3743 PH/18 If a system quota error occurred while trying to create the file for
3744 a maildir delivery, the message "Mailbox is full" was not appended to the
3745 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3746 only to a quota excession during the actual writing of the file.
3748 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3749 characters?) which causes problems in log lines. The DN values are now
3750 passed through string_printing() before being added to log lines.
3752 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3753 and InterBase are left for another time.)
3755 PH/21 Added message_body_newlines option.
3757 PH/22 Guard against possible overflow in moan_check_errorcopy().
3759 PH/23 POSIX allows open() to be a macro; guard against that.
3761 PH/24 If the recipient of an error message contained an @ in the local part
3762 (suitably quoted, of course), incorrect values were put in $domain and
3763 $local_part during the evaluation of errors_copy.
3769 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3770 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3773 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3774 issue a MAIL command.
3776 PH/02 In an ACL statement such as
3778 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3780 if a client was not listed at all, or was listed with a value other than
3781 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3782 the condition was not true (as it should be), so access was not denied.
3783 The bug was that the ! inversion was incorrectly passed on to the second
3784 item. This has been fixed.
3786 PH/03 Added additional dnslists conditions == and =& which are different from
3787 = and & when the dns lookup returns more than one IP address.
3789 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3790 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3792 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3793 FSYNC, which compiles an option called disable_fsync that allows for
3794 bypassing fsync(). The documentation is heavily laced with warnings.
3796 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3798 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3799 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3800 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3801 including adding "make clean"; (3) Added -fPIC when compiling the test
3802 dynamically loaded module, to get rid of a warning.
3804 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3805 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3806 = 0s. The bug is otherwise harmless.
3808 PH/07 There was a bug in the dovecot authenticator such that the value of
3809 $auth1 could be overwritten, and so not correctly preserved, after a
3810 successful authentication. This usually meant that the value preserved by
3811 the server_setid option was incorrect.
3813 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3815 PH/09 Installed PCRE release 7.0.
3817 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3818 run for batched SMTP input. It is now run at the start of every message
3819 in the batch. While fixing this I discovered that the process information
3820 (output by running exiwhat) was not always getting set for -bs and -bS
3821 input. This is fixed, and it now also says "batched" for BSMTP.
3823 PH/11 Added control=no_pipelining.
3825 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3826 patch, slightly modified), and move the expansion of helo_data till after
3827 the connection is made in the smtp transport (so it can use these
3830 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3832 PH/14 Added log_selector = +pid.
3834 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3836 PH/16 Add ${if forany and ${if forall.
3838 PH/17 Added dsn_from option to vary the From: line in DSNs.
3840 PH/18 Flush SMTP output before performing a callout, unless control =
3841 no_callout_flush is set.
3843 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3844 was true (the default) a successful delivery failed to delete the retry
3845 item, thus causing premature timeout of the address. The bug is now
3848 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3850 PH/21 Long custom messages for fakedefer and fakereject are now split up
3851 into multiline responses in the same way that messages for "deny" and
3852 other ACL rejections are.
3854 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3855 with slight modification.
3857 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3858 draft, changing the syntax and factoring some duplicate code".
3860 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3861 for deliveries of the second and subsequent messages over the same SMTP
3864 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3865 ${reduce, with only minor "tidies".
3867 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3869 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3870 expansion side effects.
3872 PH/27 When a message times out after an over-quota error from an Exim-imposed
3873 quota, the bounce message says "mailbox is full". This message was not
3874 being given when it was a system quota that was exceeded. It now should
3877 MH/03 Made $recipients available in local_scan(). local_scan() already has
3878 better access to the recipient list through recipients_list[], but
3879 $recipients can be useful in postmaster-provided expansion strings.
3881 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3882 in the case of a MAIL command with additional options following the
3883 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3884 were accidentally chopped off.
3886 PH/29 SMTP synchronization checks are implemented when a command is read -
3887 there is a check that no more input is waiting when there shouldn't be
3888 any. However, for some commands, a delay in an ACL can mean that it is
3889 some time before the response is written. In this time, more input might
3890 arrive, invalidly. So now there are extra checks after an ACL has run for
3891 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3892 pipelining has not been advertised.
3894 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3896 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3897 returned in the SMTP response when smtp_return_error_details was false.
3898 This has been fixed.
3900 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3901 instead of the C I/O that was originally supplied, because problems were
3902 reported on Solaris.
3904 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3905 Exim which did not show up earlier: it was assuming that a call to
3906 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3907 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3908 was a macro that became an assignment, so it seemed to work. This has
3909 changed to a proper function call with a void return, hence the compile
3910 error. Exim's code has been fixed.
3912 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3915 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3917 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3919 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3920 because it turns out to be too big or there is a timeout) there is no
3921 "Completed" line in the log. When some messages of this type were
3922 selected by exigrep, they were listed as "not completed". Others were
3923 picked up by some special patterns. I have improved the selection
3924 criteria to be more general.
3926 PH/38 The host_find_failed option in the manualroute router can now be set
3927 to "ignore", to completely ignore a host whose IP address cannot be
3928 found. If all hosts are ignored, the behaviour is controlled by the new
3929 host_all_ignored option.
3931 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3932 homing or because of multiple MX records with /mx) generated more than
3933 one IP address, and the following item turned out to be the local host,
3934 all the secondary addresses of the first item were incorrectly removed
3935 from the list, along with the local host and any following hosts (which
3936 is what is supposed to happen).
3938 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3939 whoever called Exim into the -H file. In the case of the daemon it was
3940 behaving confusingly. When first started, it used values for whoever
3941 started the daemon, but after a SIGHUP it used the Exim user (because it
3942 calls itself on a restart). I have changed the code so that it now always
3945 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3946 message are rejected with the same error (e.g. no authentication or bad
3947 sender address), and a DATA command is nevertheless sent (as can happen
3948 with PIPELINING or a stupid MUA), the error message that was given to the
3949 RCPT commands is included in the rejection of the DATA command. This is
3950 intended to be helpful for MUAs that show only the final error to their
3953 PH/42 Another patch from the Sieve maintainer.
3955 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3956 Eximstats - Fixed some broken HTML links and added missing column headers
3958 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3959 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3961 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3963 PH/43 Yet another patch from the Sieve maintainer.
3965 PH/44 I found a way to check for a TCP/IP connection going away before sending
3966 the response to the final '.' that terminates a message, but only in the
3967 case where the client has not sent further data following the '.'
3968 (unfortunately, this is allowed). However, in many cases there won't be
3969 any further data because there won't be any more messages to send. A call
3970 to select() can be used: if it shows that the input is "ready", there is
3971 either input waiting, or the socket has been closed. An attempt to read
3972 the next input character can distinguish the two cases. Previously, Exim
3973 would have sent an OK response which the client would never have see.
3974 This could lead to message repetition. This fix should cure that, at
3975 least in a lot of common cases.
3977 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3978 advertised in response to EHLO.
3984 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3985 fixed by 4.65/MH/01 (is this a record?) are fixed:
3987 (i) An empty string was always treated as zero by the numeric comparison
3988 operators. This behaviour has been restored.
3990 (ii) It is documented that the numeric comparison operators always treat
3991 their arguments as decimal numbers. This was broken in that numbers
3992 starting with 0 were being interpreted as octal.
3994 While fixing these problems I realized that there was another issue that
3995 hadn't been noticed. Values of message_size_limit (both the global option
3996 and the transport option) were treated as octal if they started with 0.
3997 The documentation was vague. These values are now always treated as
3998 decimal, and I will make that clear in the documentation.
4004 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4005 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4008 MH/01 Don't check that the operands of numeric comparison operators are
4009 integers when their expansion is in "skipping" mode (fixes bug
4010 introduced by 4.64-PH/07).
4012 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4013 child addresses, Exim now panics and dies. Previously, because the count
4014 is held in a short int, deliveries were likely to be lost. As such a
4015 large number of recipients for a single message is ridiculous
4016 (performance will be very, very poor), I have chosen to impose a limit
4017 rather than extend the field.
4023 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4024 leftover -K file (the existence of which was triggered by #402).
4025 While we were at it, introduced process PID as part of the -K
4026 filename. This should rule out race conditions when creating
4029 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4030 processing considerably. Previous code took too long for large mails,
4031 triggering a timeout which in turn triggers #401.
4033 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4034 in the DK code in transports.c. sendfile() is not really portable,
4035 hence the _LINUX specificness.
4037 TF/01 In the add_headers option to the mail command in an Exim filter,
4038 there was a bug that Exim would claim a syntax error in any
4039 header after the first one which had an odd number of characters
4042 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4043 callout verification, Exim cached a "reject" for the entire domain. This
4044 is correct for most verifications, but it is not correct for a recipient
4045 verification with use_sender or use_postmaster set, because in that case
4046 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4047 case of MAIL FROM:<> rejection from other early rejections (e.g.
4048 rejection of HELO). When verifying a recipient using a non-null MAIL
4049 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4050 Whatever the result of the callout, the value of the domain cache is
4051 left unchanged (for any other kind of callout, getting as far as trying
4052 RCPT means that the domain itself is ok).
4054 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4057 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4058 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4059 session, the daemon ignores ECONNECT errors and logs others; it now
4060 ignores EPIPE as well.
4062 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4063 (quoted-printable decoding).
4065 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4066 later the small subsequent patch to fix an introduced bug.
4068 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4070 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4072 PH/08 An error is now given if message_size_limit is specified negative.
4074 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4075 to be given (somewhat) arbitrary names.
4077 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4080 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4081 miscellaneous code fixes
4083 PH/10 Added the log_reject_target ACL modifier to specify where to log
4086 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4087 hostname. This is wrong, because it relates to the incoming message (and
4088 probably the interface on which it is arriving) and not to the outgoing
4089 callout (which could be using a different interface). This has been
4090 changed to use the value of the helo_data option from the smtp transport
4091 instead - this is what is used when a message is actually being sent. If
4092 there is no remote transport (possible with a router that sets up host
4093 addresses), $smtp_active_hostname is used.
4095 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4096 tweaks were necessary in order to get it to work (see also 21 below):
4097 (a) The code assumed that strncpy() returns a negative number on buffer
4098 overflow, which isn't the case. Replaced with Exim's string_format()
4100 (b) There were several signed/unsigned issues. I just did the minimum
4101 hacking in of casts. There is scope for a larger refactoring.
4102 (c) The code used strcasecmp() which is not a standard C function.
4103 Replaced with Exim's strcmpic() function.
4104 (d) The code set only $1; it now sets $auth1 as well.
4105 (e) A simple test gave the error "authentication client didn't specify
4106 service in request". It would seem that Dovecot has changed its
4107 interface. Fortunately there's a specification; I followed it and
4108 changed what the client sends and it appears to be working now.
4110 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4113 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4114 address A is aliased to B and C, where B exists and C does not. Without
4115 -v the output is "A verified" because verification stops after a
4116 successful redirection if more than one address is generated. However,
4117 with -v the child addresses are also verified. Exim was outputting "A
4118 failed to verify" and then showing the successful verification for C,
4119 with its parentage. It now outputs "B failed to verify", showing B's
4120 parentage before showing the successful verification of C.
4122 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4123 look up a TXT record in a specific list after matching in a combined
4126 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4127 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4128 they consult the DNS. I had assumed they would set it the way they
4129 wanted; and indeed my experiments on Linux seem to show that in some
4130 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4131 To be on the safe side, however, I have now made the interface to
4132 host_find_byname() similar to host_find_bydns(), with an argument
4133 containing the DNS resolver options. The host_find_byname() function now
4134 sets these options at its start, just as host_find_bydns() does. The smtp
4135 transport options dns_qualify_single and dns_search_parents are passed to
4136 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4137 of host_find_byname() use the default settings of RES_DEFNAMES
4138 (qualify_single) but not RES_DNSRCH (search_parents).
4140 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4141 spool_read_header() do less string testing, by means of a preliminary
4142 switch on the second character of optional "-foo" lines. (This is
4143 overdue, caused by the large number of possibilities that now exist.
4144 Originally there were few.) While I was there, I also converted the
4145 str(n)cmp tests so they don't re-test the leading "-" and the first
4146 character, in the hope this might squeeze out yet more improvement.
4148 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4149 flag allowing group syntax was set by the header_syntax check but not
4150 turned off, possible causing trouble later; (2) The flag was not being
4151 set at all for the header_verify test, causing "group"-style headers to
4152 be rejected. I have now set it in this case, and also caused header_
4153 verify to ignore an empty address taken from a group. While doing this, I
4154 came across some other cases where the code for allowing group syntax
4155 while scanning a header line wasn't quite right (mostly, not resetting
4156 the flag correctly in the right place). These bugs could have caused
4157 trouble for malformed header lines. I hope it is now all correct.
4159 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4160 with the "reply" argument non-NULL. The code, however (which originally
4161 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4162 but it didn't always do it. This confused somebody who was copying the
4163 code for some other use. I have removed all the tests.
4165 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4166 feature that was used to support insecure browsers during the U.S. crypto
4167 embargo. It requires special client support, and Exim is probably the
4168 only MTA that supported it -- and would never use it because real RSA is
4169 always available. This code has been removed, because it had the bad
4170 effect of slowing Exim down by computing (never used) parameters for the
4171 RSA_EXPORT functionality.
4173 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4174 authenticator to fail if there's a tab character in the incoming data
4175 (there should never be unless someone is messing about, as it's supposed
4176 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4177 if the connection is using TLS or if the remote IP is the same as the
4178 local IP, and the "valid-client-cert option" if a client certificate has
4181 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4182 authenticators. This can be used for authorization after authentication
4183 succeeds. (In the case of plaintext, it servers for both authentication
4186 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4187 if any retry times were supplied.
4189 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4190 connection, where there is no client IP address to check. In this
4191 situation, the verify now always succeeds.
4193 PH/25 Applied John Jetmore's -Mset patch.
4195 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4197 PH/27 In a string expansion for a processed (not raw) header when multiple
4198 headers of the same name were present, leading whitespace was being
4199 removed from all of them, but trailing whitespace was being removed only
4200 from the last one. Now trailing whitespace is removed from each header
4201 before concatenation. Completely empty headers in a concatenation (as
4202 before) are ignored.
4204 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4205 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4207 PH/29 [Removed. This was a change that I later backed out, and forgot to
4208 correct the ChangeLog entry (that I had efficiently created) before
4209 committing the later change.]
4211 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4212 address errors (4xx response to RCPT) over the same connection as other
4213 messages routed to the same hosts. Such deliveries are always "forced",
4214 so retry times are not inspected. This resulted in far too many retries
4215 for the affected addresses. The effect occurred only when there were more
4216 hosts than the hosts_max_try setting in the smtp transport when it had
4217 the 4xx errors. Those hosts that it had tried were not added to the list
4218 of hosts for which the message was waiting, so if all were tried, there
4219 was no problem. Two fixes have been applied:
4221 (i) If there are any address or message errors in an SMTP delivery, none
4222 of the hosts (tried or untried) are now added to the list of hosts
4223 for which the message is waiting, so the message should not be a
4224 candidate for sending over the same connection that was used for a
4225 successful delivery of some other message. This seems entirely
4226 reasonable: after all the message is NOT "waiting for some host".
4227 This is so "obvious" that I'm not sure why it wasn't done
4228 previously. Hope I haven't missed anything, but it can't do any
4229 harm, as the worst effect is to miss an optimization.
4231 (ii) If, despite (i), such a delivery is accidentally attempted, the
4232 routing retry time is respected, so at least it doesn't keep
4233 hammering the server.
4235 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4236 in ${readsocket because some servers need this prod.
4238 PH/32 Added some extra debug output when updating a wait-xxx database.
4240 PH/33 The hint "could be header name not terminated by colon", which has been
4241 given for certain expansion errors for a long time, was not being given
4242 for the ${if def:h_colon_omitted{... case.
4244 PH/34 The spec says: "With one important exception, whenever a domain list is
4245 being scanned, $domain contains the subject domain." There was at least
4246 one case where this was not true.
4248 PH/35 The error "getsockname() failed: connection reset by peer" was being
4249 written to the panic log as well as the main log, but it isn't really
4250 panic-worthy as it just means the connection died rather early on. I have
4251 removed the panic log writing for the ECONNRESET error when getsockname()
4254 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4255 runs only) independently of the message's sender address. This meant
4256 that, if the 4xx error was in fact related to the sender, a different
4257 message to the same recipient with a different sender could confuse
4258 things. In particular, this can happen when sending to a greylisting
4259 server, but other circumstances could also provoke similar problems.
4260 I have changed the default so that the retry time for these errors is now
4261 based a combination of the sender and recipient addresses. This change
4262 can be overridden by setting address_retry_include_sender=false in the
4265 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4266 remote server are returned as part of bounce messages. This was not
4267 happening for LMTP over a pipe (the lmtp transport), but now it is the
4268 same for both kinds of LMTP.
4270 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4271 in header lines that were in fact CNAMEs. This is no longer the case.
4273 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4274 and queue runs started by the daemon processed all messages. This has
4275 been fixed so that -R and -S can now usefully be given with -q<time>.
4277 PH/40 Import PCRE release 6.7 (fixes some bugs).
4279 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4281 PH/42 Give an error if -q is specified more than once.
4283 PH/43 Renamed the variables $interface_address and $interface_port as
4284 $received_ip_address and $received_port, to make it clear that these
4285 values apply to message reception, and not to the outgoing interface when
4286 a message is delivered. (The old names remain recognized, of course.)
4288 PH/44 There was no timeout on the connect() call when using a Unix domain
4289 socket in the ${readsocket expansion. There now is.
4291 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4292 be meaningful with "accept".
4294 SC/01 Eximstats V1.43
4295 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4297 SC/02 Eximstats V1.44
4298 Use a glob alias rather than an array ref in the generated
4299 parser. This improves both readability and performance.
4301 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4302 Collect SpamAssassin and rejection statistics.
4303 Don't display local sender or destination tables unless
4304 there is data to show.
4305 Added average volumes into the top table text output.
4307 SC/04 Eximstats V1.46
4308 Collect data on the number of addresses (recipients)
4309 as well as the number of messages.
4311 SC/05 Eximstats V1.47
4312 Added 'Message too big' to the list of mail rejection
4313 reasons (thanks to Marco Gaiarin).
4315 SC/06 Eximstats V1.48
4316 Mainlog lines which have GMT offsets and are too short to
4317 have a flag are now skipped.
4319 SC/07 Eximstats V1.49 (Alain Williams)
4320 Added the -emptyok flag.
4322 SC/08 Eximstats V1.50
4323 Fixes for obtaining the IP address from reject messages.
4325 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4326 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4327 whitespace changes from 4.64-PH/27
4329 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4332 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4333 are found, allow negative numbers in numeric criteria)
4335 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4337 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4338 to match changes made in 4.64-PH/43
4340 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4342 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4343 "domain" is actually an IP address, and force a failure. This locks out
4344 those revolvers/nameservers that support "A-for-A" lookups, in
4345 contravention of the specifications.
4347 PH/48 When a host name was looked up from an IP address, and the subsequent
4348 forward lookup of the name timed out, the host name was left in
4349 $sender_host_name, contrary to the specification.
4351 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4352 restricted to single-key lookups, Exim was not diagnosing an error if
4353 * or *@ was used with a query-style lookup.
4355 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4357 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4358 long ago, but noone interested enough thought of it. Let's just say that
4359 the "1.1" means that there are some new functions that weren't there at
4360 some point in the past.
4362 PH/51 Error processing for expansion failure of helo_data from an smtp
4363 transport during callout processing was broken.
4365 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4366 tested/used via the -bh/-bhc/-bs options.
4368 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4369 bug, fixed in subsequent PCRE releases).
4371 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4372 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4374 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4380 SC/01 Use a glob alias rather than an array ref in eximstats generated
4381 parser. This improves both readability and performance.
4383 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4384 Don't display local sender or destination tables in eximstats unless
4385 there is data to show.
4386 Added average volumes into the eximstats top table text output.
4388 SC/03 Collect data on the number of addresses (recipients) as well
4389 as the number of messages in eximstats.
4391 TF/01 Correct an error in the documentation for the redirect router. Exim
4392 does (usually) call initgroups() when daemonizing.
4394 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4395 with consistent privilege compared to when running as a daemon.
4397 TF/03 Note in the spec that $authenticated_id is not set for local
4398 submissions from trusted users.
4400 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4401 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4403 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4404 by adding some example configuration directives to the default
4405 configuration file. A little bit of work is required to uncomment the
4406 directives and define how usernames and passwords are checked, but
4407 there is now a framework to start from.
4409 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4410 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4411 without this. I don't know how relevant this is to other LDAP libraries.
4413 PH/02 Add the verb name to the "unknown ACL verb" error.
4415 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4417 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4419 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4420 directory not expanded when it should be if an expanded home directory
4421 was set for the address (which is overridden by the transport).
4423 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4426 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4427 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4428 because it is too late at that time, and has no effect.
4430 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4431 security issue with \' (bugzilla #107). I could not use the
4432 PQescapeStringConn() function, because it needs a PGconn value as one of
4435 PH/08 When testing addresses using -bt, indicate those final addresses that
4436 are duplicates that would not cause an additional delivery. At least one
4437 person was confused, thinking that -bt output corresponded to deliveries.
4438 (Suppressing duplicates isn't a good idea as you lose the information
4439 about possibly different redirections that led to the duplicates.)
4441 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4442 systems where poll() doesn't work, in particular OS X.
4444 PH/10 Added more information to debugging output for retry time not reached.
4446 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4447 operations in malware.c.
4449 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4452 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4453 syslog_duplication set false, log lines that would normally be written
4454 both the the main log and to the reject log were not written to syslog at
4457 PH/14 In the default configuration, change the use of "message" in ACL warn
4458 statements to "add_header".
4460 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4461 not followed by a command (e.g. "seen endif").
4463 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4464 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4467 PH/17 Added extra conditions to the default value of delay_warning_condition
4471 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4472 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4473 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4476 The Auto-Submitted: and various List- headers are standardised, whereas I
4477 don't think Precedence: ever was.
4479 PH/18 Refactored debugging code in route_finduser() to show more information,
4480 in particular, the error code if getpwnam() issues one.
4482 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4483 This is apparently needed in addition to the PH/07 change above to avoid
4484 any possible encoding problems.
4486 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4487 but not after initializing Perl.
4489 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4490 output them only if debugging. By default they are written stderr,
4491 apparently, which is not desirable.
4493 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4496 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4499 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4501 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4502 authenticated or an ident call has been made. Suppress the default
4503 values for $authenticated_id and $authenticated_sender (but permit -oMai
4504 and -oMas) when testing with -bh.
4506 PH/24 Re-jigged the order of the tests in the default configuration so that the
4507 tests for valid domains and recipients precede the DNS black list and CSA
4508 tests, on the grounds that those ones are more expensive.
4510 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4511 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4512 This bug exists in every version of Exim that I still have, right back to
4515 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4516 However, an attempt to turn on case-sensitivity in a regex key by
4517 including (?-i) didn't work because the subject string was already
4518 lowercased, and the effects were non-intuitive. It turns out that a
4519 one-line patch can be used to allow (?-i) to work as expected.
4525 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4526 other effects) broke the use of negated acl sub-conditions.
4528 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4531 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4532 "Deny" causes Exim to reject the incoming connection with a 554 error.
4533 Unfortunately, if there is a major crisis, such as a disk failure,
4534 tcp-wrappers gives "deny", whereas what one would like would be some
4535 kind of temporary error. A kludge has been added to help with this.
4536 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4537 554 error is used if errno is still zero or contains ENOENT (which occurs
4538 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4541 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4543 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4544 errors. Otherwise a message that provokes a temporary error (when other
4545 messages do not) can cause a whole host to time out.
4547 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4548 addresses were routed directly to files or pipes from a redirect router.
4549 File deliveries just didn't batch; pipe deliveries might have suffered
4552 PH/06 A failure to get a lock for a hints database would erroneously always say
4553 "Failed to get write lock", even when it was really a read lock.
4555 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4556 of 0600. This has been changed to use the value of the lockfile_mode
4557 option (which defaults to 0600).
4559 PH/08 Applied small patch from the Sieve maintainer.
4561 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4562 folder from quota calculations, a direct delivery into this folder messed
4563 up the contents of the maildirsize file. This was because the regex was
4564 used only to exclude .Trash (or whatever) when the size of the mailbox
4565 was calculated. There was no check that a delivery was happening into an
4566 excluded directory. This bug has been fixed by ignoring all quota
4567 processing for deliveries into excluded directories.
4569 PH/10 Added the maildirfolder_create_regex option to appendfile.
4575 PH/01 The code for finding all the local interface addresses on a FreeBSD
4576 system running IPv6 was broken. This may well have applied to all BSD
4577 systems, as well as to others that have similar system calls. The broken
4578 code found IPv4 interfaces correctly, but gave incorrect values for the
4579 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4580 that it would not match correctly against @[] and not recognize the IPv6
4583 PH/02 The ipliteral router was not recognizing addresses of the form user@
4584 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4586 PH/03 Added disable_ipv6.
4588 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4589 decoded form, because it is most often used to construct To: headers
4590 lines in autoreplies, and the decoded form may well be syntactically
4591 invalid. However, $reply_address has leading white space removed, and all
4592 newlines turned into spaces so that the autoreply transport does not
4595 PH/05 If group was specified without a user on a router, and no group or user
4596 was specified on a transport, the group from the router was ignored.
4598 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4599 for visible compile-time settings that can be used to change these
4600 numbers, for those that want even more. Backwards compatibility with old
4601 spool files has been maintained. However, going back to a previous Exim
4602 release will lost any variables that are in spool files.
4604 PH/07 Two small changes when running in the test harness: increase delay when
4605 passing a TCP/IP connection to a new process, in case the original
4606 process has to generate a bounce, and remove special handling of
4607 127.0.0.2 (sic), which is no longer necessary.
4609 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4610 be the same on different OS.
4612 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4615 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4616 whether --show-vars was specified or not
4618 JJ/02 exipick: Added support for new ACL variable spool format introduced
4621 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4622 syntactically invalid From: or Reply-to: line, and a filter used this to
4623 generate an autoreply, and therefore failed to obtain an address for the
4624 autoreply, Exim could try to deliver to a non-existent relative file
4625 name, causing unrelated and misleading errors. What now happens is that
4626 it logs this as a hard delivery error, but does not attempt to create a
4629 PH/11 The exinext utility has a -C option for testing purposes, but although
4630 the given file was scanned by exinext itself; it wasn't being passed on
4631 when Exim was called.
4633 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4634 an end-of-file indication when reading a command response.
4636 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4637 compiled. In many other places in Exim, IPv6 addresses are always
4638 recognized, so I have changed this. It also means that IPv4 domain
4639 literals of the form [IPV4:n.n.n.n] are now always recognized.
4641 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4642 used if the router is not running as root, for example, when verifying at
4643 ACL time, or when using -bh. The debugging output from this situation was
4644 non-existent - all you got was a failure to exec. I have made two
4647 (a) Failures to set uid/gid, the current directory, or a process leader
4648 in a subprocess such as that created by queryprogram now generate
4649 suitable debugging output when -d is set.
4651 (b) The queryprogram router detects when it is not running as root,
4652 outputs suitable debugging information if -d is set, and then runs
4653 the subprocess without attempting to change uid/gid.
4655 PH/15 Minor change to Makefile for building test_host (undocumented testing
4658 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4659 additional section of a DNS packet that returns MX or SRV records.
4660 Instead, it always explicitly searches for A/AAAA records. This avoids
4661 major problems that occur when a DNS server includes only records of one
4662 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4663 fixed another bug: if SRV records were looked up and the corresponding
4664 address records were *not* found in the additional section, the port
4665 values from the SRV records were lost.
4667 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4668 using the correct key (the original address) when searching the retry
4669 rules in order to find which one to use for generating the retry hint.
4671 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4672 adding the default one. Similarly, if it contains a Reply-To: header, the
4673 errors_reply_to option, if set, is not used.
4675 PH/19 When calculating a retry time, Exim used to measure the "time since
4676 failure" by looking at the "first failed" field in the retry record. Now
4677 it does not use this if it is later than than the arrival time of the
4678 message. Instead it uses the arrival time. This makes for better
4679 behaviour in cases where some deliveries succeed, thus re-setting the
4680 "first failed" field. An example is a quota failure for a huge message
4681 when small messages continue to be delivered. Without this change, the
4682 "time since failure" will always be short, possible causing more frequent
4683 delivery attempts for the huge message than are intended.
4684 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4686 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4687 $1, $2, $3) because the numerical variables can be reset during some
4688 expansion items (e.g. "match"), thereby losing the authentication data.
4690 PH/21 Make -bV show the size of off_t variables so that the test suite can
4691 decide whether to run tests for quotas > 2G.
4693 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4694 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4695 filecount value is greater than 2G or if a quota value is greater than 2G
4696 on a system where the size of off_t is not greater than 4, a panic error
4699 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4700 never match. The debug and -bh output now contains an explicit error
4701 message indicating a malformed IPv4 address or mask.
4703 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4704 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4705 PH/23 above applies.
4707 PH/25 Do not write to syslog when running in the test harness. The only
4708 occasion when this arises is a failure to open the main or panic logs
4709 (for which there is an explicit test).
4711 PH/26 Added the /no_tell option to "control=freeze".
4713 PH/27 If a host name lookup failed very early in a connection, for example, if
4714 the IP address matched host_lookup and the reverse lookup yielded a name
4715 that did not have a forward lookup, an error message of the form "no IP
4716 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4717 could be logged. Now it outputs the IP address instead of "NULL".
4719 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4720 allows the sender and the authenticated sender to be set when
4721 submitting a message from within Exim. Since child_open_exim() is
4722 documented for local_scan(), the new function should be too.
4724 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4725 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4726 results in an empty string is now treated as unset.
4728 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4730 PH/31 Added sender_verify_fail logging option.
4732 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4733 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4734 tidied the source and removed it altogether.
4736 PH/33 When a queue run was abandoned because the load average was too high, a
4737 log line was always written; now it is written only if the queue_run log
4738 selector is set. In addition, the log line for abandonment now contains
4739 information about the queue run such as the pid. This is always present
4740 in "start" and "stop" lines but was omitted from the "abandon" line.
4742 PH/34 Omit spaces between a header name and the colon in the error message that
4743 is given when verify = headers_syntax fails (if there are lots of them,
4744 the message gets confusing).
4746 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4747 names, as there are now some PTR records that contain slashes. This check
4748 is only to protect against broken name servers that fall over on strange
4749 characters, so the fact that it applies to all lookups doesn't matter.
4751 PH/36 Now that the new test suite is complete, we can remove some of the
4752 special code in Exim that was needed for the old test suite. For example,
4753 sorting DNS records because real resolvers return them in an arbitrary
4754 order. The new test suite's fake resolver always returns records in the
4757 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4758 bounces) except when queue_only is set, to avoid logging races between
4759 the different processes.
4761 PH/38 Panic-die if .include specifies a non-absolute path.
4763 PH/39 A tweak to the "H" retry rule from its user.
4765 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4766 a label. They prevented compilation on older perls.
4768 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4769 a warning to be raised on newish perls.
4771 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4772 on queue. Changes to match documented behaviour of showing count of
4773 messages matching specified criteria.
4775 PH/40 Changed the default ident timeout from 30s to 5s.
4777 PH/41 Added support for the use of login_cap features, on those BSD systems
4778 that have them, for controlling the resources used by pipe deliveries.
4780 PH/42 The content-scanning code uses fopen() to create files in which to put
4781 message data. Previously it was not paying any attention to the mode of
4782 the files. Exim runs with umask(0) because the rest of the code creates
4783 files with open(), and sets the required mode explicitly. Thus, these
4784 files were ending up world-writeable. This was not a big issue, because,
4785 being within the spool directory, they were not world-accessible. I have
4786 created a function called modefopen, which takes an additional mode
4787 argument. It sets umask(777), creates the file, chmods it to the required
4788 mode, then resets the umask. All the relevant calls to fopen() in the
4789 content scanning code have been changed to use this function.
4791 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4792 to 24 hours. This avoids potential overflow problems when processing G
4793 and H retry rules. I suspect nobody ever tinkers with this value.
4795 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4797 PH/45 When the plaintext authenticator is running as a client, the server's
4798 challenges are checked to ensure they are valid base64 strings. By
4799 default, the authentication attempt is cancelled if an invalid string is
4800 received. Setting client_ignore_invalid_base64 true ignores these errors.
4801 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4802 they are received. Thus, the responses can be made to depend on the
4803 challenges. If an invalid string is ignored, an empty string is placed in
4806 PH/46 Messages that are created by the autoreply transport now contains a
4807 References: header, in accordance with RFCs 2822 and 3834.
4809 PH/47 Added authenticated_sender_force to the smtp transport.
4811 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4813 PH/49 Installed latest patch from the Sieve maintainer.
4815 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4816 was also set, the appendfile transport was unnecessarily scanning a
4817 directory of message files (e.g. for maildir delivery) to find the count
4818 of files (along with the size), even though it did not need this
4819 information. It now does the scan only if it needs to find either the
4820 size of the count of files.
4822 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4824 PH/52 Two bugs concerned with error handling when the smtp transport is
4827 (i) Exim was not creating retry information for temporary errors given
4828 for individual recipients after the DATA command when the smtp transport
4829 was used in LMTP mode. This meant that they could be retried too
4830 frequently, and not timed out correctly.
4832 (ii) Exim was setting the flag that allows error details to be returned
4833 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4834 recipients that were returned after the DATA command.
4836 PH/53 This is related to PH/52, but is more general: for any failing address,
4837 when detailed error information was permitted to be returned to the
4838 sender, but the error was temporary, then after the final timeout, only
4839 "retry timeout exceeded" was returned. Now it returns the full error as
4840 well as "retry timeout exceeded".
4842 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4843 do this, and (what is worse) MTAs that accept it.
4845 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4846 will now be deprecated.
4848 PH/56 New os.c-cygwin from the Cygwin maintainer.
4850 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4851 formats (previously only available in exim formats via -bpr, -bpru,
4852 and -bpra. Now also available in native and exiqgrep formats)
4854 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4855 with very large, slow to parse queues
4857 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4859 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4861 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4862 responses to authentication challenges, though it was showing the
4863 challenges; (ii) I've removed the CR characters from the debug output for
4866 PH/58 Allow for the insertion of a newline as well as a space when a string
4867 is turned into more than one encoded-word during RFC 2047 encoding. The
4868 Sieve code now uses this.
4870 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4871 data_4xx, lost_connection, tls_required.
4873 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4874 message was being sent as an SMTP response.
4876 PH/61 Add -l and -k options to exicyclog.
4878 PH/62 When verifying, if an address was redirected to one new address, so that
4879 verification continued, and the new address failed or deferred after
4880 having set something in $address_data, the value of $address_data was not
4881 passed back to the ACL. This was different to the case when no
4882 redirection occurred. The value is now passed back in both cases.
4884 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4885 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4886 use login_cap.h, so on its own it isn't the distinguishing feature. The
4887 new name refers directly to the setclassresources() function.
4889 PH/65 Added configuration files for NetBSD3.
4891 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4893 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4894 is preferred over IPv4.
4896 PH/68 The bounce_return_message and bounce_return_body options were not being
4897 honoured for bounces generated during the reception of non-SMTP messages.
4898 In particular, this applied to messages rejected by the ACL. This bug has
4899 been fixed. However, if bounce_return_message is true and bounce_return_
4900 body is false, the headers that are returned for a non-SMTP message
4901 include only those that have been read before the error was detected.
4902 (In the case of an ACL rejection, they have all been read.)
4904 PH/69 The HTML version of the specification is now built in a directory called
4905 spec_html instead of spec.html, because the latter looks like a path with
4906 a MIME-type, and this confuses some software.
4908 PH/70 Catch two compiler warnings in sieve.c.
4910 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4911 function verify_get_ident() calls ip_connect() to connect a socket, but
4912 if the "connect()" function timed out, ip_connect() used to close the
4913 socket. However, verify_get_ident() also closes the socket later, and in
4914 between Exim writes to the log, which may get opened at this point. When
4915 the socket was closed in ip_connect(), the log could get the same file
4916 descriptor number as the socket. This naturally causes chaos. The fix is
4917 not to close the socket in ip_connect(); the socket should be closed by
4918 the function that creates it. There was only one place in the code where
4919 this was missing, in the iplookup router, which I don't think anybody now
4920 uses, but I've fixed it anyway.
4922 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4923 well as to direct DNS lookups. Otherwise the handling of names in host
4924 lists is inconsistent and therefore confusing.
4930 PH/01 Two changes to the default runtime configuration:
4932 (1) Move the checks for relay_from_hosts and authenticated clients from
4933 after to before the (commented out) DNS black list checks.
4935 (2) Add control=submission to the relay_from_hosts and authenticated
4936 clients checks, on the grounds that messages accepted by these
4937 statements are most likely to be submissions.
4939 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4941 (1) Generate an error if the third argument for the ${prvs expansion is
4944 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4947 (3) Reset the variables that are obtained from the first argument of
4948 ${prvscheck and used in the second argument before leaving the code,
4949 because their memory is reclaimed, so using them afterwards may do
4952 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4953 one (it's much easier than Tom thought :-).
4955 (5) Because of (4), we can now allow for the use of $prvscheck_result
4956 inside the third argument.
4958 PH/03 For some reason, the default setting of PATH when running a command from
4959 a pipe transport was just "/usr/bin". I have changed it to
4962 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4963 anything to be listed in the output from -bV.
4965 PH/05 When a filter generated an autoreply, the entire To: header line was
4966 quoted in the delivery log line, like this:
4968 => >A.N.Other <ano@some.domain> <original@ddress> ...
4970 This has been changed so that it extracts the operative address. There
4971 may be more than one such address. If so, they are comma-separated, like
4974 => >ano@some.domain,ona@other.domain <original@ddress> ...
4976 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4977 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4978 not being looked up in the rDNS to get a host name, Exim was showing the
4979 IP address twice in Received: lines, even though the IP addresses were
4980 identical. For example:
4982 Received: from [1.2.3.4] (helo=[1.2.3.4])
4984 However, if the real host name was known, it was omitting the HELO data
4985 if it matched the actual IP address. This has been tidied up so that it
4986 doesn't show the same IP address twice.
4988 PH/07 When both +timestamp and +memory debugging was on, the value given by
4989 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4990 called by the debug printing, thereby overwriting the timestamp buffer.
4991 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4993 PH/08 When the original message was included in an autoreply transport, it
4994 always said "this is a copy of the message, including all the headers",
4995 even if body_only or headers_only was set. It now gives an appropriate
4998 PH/09 Applied a patch from the Sieve maintainer which:
5000 o fixes some comments
5001 o adds the (disabled) notify extension core
5002 o adds some debug output for the result of if/elsif tests
5003 o points to the current vacation draft in the documentation
5004 and documents the missing references header update
5008 o fixes a bug in processing the envelope test (when testing
5009 multiple envelope elements, the last element determined the
5012 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5013 Electronic Mail") by including:
5015 Auto-submitted: auto-generated
5017 in the messages that it generates (bounce messages and others, such as
5018 warnings). In the case of bounce messages for non-SMTP messages, there was
5019 also a typo: it was using "Auto_submitted" (underscore instead of
5020 hyphen). Since every message generated by Exim is necessarily in response
5021 to another message, thes have all been changed to:
5023 Auto-Submitted: auto-replied
5025 in accordance with these statements in the RFC:
5027 The auto-replied keyword:
5029 - SHOULD be used on messages sent in direct response to another
5030 message by an automatic process,
5032 - MUST NOT be used on manually-generated messages,
5034 - MAY be used on Delivery Status Notifications (DSNs) and Message
5035 Disposition Notifications (MDNs),
5037 - MUST NOT be used on messages generated by automatic or periodic
5038 processes, except for messages which are automatic responses to
5041 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5042 to the default Received: header definition.
5044 PH/12 Added log selector acl_warn_skipped (default on).
5046 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5047 variables because (a) they are in the wrong storage pool and (b) even if
5048 they were copied, it wouldn't work properly because of the caching.
5050 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5051 checking when decoding. Apparently there are clients that generate
5052 overlong encoded strings. Why am I not surprised?
5054 PH/15 If the first argument of "${if match_address" was not empty, but did not
5055 contain an "@" character, Exim crashed. Now it writes a panic log message
5056 and treats the condition as false.
5058 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5060 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5061 extension "envelope-auth". The code is finished and in agreement with
5062 other implementations, but there is no documentation so far and in fact,
5063 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5064 not changing the active code.
5066 Print executed "if" and "elsif" statements when debugging is used. This
5067 helps a great deal to understand what a filter does.
5069 Document more things not specified clearly in RFC3028. I had all this
5070 sorted out, when out of a sudden new issues came to my mind. Oops."
5072 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5075 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5076 canonical form (as documented). However, after a host name lookup from
5077 the IP address, check_host() was doing a simple string comparison with
5078 addresses acquired from the DNS when checking that the found name did
5079 have the original IP as one of its addresses. Since any found IPv6
5080 addresses are likely to be in abbreviated form, the comparison could
5081 fail. Luckily, there already exists a function for doing the comparison
5082 by converting both addresses to binary, so now that is used instead of
5083 the text comparison.
5085 PH/20 There was another similar case to PH/19, when a complete host name was
5086 given in a host list; looking up its IP address could give an abbreviated
5087 form, whereas the current host's name might or might not be abbreviated.
5088 The same fix has been applied.
5094 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5095 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5098 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5099 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5101 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5103 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5104 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5105 Also, for "Auto-Submitted", treat anything other than "no" as
5106 non-personal, in accordance with RFC 3834. (Previously it treated
5107 anything starting "auto-" as non-personal.)
5109 TF/01 The control=submission/name=... option had a problem with syntax
5110 errors if the name included a slash character. The /name= option
5111 now slurps the rest of the string, so it can include any characters
5112 but it must come last in the list of options (after /sender_retain
5115 PH/05 Some modifications to the interface to the fake nameserver for the new
5123 TK/01 Added the "success_on_redirect" address verification option. See
5124 NewStuff for rationale and an example.
5126 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5128 PH/02 Patch to exigrep to allow it to work on syslog lines.
5130 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5131 fread() to skip over the body file's header line, because in Cygwin the
5132 header line is locked and is inaccessible.
5134 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5135 co-exist for some time) to make it clear that it is the Exim ID that is
5136 referenced, not the Message-ID: header line.
5138 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5139 string_format() function, because snprintf() does not exist on all
5142 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5143 ${stat: expansion item.
5145 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5146 protocol synchronization error", to keep the pedants happy.
5148 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5149 well as for IRIX systems, when gcc is being used. See the host.c source
5152 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5154 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5157 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5158 transport and to the smtp transport in LMTP mode.
5160 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5162 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5163 run a filter in a subprocess. This could lead to confusion in subsequent
5164 lookups in the parent process. There should also be a search_tidyup() at
5165 the end of the subprocess.
5167 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5168 only if the host matched helo_try_verify_hosts, which caused the
5169 verification to occur when the EHLO/HELO command was issued. The ACL just
5170 tested the remembered result. Now, if a previous verification attempt has
5171 not happened, "verify = helo" does it there and then.
5173 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5175 TK/03 Fix log output including CR from clamd.
5177 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5178 exist provoked a memory error which could cause a segfault.
5180 PH/15 Installed PCRE 6.2
5182 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5184 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5185 of the problem. Specifically, suggested +O2 rather than +O1 for the
5188 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5190 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5191 record, the port was not correctly logged when the outgoing_port log
5192 selector was set (it logged the transort's default port).
5194 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5195 fallback_hosts, and "hosts" in the smtp transport.
5197 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5198 host errors such as "Connection refused".
5200 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5201 authentication with radiusclient 0.4.9:
5203 - Error returned from rc_read_config was caught wrongly
5204 - Username/password not passed on to radius server due to wrong length.
5206 The presumption is that some radiusclient API changes for 4.51/PH/17
5207 were not taken care of correctly. The code is still untested by me (my
5208 Linux distribution still has 0.3.2 of radiusclient), but it was
5209 contributed by a Radius user.
5211 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5212 expanding the "port" option of the smtp transport.
5214 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5215 while reading a MIME header. Thanks to Tom Hughes for a patch.
5217 PH/24 Include config.h inside local_scan.h so that configuration settings are
5220 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5221 that in an ACL for RCPT (for example), you can examine exactly what was
5224 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5225 commands, but it was not correctly comparing the address with the actual
5226 client host address. Thus, it would show the EHLO address in Received:
5227 header lines when this was not necessary.
5229 PH/27 Added the % operator to ${eval:}.
5231 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5232 it should be ignoring failures (because with -C, for example, it has lost
5233 privilege). It wasn't ignoring creation failures other than "already
5236 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5239 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5240 because an input line was too long, either on its own, or by virtue of
5241 too many continuations, the temporary file was not being removed, and the
5242 return code was incorrect.
5244 PH/31 Missing "BOOL" in function definition in filtertest.c.
5246 PH/32 Applied Sieve patches from the maintainer.
5248 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5250 PH/33 Added "verify = not_blind".
5252 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5253 Local/Makefile (with some defaults set). These are used in built scripts
5254 such as exicyclog, but they have never been used in the exim_install
5255 script (though there are many overriding facilities there). I have
5256 arranged that the exim_install script now takes note of these two
5259 PH/35 Installed configuration files for Dragonfly.
5261 PH/36 When a locally submitted message by a trusted user did not contain a
5262 From: header, and the sender address was obtained from -f or from an SMTP
5263 MAIL command, and the trusted user did not use -F to supply a sender
5264 name, $originator_name was incorrectly used when constructing a From:
5265 header. Furthermore, $originator_name was used for submission mode
5266 messages from external hosts without From: headers in a similar way,
5267 which is clearly wrong.
5269 PH/37 Added control=suppress_local_fixups.
5271 PH/38 When log_selector = +received_sender was set, and the addition of the
5272 sender made the log line's construction buffer exactly full, or one byte
5273 less than full, an overflow happened when the terminating "\n" was
5276 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5277 when the result of a list match is failure because a DNS lookup failed.
5279 PH/40 RM_COMMAND is now used in the building process.
5281 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5282 the "build-* directories that it finds.
5284 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5285 address in a domain literal was a prefix of an interface address.
5287 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5288 when verifying a sender address, unless rewrite_headers is false.
5290 PH/44 Wrote a long comment about why errors_to addresses are verified as
5291 recipients, not senders.
5293 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5294 the ratelimit ACL was added.
5296 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5298 PH/46 Added extra information about PostgreSQL errors to the error string.
5300 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5301 avoiding the need to install special zones in a real server. This is
5302 backwards compatible; if it can't find the fake resolver, it drops back.
5303 Thus, both old and new test suites can be run.
5305 TF/02 Added util/ratelimit.pl
5307 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5308 clock is set back in time.
5310 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5311 Candler <B.Candler@pobox.com>.
5313 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5314 OK for addresses that are the result of redirections.
5316 PH/48 A number of further additions for the benefit of the new test suite,
5317 including a fake gethostbyname() that interfaces to the fake DNS resolver
5320 TF/06 The fix for widen_domains has also been applied to qualify_single and
5321 search_parents which are the other dnslookup options that can cause
5324 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5327 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5329 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5330 service" if it encountered an MX record with an empty target hostname.
5331 The message is now "an MX or SRV record indicated no SMTP service".
5333 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5334 if the DNS of the sending site is misconfigured. This is quite a
5335 common situation. This change restores the behaviour of treating a
5336 helo verification defer as a failure.
5338 PH/51 If self=fail was set on a router, the bounce message did not include the
5339 actual error message.
5345 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5347 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5348 command itself ended in error, the underlying message about the transport
5349 filter timeout was being overwritten with the pipe command error. Now the
5350 underlying error message should be appended to the second error message.
5352 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5354 PH/02 Reduce the amount of output that "make" produces by default. Full output
5355 can still be requested.
5357 PH/03 The warning log line about a condition test deferring for a "warn" verb
5358 was being output only once per connection, rather than after each
5359 occurrence (because it was using the same function as for successful
5360 "warn" verbs). This seems wrong, so I have changed it.
5362 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5363 it should not have, which might have caused a crash in the right
5364 circumstances, but probably never did.
5366 PH/04 Installed a modified version of Tony Finch's patch to make submission
5367 mode fix the return path as well as the Sender: header line, and to
5368 add a /name= option so that you can make the user's friendly name appear
5371 TF/03 Added the control = fakedefer ACL modifier.
5373 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5374 Mark Lowes for thorough testing.
5376 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5378 TK/03 Merged latest SRS patch from Miles Wilton.
5380 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5381 with the definition in sysexits.h (which is #included earlier).
5382 Fortunately, Exim does not actually use EX_OK. The code used to try to
5383 preserve the sysexits.h value, by assuming that macro definitions were
5384 scanned for macro replacements. I have been disabused of this notion,
5385 so now the code just undefines EX_OK before #including unistd.h.
5387 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5388 in the smtp transport. When a block could not be written in a single
5389 write() function, the timeout was being re-applied to each part-write.
5390 This seems wrong - if the receiver was accepting one byte at a time it
5391 would take for ever. The timeout is now adjusted when this happens. It
5392 doesn't have to be particularly precise.
5394 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5395 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5397 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5398 without a domain if the check to <postmaster@domain> fails.
5400 SC/01 Eximstats: added -xls and the ability to specify output files
5401 (patch written by Frank Heydlauf).
5403 SC/02 Eximstats: use FileHandles for outputting results.
5405 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5407 SC/04 Eximstats: fixed display of large numbers with -nvr option
5409 SC/05 Eximstats: fixed merging of reports with empty tables.
5411 SC/06 Eximstats: added the -include_original_destination flag
5413 SC/07 Eximstats: removed tabs and trailing whitespace.
5415 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5417 TK/06 MBOX spool code: Add real "From " MBOX separator line
5418 so the .eml file is really in mbox format (even though
5419 most programs do not really care). Patch from Alex Miller.
5421 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5422 The latter is generated from $received_to and is only set if the
5423 message has one envelope recipient. SA can use these headers,
5424 obviously out-of-the-box. Patch from Alex Miller.
5426 PH/08 The ${def test on a variable was returning false if the variable's
5427 value was "0", contrary to what the specification has always said!
5428 The result should be true unless the variable is empty.
5430 PH/09 The syntax error of a character other than { following "${if
5431 def:variable_name" (after optional whitespace) was not being diagnosed.
5432 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5433 accidental colon was present, for example, could give incorrect results.
5435 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5436 result is used (not including appendfile, where other changes are about
5439 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5440 This involved changing a lot of size variables from int to off_t. It
5441 should work with maildirs and everything.
5443 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5444 spamd dying while we are connected to it.
5446 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5449 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5450 function for BDB 4.3.
5452 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5454 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5455 the sources for PCRE have been re-arranged and more files are now
5458 PH/15 The code I had for printing potentially long long variables in PH/11
5459 above was not the best (it lost precision). The length of off_t variables
5460 is now inspected at build time, and an appropriate printing format (%ld
5461 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5462 to be "long long int" or "long int". This is needed for the internal
5463 formatting function string_vformat().
5465 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5466 the configuration file to be ":syslog", then the script "guesses" where
5467 the logs files are, rather than using the compiled in default. In our
5468 case the guess is not the same as the compiled default, so the script
5469 suddenly stopped working when I started to use syslog. The patch checks
5470 to see if log_file_path is "". If so, it attempts to read it from exim
5471 with no configuration file to get the compiled in version, before it
5472 falls back to the previous guessing code."
5474 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5475 implementing BATV in an Exim configuration. See NewStuff for the gory
5478 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5479 Makefile that are specific to HP-UX.
5481 PH/18 If the "use_postmaster" option was set for a recipient callout together
5482 with the "random" option, the postmaster address was used as the MAIL
5483 FROM address for the random test, but not for the subsequent recipient
5484 test. It is now used for both.
5486 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5487 patch removes a few documentation additions to RFC 3028, because the
5488 latest draft now contains them. It adds the new en;ascii-case comparator
5489 and a new error check for 8bit text in MIME parts. Comparator and
5490 require names are now matched exactly. I enabled the subaddress
5491 extension, but it is not well tested yet (read: it works for me)."
5493 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5494 rework some of the code of TK/09 above to avoid the hardwired use of
5495 "%lld" and "long long". Replaced the call to snprintf() with a call to
5498 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5499 records point to non-existent hosts", "retry timeout exceeded", and
5500 "retry time not reached for any host after a long failure period".
5502 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5503 experimental DomainKeys support:
5505 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5506 (2) On an error such as an illegally used "control", the wrong name for
5507 the control was given.
5509 These problems did NOT occur unless DomainKeys support was compiled.
5511 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5513 PH/24 Added ${if match_ip condition.
5515 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5516 databases so that it will be absolutely obvious if a crash occurs in the
5517 DB library. This is a regular occurrence (often caused by mis-matched
5520 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5521 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5522 picked up on a user's system that detects such things. There doesn't seem
5523 to be a gcc warning option for this - only an attribute that has to be
5524 put on the function's prototype. It seems that in Fedora Core 4 they have
5525 set this on a number of new functions. No doubt there will be more in due
5528 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5529 specify a transport. However, if an address that was verified by such a
5530 router was the subject of a callout, Exim crashed because it tried to
5531 read the rcpt_include_affixes from the non-existent transport. Now it
5532 just assumes that the setting of that option is false. This bug was
5533 introduced by 4.51/PH/31.
5535 PH/28 Changed -d+all to exclude +memory, because that information is very
5536 rarely of interest, but it makes the output a lot bigger. People tend to
5537 do -d+all out of habit.
5539 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5540 code in os-type was giving problems when libc.so lives in lib64, like on
5543 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5544 aren't the modern standard, and it seems that some systems' include files
5545 don't always have them. Exim was already checking for some of the newer
5546 ones like T_AAAA, and defining it itself. I've added checks for all the
5547 record types that Exim uses.
5549 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5550 not automatically generating a new one, as it is supposed to. This
5551 prevented TLS from working. If the file did exist, but contained invalid
5552 data, a new version was generated, as expected. It was only the case of a
5553 non-existent file that was broken.
5555 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5556 with a change in libdomainkeys > 0.64.
5558 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5559 from DNS. If the selector record carries the flag, it now has
5560 precedence over the domain-wide flag.
5562 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5564 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5565 the use of a port name that isn't defined in /etc/services) occurred, the
5566 message was deferred as in a normal delivery, and thus remained on the
5567 spool, instead of being failed because of the mua_wrapper setting. This
5568 is now fixed, and I tidied up some of the mua_wrapper messages at the
5571 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5572 the messages in a hash of arrays rather than using individual hashes.
5573 This is a bit cleaner and results in dramatic memory savings, albeit
5574 at a slight CPU cost.
5576 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5577 as requested by Marc Sherman.
5579 SC/10 Eximstats: added histograms for user specified patterns as requested
5582 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5584 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5585 fopen() in the content-scanning modules that did not already have it.
5591 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5592 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5594 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5596 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5598 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5599 to test Sieve filters that use "vacation".
5601 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5602 that changes the way the GnuTLS parameters are stored in the cache file.
5603 The new format can be generated externally. For backward compatibility,
5604 if the data in the cache doesn't make sense, Exim assumes it has read an
5605 old-format file, and it generates new data and writes a new file. This
5606 means that you can't go back to an older release without removing the
5609 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5610 work if there are any delivery delays because "one_time" forces the
5611 parent to be marked "delivered", so its unseen clone is never tried
5612 again. For this reason, Exim now forbids the simultaneous setting of
5615 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5616 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5617 entry if you want to know the details. The fix, however, neglected to
5618 consider the case when local delivery batching is involved. The test for
5619 "previously delivered" was not happening when checking to see if an
5620 address could be batched with a previous (undelivered) one; under
5621 certain circumstances this could lead to multiple deliveries to the same
5624 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5625 in its include files, and this causes problems building Exim.
5627 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5628 header_syntax) but Exim was just ignoring anything given after a slash.
5629 In particular, this caused confusion with an attempt to use "verify =
5630 reverse_host_lookup/defer_ok". An error is now given when options are
5631 supplied for verify items that do not have them. (Maybe reverse_host_
5632 lookup should have a defer_ok option, but that's a different point.)
5634 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5635 defined by RFC 821) to 2048, because there were problems with some AUTH
5636 commands, and RFC 1869 says the size should be increased for extended
5637 SMTP commands that take arguments.
5639 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5642 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5643 "unknown" error; now it says that the functionality isn't in the binary.
5645 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5646 an address' error message when a string expansion fails (syntax or
5647 whatever). Otherwise the password may appear in the log. Following change
5648 PH/42 below, there is no longer a chance of it appearing in a bounce
5651 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5653 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5654 of its addresses was ever tried. (Bugzilla bug #2.)
5656 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5657 the result incorrectly in the debug output. (It correctly added a newline
5658 to what was transported.)
5660 TF/01 Added $received_time.
5662 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5663 commented out examples of how to interface to a virus scanner and to
5664 SpamAssassin. Also added commented examples of av_scanner and
5665 spamd_address settings.
5667 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5668 and controls are allowed in which ACLs. There were a couple of minor
5669 errors. Some of the entries in the conditions table (which is a table of
5670 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5671 negation of where the condition IS allowed.
5673 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5675 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5676 header file does not have a version number, so I've had to invent a new
5677 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5678 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5679 radiusclient), but it was contributed by a Radius user.
5681 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5682 files or directories, for OpenSSL.
5684 PH/19 When an Exim process that is running as root has to create an Exim log
5685 file, it does so in a subprocess that runs as exim:exim so as to get the
5686 ownership right at creation (otherwise, other Exim processes might see
5687 the file with the wrong ownership). There was no test for failure of this
5688 fork() call, which would lead to the process getting stuck as it waited
5689 for a non-existent subprocess. Forks do occasionally fail when resources
5690 run out. I reviewed all the other calls to fork(); they all seem to check
5693 PH/20 When checking for unexpected SMTP input at connect time (before writing
5694 the banner), Exim was not dealing correctly with a non-positive return
5695 from the read() function. If the client had disconnected by this time,
5696 the result was a log entry for a synchronization error with an empty
5697 string after "input=" when read() returned zero. If read() returned -1
5698 (an event I could not check), uninitialized data bytes were printed.
5699 There were reports of junk text (parts of files, etc) appearing after
5702 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5704 PH/22 Added support for macro redefinition, and (re)definition in between
5705 driver and ACL definitions.
5707 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5708 forgetting to use the resulting value; it was using the unexpanded value.
5710 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5711 hadn't been configured. The fix is from Juergen Kreileder, who
5712 understands it better than I do:
5714 "Here's what I see happening with three configured cyrus_sasl
5715 authenticators configured (plain, login, cram-md5):
5717 On startup auth_cyrus_sasl_init() gets called for each of these.
5718 This means three calls to sasl_listmech() without a specified mech_list.
5719 => SASL tests which mechs of all available mechs actually work
5720 => three warnings about OTP not working
5721 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5723 With the patch, sasl_listmech() also gets called three times. But now
5724 SASL's mech_list option is set to the server_mech specified in the the
5725 authenticator. Or in other words, the answer from sasl_listmech()
5726 gets limited to just the mech you're testing for (which is different
5728 => the return list contains just 'plain' or 'login', 'cram-md5' or
5729 nothing depending on the value of ob->server_mech.
5731 I've just tested the patch: Authentication still works fine,
5732 unavailable mechs specified in the exim configuration are still
5733 caught, and the auth.log warnings about OTP are gone."
5735 PH/25 When debugging is enabled, the contents of the command line are added
5736 to the debugging output, even when log_selector=+arguments is not
5739 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5740 answer is "GNU", and only if the return is "GNU/something" is the answer
5743 PH/27 $acl_verify_message is now set immediately after the failure of a
5744 verification in an ACL, and so is available in subsequent modifiers. In
5745 particular, the message can be preserved by coding like this:
5747 warn !verify = sender
5748 set acl_m0 = $acl_verify_message
5750 Previously, $acl_verify_message was set only while expanding "message"
5751 and "log_message" when a very denied access.
5753 PH/28 Modified OS/os.c-Linux with
5755 -#ifndef OS_LOAD_AVERAGE
5756 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5758 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5759 nomenclature these days.)
5761 PH/29 Installed patch from the Sieve maintainer that adds the options
5762 sieve_useraddress and sieve_subaddress to the redirect router.
5764 PH/30 In these circumstances:
5765 . Two addresses routed to the same list of hosts;
5766 . First host does not offer TLS;
5767 . First host accepts first address;
5768 . First host gives temporary error to second address;
5769 . Second host offers TLS and a TLS session is established;
5770 . Second host accepts second address.
5771 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5772 and peerdn, if requested) that were in fact used only for the second
5775 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5776 attention to any local part prefix or suffix that was matched by the
5777 router that accepted the address. It now behaves in the same way as it
5778 does for delivery: the affixes are removed from the local part unless
5779 rcpt_include_affixes is set on the transport.
5781 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5782 timeout during the DATA phase of an incoming message.
5784 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5785 applied a patch sanctioned by the Sieve maintainer.
5787 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5788 the uid or gid is negative. A case of a negative gid caused this to be
5789 noticed. The fix allows for either to be negative.
5791 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5792 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5795 PH/36 The change PH/12 above was broken. Fixed it.
5797 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5798 the grounds that routing the same address twice would always produce the
5799 same answer. This might have been true once, but it is certainly no
5800 longer true now. Routing a child address may depend on the previous
5801 routing that produced that child. Some complicated redirection strategies
5802 went wrong when messages had multiple recipients, and made Exim's
5803 behaviour dependent on the order in which the addresses were given.
5805 I have moved the duplicate checking until after the routing is complete.
5806 Exim scans the addresses that are assigned to local and remote
5807 transports, and removes any duplicates. This means that more work will be
5808 done, as duplicates will always all be routed, but duplicates are
5809 presumably rare, so I don't expect this is of any significance.
5811 For deliveries to pipes, files, and autoreplies, the duplicate checking
5812 still happens during the routing process, since they are not going to be
5815 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5816 It corrects a timeout issue with spamd. This is Ian's comment: "The
5817 background is that sometimes spamd either never reads data from a
5818 connection it has accepted, or it never writes response data. The exiscan
5819 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5820 blindly assumes that writes won't block so it may never time out."
5822 PH/39 Allow G after quota size as well as K and M.
5824 PH/40 The value set for $authenticated_id in an authenticator may not contain
5825 binary zeroes or newlines because the value is written to log lines and
5826 to spool files. There was no check on this. Now the value is run through
5827 the string_printing() function so that such characters are converted to
5828 printable escape sequences.
5830 PH/41 $message_linecount is a new variable that contains the total number of
5831 lines in the message. Compare $body_linecount, which is the count for the
5834 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5835 bounce and delay warning messages, except in certain special cases, which
5838 (a) An SMTP error message from a remote host;
5839 (b) A message specified in a :fail: redirection;
5840 (c) A message specified in a "fail" command in a system filter;
5841 (d) A message specified in a FAIL return from the queryprogram router;
5842 (e) A message specified by the cannot_route_message router option.
5844 In these cases only, Exim does include the error details in bounce and
5845 warning messages. There are also a few cases where bland messages such
5846 as "unrouteable address" or "local delivery error" are given.
5848 PH/43 $value is now also set for the "else" part of a ${run expansion.
5850 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5851 being worked on, but at least Exim now implements the latest version to
5854 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5855 process to complete was treated as a delivery failure, a timeout while
5856 writing the message to the pipe was logged, but erroneously treated as a
5857 successful delivery. Such timeouts include transport filter timeouts. For
5858 consistency with the overall process timeout, these timeouts are now
5859 treated as errors, giving rise to delivery failures by default. However,
5860 there is now a new Boolean option for the pipe transport called
5861 timeout_defer, which, if set TRUE, converts the failures into defers for
5862 both kinds of timeout. A transport filter timeout is now identified in
5865 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5866 systems where "make" and "gmake" are different, calling "gmake" at top
5867 level broke things. I've arranged for the value of $(MAKE) to be passed
5868 from the Makefile to this script so that it can call the same version of
5872 A note about Exim versions 4.44 and 4.50
5873 ----------------------------------------
5875 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5876 changes of various kinds. As a consequence, a big documentation update was
5877 needed. This delayed the release for rather longer than seemed good, especially
5878 in the light of a couple of (minor) security issues. Therefore, the changes
5879 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5880 release. So 4.44 and 4.50 are in effect two different branches that both start
5883 I have left the 4.50 change log unchanged; it contains all the changes since
5884 4.43. The change log for 4.44 is below; many of its items are identical to
5885 those for 4.50. This seems to be the most sensible way to preserve the
5886 historical information.
5892 1. Minor wording change to the doc/README.SIEVE file.
5894 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5895 computation of the current number of files was incorrect.
5897 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5898 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5901 4. Give more explanation in the error message when the command for a transport
5902 filter fails to execute.
5904 5. There are several places where Exim runs a non-Exim command in a
5905 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5906 was being done only for the command run by the queryprogram router. It is
5907 now done for all such subprocesses. The other cases are: ${run, transport
5908 filters, and the commands run by the lmtp and pipe transports.
5910 6. Added CONFIGURE_GROUP build-time option.
5912 7. Some older OS have a limit of 256 on the maximum number of file
5913 descriptors. Exim was using setrlimit() to set 1000 as a large value
5914 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5915 systems. I've change it so that if it can't get 1000, it tries for 256.
5917 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5918 was an oversight, and furthermore, ever since the addition of extra
5919 controls (e.g. 4.43/32), the checks on when to allow different forms of
5920 "control" were broken. There should now be diagnostics for all cases when a
5921 control that does not make sense is encountered.
5923 9. Added the /retain_sender option to "control=submission".
5925 10. $recipients is now available in the predata ACL (oversight).
5927 11. Tidy the search cache before the fork to do a delivery from a message
5928 received from the command line. Otherwise the child will trigger a lookup
5929 failure and thereby defer the delivery if it tries to use (for example) a
5930 cached ldap connection that the parent has called unbind on.
5932 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5933 of $address_data from the recipient verification was clobbered by the
5934 sender verification.
5936 13. The value of address_data from a sender verification is now available in
5937 $sender_address_data in subsequent conditions in the ACL statement.
5939 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5941 15. Added a new option "connect=<time>" to callout options, to set a different
5944 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5945 was its contents. (It was OK if the option was not defined at all.)
5947 17. A "Completed" log line is now written for messages that are removed from
5948 the spool by the -Mrm option.
5950 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5951 information about exactly what failed.
5953 19. Added -dd to debug only the daemon process.
5955 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5956 handles timeouts, both on the server side and network timeouts. Renamed the
5957 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5959 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5960 It is now set to "smtps".
5962 22. $host_address is now set to the target address during the checking of
5963 ignore_target_hosts.
5965 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5966 being passed; this would have caused $sender_host_name to have been used if
5967 matching the list had actually called for a host name (not very likely,
5968 since this list is usually IP addresses). A host name is now passed as
5971 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5972 code that specifies a non-restarting handler (typically sigaction() in
5973 modern systems) in an attempt to fix a rare and obscure crash bug.
5975 25. Narrowed the window for a race in the daemon that could cause it to ignore
5976 SIGCHLD signals. This is not a major problem, because they are used only to
5977 wake it up if nothing else does.
5979 26. A malformed maildirsize file could cause Exim to calculate negative values
5980 for the mailbox size or file count. Odd effects could occur as a result.
5981 The maildirsize information is now recalculated if the size or filecount
5984 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5985 support for a long time. Removed HAVE_SYS_VFS_H.
5987 28. Installed the latest version of exipick from John Jetmore.
5989 29. In an address list, if the pattern was not a regular expression, an empty
5990 subject address (from a bounce message) matched only if the pattern was an
5991 empty string. Non-empty patterns were not even tested. This was the wrong
5992 because it is perfectly reasonable to use an empty address as part of a
5993 database query. An empty address is now tested by patterns that are
5994 lookups. However, all the other forms of pattern expect the subject to
5995 contain a local part and a domain, and therefore, for them, an empty
5996 address still always fails if the pattern is not itself empty.
5998 30. Exim went into a mad DNS loop when attempting to do a callout where the
5999 host was specified on an smtp transport, and looking it up yielded more
6000 than one IP address.
6002 31. Re-factored the code for checking spool and log partition space into a
6003 function that finds that data and another that does the check. The former
6004 is then used to implement four new variables: $spool_space, $log_space,
6005 $spool_inodes, and $log_inodes.
6007 32. The RFC2047 encoding function was originally intended for short strings
6008 such as real names; it was not keeping to the 75-character limit for
6009 encoded words that the RFC imposes. It now respects the limit, and
6010 generates multiple encoded words if necessary. To be on the safe side, I
6011 have increased the buffer size for the ${rfc2047: expansion operator from
6014 33. It is now permitted to omit both strings after an "if" condition; if the
6015 condition is true, the result is "true". As before, when the second string
6016 is omitted, a false condition yields an empty string. This makes it less
6017 cumbersome to write custom ACL and router conditions.
6019 34. Failure to deliver a bounce message always caused it to be frozen, even if
6020 there was an errors_to setting on the router. The errors_to setting is now
6023 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6024 canonical form (fully expanded) before being placed in
6025 $sender_host_address.
6027 36. The table in the code that translates DNS record types into text (T_A to
6028 "A" for instance) was missing entries for NS and CNAME. It is just possible
6029 that this could have caused confusion if both these types were looked up
6030 for the same domain, because the text type is used as part of Exim's
6031 per-process caching. But the chance of anyone hitting this buglet seems
6034 37. The dnsdb lookup has been extended in a number of ways.
6036 (1) There is a new type, "zns", which walks up the domain tree until it
6037 finds some nameserver records. It should be used with care.
6039 (2) There is a new type, "mxh", which is like "mx" except that it returns
6040 just the host names, not the priorities.
6042 (3) It is now possible to give a list of domains (or IP addresses) to be
6043 looked up. The behaviour when one of the lookups defers can be
6044 controlled by a keyword.
6046 (4) It is now possible to specify the separator character for use when
6047 multiple records are returned.
6049 38. The dnslists ACL condition has been extended: it is now possible to supply
6050 a list of IP addresses and/or domains to be looked up in a particular DNS
6053 39. Added log_selector=+queue_time_overall.
6055 40. When running the queue in the test harness, wait just a tad after forking a
6056 delivery process, to get repeatability of debugging output.
6058 41. Include certificate and key file names in error message when GnuTLS fails
6059 to set them up, because the GnuTLS error message doesn't include the name
6060 of the failing file when there is a problem reading it.
6062 42. Allow both -bf and -bF in the same test run.
6064 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6066 44. The "Exiscan patch" is now merged into the mainline Exim source.
6068 45. Sometimes the final signoff response after QUIT could fail to get
6069 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6070 < 0 before doing a fflush(). This bug looks as though it goes back to the
6071 introduction of TLS in release 3.20, but "sometimes" must have been rare
6072 because the tests only now provoked it.
6074 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6075 (this can affect the format of dates).
6077 47. exim_tidydb, when checking for the continued existence of a message for
6078 which it has found a message-specific retry record, was not finding
6079 messages that were in split spool directories. Consequently, it was
6080 deleting retry records that should have stayed in existence.
6082 48. Steve fixed some bugs in eximstats.
6084 49. The SPA authentication driver was not abandoning authentication and moving
6085 on to the next authenticator when an expansion was forced to fail,
6086 contradicting the general specification for all authenticators. Instead it
6087 was generating a temporary error. It now behaves as specified.
6089 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6090 (the order specifies the preference for clients). The order is now AES256,
6091 AES128, 3DES, ARCFOUR128.
6093 51. Small patch to Sieve code - explicitly set From: when generating an
6096 52. Exim crashed if a remote delivery caused a very long error message to be
6097 recorded - for instance if somebody sent an entire SpamAssassin report back
6098 as a large number of 550 error lines. This bug was coincidentally fixed by
6099 increasing the size of one of Exim's internal buffers (big_buffer) that
6100 happened as part of the Exiscan merge. However, to be on the safe side, I
6101 have made the code more robust (and fixed the comments that describe what
6104 53. Now that there can be additional text after "Completed" in log lines (if
6105 the queue_time_overall log selector is set), a one-byte patch to exigrep
6106 was needed to allow it to recognize "Completed" as not the last thing in
6109 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6110 patch that reportedly fixes this has been added. I am not expert enough to
6111 create a test for it. This is what the patch creator wrote:
6113 "I found a little strange behaviour of ldap code when working with
6114 Windows 2003 AD Domain, where users was placed in more than one
6115 Organization Units. When I tried to give exim partial DN, the exit code
6116 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6117 But simultaneously result of request was absolutely normal ldap result,
6118 so I produce this patch..."
6120 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6121 REFERENCE, so I have modified the code to exclude the patch when that macro
6124 55. Some experimental protocols are using DNS PTR records for new purposes. The
6125 keys for these records are domain names, not reversed IP addresses. The
6126 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6127 leaves it alone. Component reversal etc. now happens only for IP addresses.
6130 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6132 57. Double the size of the debug message buffer (to 2048) so that more of very
6133 long debug lines gets shown.
6135 58. The exicyclog utility now does better if the number of log files to keep
6136 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6138 59. Two changes related to the smtp_active_hostname option:
6140 (1) $smtp_active_hostname is now available as a variable.
6141 (2) The default for smtp_banner uses $smtp_active_hostname instead
6142 of $primary_hostname.
6144 60. The host_aton() function is supposed to be passed a string that is known
6145 to be a valid IP address. However, in the case of IPv6 addresses, it was
6146 not checking this. This is a hostage to fortune. Exim now panics and dies
6147 if the condition is not met. A case was found where this could be provoked
6148 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6149 components; fortuitously, this particular loophole had already been fixed
6150 by change 4.50/55 above.
6152 If there are any other similar loopholes, the new check in host_aton()
6153 itself should stop them being exploited. The report I received stated that
6154 data on the command line could provoke the exploit when Exim was running as
6155 exim, but did not say which command line option was involved. All I could
6156 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6157 running as the user.
6160 61. There was a buffer overflow vulnerability in the SPA authentication code
6161 (which came originally from the Samba project). I have added a test to the
6162 spa_base64_to_bits() function which I hope fixes it.
6165 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6166 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6168 63. The daemon start-up calls getloadavg() while still root for those OS that
6169 need the first call to be done as root, but it missed one case: when
6170 deliver_queue_load_max is set with deliver_drop_privilege. This is
6171 necessary for the benefit of the queue runner, because there is no re-exec
6172 when deliver_drop_privilege is set.
6174 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6175 This has been fixed.
6177 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6178 was in use, was not putting the data itself into the right store pool;
6179 consequently, it could be overwritten for a subsequent message in the same
6180 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6183 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6185 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6186 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6187 the function were treating the return as a boolean value, which happened to
6188 work because 0=false and not-0=true, but is not correct code.
6190 68. The host_aton() function was not handling scoped IPv6 addresses (those
6191 with, for example, "%eth0" on the end) correctly.
6193 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6194 negated items (that is, ~something) in unsigned ints. Some compilers
6195 apparently mutter when there is no cast.
6197 70. If an address verification called from an ACL failed, and did not produce a
6198 user-specific message (i.e. there was only a "system" message), nothing was
6199 put in $acl_verify_message. In this situation, it now puts the system
6202 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6203 session; change 4.31/43 added the unwanted input to the log line - except
6204 that it did not do this in the start of session case. It now does.
6206 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6207 This is wrong and can cause the other end to generate a synchronization
6208 error if it is another Exim or anything else that does the synchronization
6209 check. A QUIT command is no longer sent after a timeout.
6211 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6212 during host lookups.
6214 74. The defer_ok option of callout verification was not working if it was used
6215 when verifying addresses in header lines, that is, for this case:
6217 verify = header_sender/callout=defer_ok
6219 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6220 those file descriptors could be used for SMTP connections. If anything
6221 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6222 could be sent to the SMTP client, causing chaos. The daemon now opens
6223 stdin, stdout, and stderr to /dev/null when it puts itself into the
6226 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6227 log by default. The user can override this with suitable Perl magic.
6229 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6230 the log message when discard triggers, was not working for the DATA ACL or
6231 for the non-SMTP ACL.
6233 78. Error message wording change in sieve.c.
6235 79. If smtp_accept_max_per_host was set, the number of connections could be
6236 restricted to fewer than expected, because the daemon was trying to set up
6237 a new connection before checking whether the processes handling previous
6238 connections had finished. The check for completed processes is now done
6239 earlier. On busy systems, this bug wouldn't be noticed because something
6240 else would have woken the daemon, and it would have reaped the completed
6243 80. If a message was submitted locally by a user whose login name contained one
6244 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6245 It caused a spool format error. I have fixed the spool reading code. A
6246 related problem was that the "from" clause in the Received: line became
6247 illegal because of the space(s). It is now covered by ${quote_local_part.
6249 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6252 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6254 83. Patch from Sieve maintainer to fix unterminated string problem in
6255 "vacation" handling.
6257 84. Some minor changes to the Linux configuration files to help with other
6258 OS variants using glibc.
6260 85. One more patch for Sieve to update vacation handling to latest spec.
6263 ----------------------------------------------------
6264 See the note above about the 4.44 and 4.50 releases.
6265 ----------------------------------------------------
6271 1. Change 4.43/35 introduced a bug that caused file counts to be
6272 incorrectly computed when quota_filecount was set in an appendfile
6275 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6276 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6279 3. Give more explanation in the error message when the command for a transport
6280 filter fails to execute.
6282 4. There are several places where Exim runs a non-Exim command in a
6283 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6284 was being done only for the command run by the queryprogram router. It is
6285 now done for all such subprocesses. The other cases are: ${run, transport
6286 filters, and the commands run by the lmtp and pipe transports.
6288 5. Some older OS have a limit of 256 on the maximum number of file
6289 descriptors. Exim was using setrlimit() to set 1000 as a large value
6290 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6291 systems. I've change it so that if it can't get 1000, it tries for 256.
6293 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6294 was an oversight, and furthermore, ever since the addition of extra
6295 controls (e.g. 4.43/32), the checks on when to allow different forms of
6296 "control" were broken. There should now be diagnostics for all cases when a
6297 control that does not make sense is encountered.
6299 7. $recipients is now available in the predata ACL (oversight).
6301 8. Tidy the search cache before the fork to do a delivery from a message
6302 received from the command line. Otherwise the child will trigger a lookup
6303 failure and thereby defer the delivery if it tries to use (for example) a
6304 cached ldap connection that the parent has called unbind on.
6306 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6307 of $address_data from the recipient verification was clobbered by the
6308 sender verification.
6310 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6311 was its contents. (It was OK if the option was not defined at all.)
6313 11. A "Completed" log line is now written for messages that are removed from
6314 the spool by the -Mrm option.
6316 12. $host_address is now set to the target address during the checking of
6317 ignore_target_hosts.
6319 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6320 being passed; this would have caused $sender_host_name to have been used if
6321 matching the list had actually called for a host name (not very likely,
6322 since this list is usually IP addresses). A host name is now passed as
6325 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6326 code that specifies a non-restarting handler (typically sigaction() in
6327 modern systems) in an attempt to fix a rare and obscure crash bug.
6329 15. Narrowed the window for a race in the daemon that could cause it to ignore
6330 SIGCHLD signals. This is not a major problem, because they are used only to
6331 wake it up if nothing else does.
6333 16. A malformed maildirsize file could cause Exim to calculate negative values
6334 for the mailbox size or file count. Odd effects could occur as a result.
6335 The maildirsize information is now recalculated if the size or filecount
6338 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6339 support for a long time. Removed HAVE_SYS_VFS_H.
6341 18. Updated exipick to current release from John Jetmore.
6343 19. Allow an empty sender to be matched against a lookup in an address list.
6344 Previously the only cases considered were a regular expression, or an
6347 20. Exim went into a mad DNS lookup loop when doing a callout where the
6348 host was specified on the transport, if the DNS lookup yielded more than
6351 21. The RFC2047 encoding function was originally intended for short strings
6352 such as real names; it was not keeping to the 75-character limit for
6353 encoded words that the RFC imposes. It now respects the limit, and
6354 generates multiple encoded words if necessary. To be on the safe side, I
6355 have increased the buffer size for the ${rfc2047: expansion operator from
6358 22. Failure to deliver a bounce message always caused it to be frozen, even if
6359 there was an errors_to setting on the router. The errors_to setting is now
6362 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6363 canonical form (fully expanded) before being placed in
6364 $sender_host_address.
6366 24. Updated eximstats to version 1.33
6368 25. Include certificate and key file names in error message when GnuTLS fails
6369 to set them up, because the GnuTLS error message doesn't include the name
6370 of the failing file when there is a problem reading it.
6372 26. Expand error message when OpenSSL has problems setting up cert/key files.
6375 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6376 (this can affect the format of dates).
6378 28. exim_tidydb, when checking for the continued existence of a message for
6379 which it has found a message-specific retry record, was not finding
6380 messages that were in split spool directories. Consequently, it was
6381 deleting retry records that should have stayed in existence.
6383 29. eximstats updated to version 1.35
6384 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6385 1.35 - bugfix such that pie charts by volume are generated correctly
6387 30. The SPA authentication driver was not abandoning authentication and moving
6388 on to the next authenticator when an expansion was forced to fail,
6389 contradicting the general specification for all authenticators. Instead it
6390 was generating a temporary error. It now behaves as specified.
6392 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6393 (the order specifies the preference for clients). The order is now AES256,
6394 AES128, 3DES, ARCFOUR128.
6396 31. Small patch to Sieve code - explicitly set From: when generating an
6399 32. Exim crashed if a remote delivery caused a very long error message to be
6400 recorded - for instance if somebody sent an entire SpamAssassin report back
6401 as a large number of 550 error lines. This bug was coincidentally fixed by
6402 increasing the size of one of Exim's internal buffers (big_buffer) that
6403 happened as part of the Exiscan merge. However, to be on the safe side, I
6404 have made the code more robust (and fixed the comments that describe what
6407 33. Some experimental protocols are using DNS PTR records for new purposes. The
6408 keys for these records are domain names, not reversed IP addresses. The
6409 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6410 leaves it alone. Component reversal etc. now happens only for IP addresses.
6413 34. The host_aton() function is supposed to be passed a string that is known
6414 to be a valid IP address. However, in the case of IPv6 addresses, it was
6415 not checking this. This is a hostage to fortune. Exim now panics and dies
6416 if the condition is not met. A case was found where this could be provoked
6417 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6418 components; fortuitously, this particular loophole had already been fixed
6419 by change 4.50/55 or 4.44/33 above.
6421 If there are any other similar loopholes, the new check in host_aton()
6422 itself should stop them being exploited. The report I received stated that
6423 data on the command line could provoke the exploit when Exim was running as
6424 exim, but did not say which command line option was involved. All I could
6425 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6426 running as the user.
6429 35. There was a buffer overflow vulnerability in the SPA authentication code
6430 (which came originally from the Samba project). I have added a test to the
6431 spa_base64_to_bits() function which I hope fixes it.
6434 36. The daemon start-up calls getloadavg() while still root for those OS that
6435 need the first call to be done as root, but it missed one case: when
6436 deliver_queue_load_max is set with deliver_drop_privilege. This is
6437 necessary for the benefit of the queue runner, because there is no re-exec
6438 when deliver_drop_privilege is set.
6440 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6441 was in use, was not putting the data itself into the right store pool;
6442 consequently, it could be overwritten for a subsequent message in the same
6443 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6446 38. Sometimes the final signoff response after QUIT could fail to get
6447 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6448 < 0 before doing a fflush(). This bug looks as though it goes back to the
6449 introduction of TLS in release 3.20, but "sometimes" must have been rare
6450 because the tests only now provoked it.
6456 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6457 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6458 Now it returns an integer. A number of calls were still expecting a T/F
6459 return. Fortuitously, in all cases, the tests worked in OK situations,
6460 which is the norm. However, things would have gone wrong on any write
6461 failures on the smtp file descriptor. This function is used when sending
6462 messages over SMTP and also when doing verify callouts.
6464 2. When Exim is called to do synchronous delivery of a locally submitted
6465 message (the -odf or -odi options), it no longer closes stderr before doing
6468 3. Implemented the mua_wrapper option.
6470 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6472 5. Implemented the functions header_remove(), header_testname(),
6473 header_add_at_position(), and receive_remove_recipient(), and exported them
6476 6. If an ACL "warn" statement specified the addition of headers, Exim already
6477 inserted X-ACL-Warn: at the start if there was no header name. However, it
6478 was not making this test for the second and subsequent header lines if
6479 there were newlines in the string. This meant that an invalid header could
6480 be inserted if Exim was badly configured.
6482 7. Allow an ACL "warn" statement to add header lines at the start or after all
6483 the Received: headers, as well as at the end.
6485 8. Added the rcpt_4xx retry error code.
6487 9. Added postmaster_mailfrom=xxx to callout verification option.
6489 10. Added mailfrom=xxxx to the callout verification option, for verify=
6492 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6493 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6495 12. Inserted some casts to stop certain compilers warning when using pointer
6496 differences as field lengths or precisions in printf-type calls (mostly
6497 affecting debugging statements).
6499 13. Added optional readline() support for -be (dynamically loaded).
6501 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6502 same clock tick as a message's arrival, so that its received time was the
6503 same as the "first fail" time on the retry record, and that message
6504 remained on the queue past the ultimate address timeout, every queue runner
6505 would try a delivery (because it was past the ultimate address timeout) but
6506 after another failure, the ultimate address timeout, which should have then
6507 bounced the address, did not kick in. This was a "< instead of <=" error;
6508 in most cases the first failure would have been in the next clock tick
6509 after the received time, and all would be well.
6511 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6512 being recognized when the domain list was tested by the match_domain
6513 condition in an expansion string.
6515 16. Added the ${str2b64: operator.
6517 17. Exim was always calling setrlimit() to set a large limit for the number of
6518 processes, without checking whether the existing limit was already
6519 adequate. (It did check for the limit on file descriptors.) Furthermore,
6520 errors from getrlimit() and setrlimit() were being ignored. Now they are
6521 logged to the main and panic logs, but Exim does carry on, to try to do its
6522 job under whatever limits there are.
6524 18. Imported PCRE 5.0.
6526 19. Trivial typo in log message " temporarily refused connection" (the leading
6529 20. If the log selector return_path_on_delivery was set and an address was
6530 redirected to /dev/null, the delivery process crashed because it assumed
6531 that a return path would always be set for a "successful" delivery. In this
6532 case, the whole delivery is bypassed as an optimization, and therefore no
6535 21. Internal re-arrangement: the function for sending a challenge and reading
6536 a response while authentication was assuming a zero-terminated challenge
6537 string. It's now changed to take a pointer and a length, to allow for
6538 binary data in such strings.
6540 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6542 23. Exim was not respecting finduser_retries when seeking the login of the
6543 uid under which it was called; it was always trying 10 times. (The default
6544 setting of finduser_retries is zero.) Also, it was sleeping after the final
6545 failure, which is pointless.
6547 24. Implemented tls_on_connect_ports.
6549 25. Implemented acl_smtp_predata.
6551 26. If the domain in control=submission is set empty, Exim assumes that the
6552 authenticated id is a complete email address when it generates From: or
6553 Sender: header lines.
6555 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6556 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6557 chown and chgrp in /bin and hostname in /usr/bin.
6559 28. Exim was keeping the "process log" file open after each use, just as it
6560 does for the main log. This opens the possibility of it remaining open for
6561 long periods when the USR1 signal hits a daemon. Occasional processlog
6562 errors were reported, that could have been caused by this. Anyway, it seems
6563 much more sensible not to leave this file open at all, so that is what now
6566 29. The long-running daemon process does not normally write to the log once it
6567 has entered its main loop, and it closes the log before doing so. This is
6568 so that log files can straightforwardly be renamed and moved. However,
6569 there are a couple of unusual error situations where the daemon does write
6570 log entries, and I had neglected to close the log afterwards.
6572 30. The text of an SMTP error response that was received during a remote
6573 delivery was being truncated at 512 bytes. This is too short for some of
6574 the long messages that one sometimes sees. I've increased the limit to
6577 31. It is now possible to make retry rules that apply only when a message has a
6578 specific sender, in particular, an empty sender.
6580 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6581 it possible to be selective about when SMTP synchronization is enforced.
6583 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6585 32. Implemented hosts_connection_nolog.
6587 33. Added an ACL for QUIT.
6589 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6592 35. Added mailbox_size and mailbox_filecount to appendfile.
6594 36. Added control = no_multiline_responses to ACLs.
6596 37. There was a bug in the logic of the code that waits for the clock to tick
6597 in the case where the clock went backwards by a substantial amount such
6598 that the microsecond fraction of "now" was more than the microsecond
6599 fraction of "then" (but the whole seconds number was less).
6601 38. Added support for the libradius Radius client library this is found on
6602 FreeBSD (previously only the radiusclient library was supported).
6608 1. When certain lookups returned multiple values in the form name=value, the
6609 quoting of the values was not always being done properly. Specifically:
6610 (a) If the value started with a double quote, but contained no whitespace,
6612 (b) If the value contained whitespace other than a space character (i.e.
6613 tabs or newlines or carriage returns) it was not quoted.
6614 This fix has been applied to the mysql and pgsql lookups by writing a
6615 separate quoting function and calling it from the lookup code. The fix
6616 should probably also be applied to nisplus, ibase and oracle lookups, but
6617 since I cannot test any of those, I have not disturbed their existing code.
6619 2. A hit in the callout cache for a specific address caused a log line with no
6620 reason for rejecting RCPT. Now it says "Previous (cached) callout
6621 verification failure".
6623 3. There was an off-by-one bug in the queryprogram router. An over-long
6624 return line was truncated at 256 instead of 255 characters, thereby
6625 overflowing its buffer with the terminating zero. As well as fixing this, I
6626 have increased the buffer size to 1024 (and made a note to document this).
6628 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6629 when Exim is waiting for an SMTP response from a remote server, Exim
6630 restarts its select() call on the socket, thereby resetting its timeout.
6631 This is not a problem when such interrupts are rare. Somebody set up a cron
6632 job to run exiwhat every 2 minutes, which is less than the normal select()
6633 timeout (5 or 10 minutes). This meant that the select() timeout never
6634 kicked in because it was always reset. I have fixed this by comparing the
6635 time when an interrupt arrives with the time at the start of the first call
6636 to select(). If more time than the timeout has elapsed, the interrupt is
6637 treated as a timeout.
6639 5. Some internal re-factoring in preparation for the addition of Sieve
6640 extensions (by MH). In particular, the "personal" test is moved to a
6641 separate function, and given an option for scanning Cc: and Bcc: (which is
6642 not set for Exim filters).
6644 6. When Exim created an email address using the login of the caller as the
6645 local part (e.g. when creating a From: or Sender: header line), it was not
6646 quoting the local part when it contained special characters such as @.
6648 7. Installed new OpenBSD configuration files.
6650 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6651 try to make them clearer.
6653 9. Callout options, other than the timeout value, were being ignored when
6654 verifying sender addresses in header lines. For example, when using
6656 verify = header_sender/callout=no_cache
6658 the cache was (incorrectly) being used.
6660 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6661 only the Cygwin environment.
6663 11. When return_path_on_delivery was set as a log selector, if different remote
6664 addresses in the same message used different return paths and parallel
6665 remote delivery occurred, the wrong values would sometimes be logged.
6666 (Whenever a remote delivery process finished, the return path value from
6667 the most recently started remote delivery process was logged.)
6669 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6670 lines when AUTH and/or TLS are in use. This is the "received protocol"
6671 field. Exim used to use "asmtp" for authenticated SMTP, without any
6672 indication (in the protocol name) for TLS use. Now it follows the RFC and
6673 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6674 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6675 names appear in log lines as well as in Received: header lines.
6677 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6678 extensions, and comparison tests, and to fix some bugs.
6680 14. Changes to the "personal" filter test:
6682 (1) The test was buggy in that it was just doing the equivalent of
6683 "contains" tests on header lines. For example, if a user's address was
6684 anne@some.where, the "personal" test would incorrectly be true for
6686 To: susanne@some.where
6688 This test is now done by extracting each address from the header in turn,
6689 and checking the entire address. Other tests that are part of "personal"
6690 are now done using regular expressions (for example, to check local parts
6691 of addresses in From: header lines).
6693 (2) The list of non-personal local parts in From: addresses has been
6694 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6695 taken from the Sieve specification recommendations.
6697 (3) If the message contains any header line starting with "List-" it is
6698 treated as non-personal.
6700 (4) The test for "circular" in the Subject: header line has been removed
6701 because it now seems ill-conceived.
6703 15. Minor typos in src/EDITME comments corrected.
6705 16. Installed latest exipick from John Jetmore.
6707 17. If headers_add on a router specified a text string that was too long for
6708 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6709 of string_sprintf() is now avoided.
6711 18. $message_body_size was not set (it was always zero) when running the DATA
6712 ACL and the local_scan() function.
6714 19. For the "mail" command in an Exim filter, no default was being set for
6715 the once_repeat time, causing a random time value to be used if "once" was
6716 specified. (If the value happened to be <= 0, no repeat happened.) The
6717 default is now 0s, meaning "never repeat". The "vacation" command was OK
6718 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6719 (I found it when inspecting the code).
6721 20. There is now an overall timeout for performing a callout verification. It
6722 defaults to 4 times the callout timeout, which applies to individual SMTP
6723 commands during the callout. The overall timeout applies when there is more
6724 than one host that can be tried. The timeout is checked before trying the
6725 next host. This prevents very long delays if there are a large number of
6726 hosts and all are timing out (e.g. when the network connections are timing
6727 out). The value of the overall timeout can be changed by specifying an
6728 additional sub-option for "callout", called "maxwait". For example:
6730 verify = sender/callout=5s,maxwait=20s
6732 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6733 to the end before writing, but this should make it even safer).
6735 22. Exim was forgetting that it had advertised PIPELINING for the second and
6736 subsequent messages on an SMTP connection. It was also not resetting its
6737 memory on STARTTLS and an internal HELO.
6739 23. When Exim logs an SMTP synchronization error within a session, it now
6740 records whether PIPELINING has been advertised or not.
6742 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6743 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6744 rather than long int.
6746 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6748 26. Added the never_mail option to autoreply.
6754 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6755 crash if the getsockname() call failed; this can happen if a connection is
6756 closed very soon after it is established. The problem was simply in the
6757 order in which certain operations were done, causing Exim to try to write
6758 to the SMTP stream before it had set up the file descriptor. The bug has
6759 been fixed by making things happen in the correct order.
6765 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6766 before the connection was closed, thus losing the rejection response.
6768 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6769 some early Solaris releases, but causes trouble in current releases where
6770 socklen_t is defined.
6772 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6775 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6778 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6779 that releasing the top of it at the end releases what was used for sub-
6780 expansions (unless the block got too big). However, discard this block if
6781 the first thing is a variable or header, so that we can use its block when
6782 it is dynamic (useful for very large $message_headers, for example).
6784 6. Lookups now cache *every* query, not just the most recent. A new, separate
6785 store pool is used for this. It can be recovered when all lookup caches are
6786 flushed. Lookups now release memory at the end of their result strings.
6787 This has involved some general refactoring of the lookup sources.
6789 7. Some code has been added to the store_xxx() functions to reduce the amount
6790 of flapping under certain conditions.
6792 8. log_incoming_interface used to affect only the <= reception log lines. Now
6793 it causes the local interface and port to be added to several more SMTP log
6794 lines, for example "SMTP connection from", and rejection lines.
6796 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6798 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6800 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6801 could be overwritten at the end of the current message (or the start of a
6802 new message if it was set in a HELO ACL). The value is now preserved for
6803 the duration of the SMTP connection.
6805 12. If a transport had a headers_rewrite setting, and a matching header line
6806 contained an unqualified address, that address was qualified, even if it
6807 did not match any rewriting rules. The underlying bug was that the values
6808 of the flags that permit the existence of unqualified sender and recipient
6809 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6810 non-local messages, and by -bnq for local messages) were not being
6811 preserved with the message after it was received.
6813 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6814 "next input=" as part of the text comprising the host identity instead of
6815 the correct text. The code was using the same buffer for two different
6816 strings. However, depending on which order the printing function evaluated
6817 its arguments, the bug did not always show up. Under Linux, for example, my
6818 test suite worked just fine.
6820 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6821 doesn't work with some older versions of Perl. It has been changed to "my",
6822 which in any case is probably the better facility to use.
6824 15. A really picky compiler found some instances of statements for creating
6825 error messages that either had too many or two few arguments for the format
6828 16. The size of the buffer for calls to the DNS resolver has been increased
6829 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6830 for addresses that have a lot of PTR records. This alleviates a problem; it
6831 does not fully solve it.
6833 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6834 buffer now truncates the list and logs the incident, which is the same
6835 action as happens when Exim is looking up a host name and its aliases.
6836 Previously in this situation something unpredictable would happen;
6837 sometimes it was "internal error: store_reset failed".
6839 18. If a server dropped the connection unexpectedly when an Exim client was
6840 using GnuTLS and trying to read a response, the client delivery process
6841 crashed while trying to generate an error log message.
6843 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6844 string, for example:
6846 warn message = H1: something\nH2: something
6848 the text was added as a single header line from Exim's point of view
6849 though it ended up OK in the delivered message. However, searching for the
6850 second and subsequent header lines using $h_h2: did not work. This has been
6851 fixed. Similarly, if a system filter added multiple headers in this way,
6852 the routers could not see them.
6854 20. Expanded the error message when iplsearch is called with an invalid key to
6855 suggest using net-iplsearch in a host list.
6857 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6858 ACLs are no longer actually imposed (and a message to that effect is
6861 22. If a "gecos" field in a passwd entry contained escaped characters, in
6862 particular, if it contained a \" sequence, Exim got it wrong when building
6863 a From: or a Sender: header from that name. A second bug also caused
6864 incorrect handling when an unquoted " was present following a character
6865 that needed quoting.
6867 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6868 was not being matched caselessly.
6870 24. Arranged for all hyphens in the exim.8 source to be escaped with
6873 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6874 properly. Recipient callout cache records were still being keyed to include
6875 the sender, even when use_sender was set false. This led to far more
6876 callouts that were necessary. The sender is no longer included in the key
6877 when use_sender is false.
6879 26. Added "control = submission" modifier to ACLs.
6881 27. Added the ${base62d: operator to decode base 62 numbers.
6883 28. dnsdb lookups can now access SRV records.
6885 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6886 the configuration file.
6888 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6889 (-v) mode. This makes the output for a verbose queue run more intelligible.
6891 31. Added a use_postmaster feature to recipient callouts.
6893 32. Added the $body_zerocount variable, containing the number of binary zero
6894 bytes in the message body.
6896 33. The time of last modification of the "new" subdirectory is now used as the
6897 "mailbox time last read" when there is a quota error for a maildir
6900 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6902 35. Added +ignore_unknown as a special item in host lists.
6904 36. Code for decoding IPv6 addresses in host lists is now included, even if
6905 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6906 address was recognized as an IP address, but was then not correctly decoded
6907 into binary, causing unexpected and incorrect effects when compared with
6914 1. Very minor rewording of debugging text in manualroute to say "list of
6915 hosts" instead of "hostlist".
6917 2. If verify=header_syntax was set, and a header line with an unqualified
6918 address (no domain) and a large number of spaces between the end of the
6919 name and the colon was received, the reception process suffered a buffer
6920 overflow, and (when I tested it) crashed. This was caused by some obsolete
6921 code that should have been removed. The fix is to remove it!
6923 3. When running in the test harness, delay a bit after writing a bounce
6924 message to get a bit more predictability in the log output.
6926 4. Added a call to search_tidyup() just before forking a reception process. In
6927 theory, someone could use a lookup in the expansion of smtp_accept_max_
6928 per_host which, without the tidyup, could leave open a database connection.
6930 5. Added the variables $recipient_data and $sender_data which get set from a
6931 lookup success in an ACL "recipients" or "senders" condition, or a router
6932 "senders" option, similar to $domain_data and $local_part_data.
6934 6. Moved the writing of debug_print from before to after the "senders" test
6937 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6938 problems for message scanning, either using a data ACL, or using
6939 local_scan() because the Received: header was not generated till after they
6940 were called (in order to set the time as the time of reception completion).
6941 I have revised the way this works. The header is now generated after the
6942 body is received, but before the ACL or local_scan() are called. After they
6943 are run, the timestamp in the header is updated.
6949 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6950 before starting a queue runner without re-exec. This happened only when
6951 deliver_drop_privilege was set or when the Exim user was set to root. The
6952 effect of the bug was that timeouts during subsequent deliveries caused
6953 crashes instead of being properly handled. The handler is now left at its
6954 default (and expected) setting.
6956 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6957 message, again when deliver_drop_privilege is set or Exim is run as root.
6958 The bug described in (1) was not present in this case, but the tidying up
6959 of the other signals was missing. I have made the two cases consistent.
6961 3. The ignore_target_hosts setting on a manualroute router was being ignored
6962 for hosts that were looked up using the /MX notation.
6964 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6967 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6968 operated on the sender address. After changing the $sender_address to <>
6969 for the sender address verify, Exim was re-instated it as the original
6970 (before rewriting) address, but remembering that it had rewritten it, so it
6971 wasn't rewriting it again. This bug also had the effect of breaking the
6972 sender address verification caching when the sender address was rewritten.
6974 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6975 This has been changed so that if the ip literal address matches
6976 ignore_target_hosts, the router declines.
6978 7. Added expansion conditions match_domain, match_address, and match_local_
6979 part (NOT match_host).
6981 8. The placeholder for the Received: header didn't have a length field set.
6983 9. Added code to Exim itself and to exim_lock to test for a specific race
6984 condition that could lead to file corruption when using MBX delivery. The
6985 issue is with the lockfile that is created in /tmp. If this file is removed
6986 after a process has opened it but before that process has acquired a lock,
6987 there is the potential for a second process to recreate the file and also
6988 acquire a lock. This could lead to two Exim processes writing to the file
6989 at the same time. The added code performs the same test as UW imapd; it
6990 checks after acquiring the lock that its file descriptor still refers to
6991 the same named file.
6993 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6994 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6997 11. Added the smtp_active_hostname option. If used, this will typically be made
6998 to depend on the incoming interface address. Because $interface_address is
6999 not set up until the daemon has forked a reception process, error responses
7000 that can happen earlier (such as "too many connections") no longer contain
7003 12. If an expansion in a condition on a "warn" statement fails because a lookup
7004 defers, the "warn" statement is abandoned, and the next ACL statement is
7005 processed. Previously this caused the whole ACL to be aborted.
7007 13. Added the iplsearch lookup type.
7009 14. Added ident_timeout as a log selector.
7011 15. Added tls_certificate_verified as a log selector.
7013 16. Added a global option tls_require_ciphers (compare the smtp transport
7014 option of the same name). This controls incoming TLS connections.
7016 17. I finally figured out how to make tls_require_ciphers do a similar thing
7017 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7018 before starting the TLS session.
7020 18. Tabs are now shown as \t in -bP output.
7022 19. If the log selector return_path_on_delivery was set, Exim crashed when
7023 bouncing a message because it had too many Received: header lines.
7025 20. If two routers both had headers_remove settings, and the first one included
7026 a superfluous trailing colon, the final name in the first list and the
7027 first name in the second list were incorrectly joined into one item (with a
7028 colon in the middle).
7034 1. Added -C and -D options to the exinext utility, mainly to make it easier
7035 to include in the automated testing, but these could be helpful when
7036 multiple configurations are in use.
7038 2. The exinext utility was not formatting the output nicely when there was
7039 an alternate port involved in the retry record key, nor when there was a
7040 message id as well (for retries that were specific to a specific message
7041 and a specific host). It was also confused by IPv6 addresses, because of
7042 the additional colons they contain. I have fixed the IPv4 problem, and
7043 patched it up to do a reasonable job for IPv6.
7045 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7046 delivery, the log line now contains "pipelined" if PIPELINING was used.
7048 4. An SMTP transport process used to panic and die if the bind() call to set
7049 an explicit outgoing interface failed. This has been changed; it is now
7050 treated in the same way as a connect() failure.
7052 5. A reference to $sender_host_name in the part of a conditional expansion
7053 that was being skipped was still causing a DNS lookup. This no longer
7056 6. The def: expansion condition was not recognizing references to header lines
7057 that used bh_ and bheader_.
7059 7. Added the _cache feature to named lists.
7061 8. The code for checking quota_filecount in the appendfile transport was
7062 allowing one more file than it should have been.
7064 9. For compatibility with Sendmail, the command line option
7072 and sets the incoming protocol and host name (for trusted callers). The
7073 host name and its colon can be omitted when only the protocol is to be set.
7074 Note the Exim already has two private options, -pd and -ps, that refer to
7075 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7076 "s", but I don't think that's a major issue.
7078 10. A number of refactoring changes to the code, none of which should affect
7081 (a) The number of logging options was getting close to filling up the
7082 32-bit word that was used as a bit map. I have split them into two classes:
7083 those that are passed in the argument to log_write(), and those that are
7084 only ever tested independently outside of that function. These are now in
7085 separate 32-bit words, so there is plenty of room for expansion again.
7086 There is no change in the user interface or the logging behaviour.
7088 (b) When building, for example, log lines, the code previously used a
7089 macro that called string_cat() twice, in order to add two strings. This is
7090 not really sufficiently general. Furthermore, there was one instance where
7091 it was actually wrong because one of the argument was used twice, and in
7092 one call a function was used. (As it happened, calling the function twice
7093 did not affect the overall behaviour.) The macro has been replaced by a
7094 function that can join an arbitrary number of extra strings onto a growing
7097 (c) The code for expansion conditions now uses a table and a binary chop
7098 instead of a serial search (which was left over from when there were very
7099 few conditions). Also, it now recognizes conditions like "pam" even when
7100 the relevant support is not compiled in: a suitably worded error message is
7101 given if an attempt is made to use such a condition.
7103 11. Added ${time_interval:xxxxx}.
7105 12. A bug was causing one of the ddress fields not to be passed back correctly
7106 from remote delivery subprocesses. The field in question was not being
7107 subsequently used, so this caused to problems in practice.
7109 13. Added new log selectors queue_time and deliver_time.
7111 14. Might have fixed a bug in maildirsizefile handling that threw up
7112 "unexpected character" debug warnings, and recalculated the data
7113 unnecessarily. In any case, I expanded the warning message to give more
7116 15. Added the message "Restricted characters in address" to the statements in
7117 the default ACL that block characters like @ and % in local parts.
7119 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7120 Three changes have been made:
7122 (a) There was a serious bug; a negative response to MAIL caused the whole
7123 recipient domain to be cached as invalid, thereby blocking all messages
7124 to all local parts at the same domain, from all senders. This bug has
7125 been fixed. The domain is no longer cached after a negative response to
7126 MAIL if the sender used is not empty.
7128 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7131 (c) A new callout option, "use_sender" has been added for people who want
7132 the modified behaviour.
7138 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7141 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7142 indeed breaks things for older releases.
7144 3. Added additional logging to the case where there is a problem reading data
7145 from a filter that is running in a subprocess using a pipe, in order to
7146 try to track down a specific problem.
7148 4. Testing facility fudge: when running in the test harness and attempting
7149 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7150 getting "No route to host". Convert this to a timeout.
7152 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7155 6. Some OS don't have socklen_t but use size_t instead. This affects the
7156 fifth argument of getsockopt() amongst other things. This is now
7157 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7158 can be set for individual OS. I have set it for SunOS5, OSF1, and
7159 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7160 some earlier ones do not.
7162 7. Change 4.30/15 was not doing the test caselessly.
7164 8. The standard form for an IPv6 address literal was being rejected by address
7165 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7166 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7167 this, as well as the form without the "IPv6" on the front (but only when
7168 address literals are enabled, of course).
7170 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7172 10. Exim crashed if a message with an empty sender address specified by -f
7173 encountered a router with an errors_to setting. This could be provoked only
7174 by a command such as
7178 where an empty string was supplied; "<>" did not hit this bug.
7180 11. Installed PCRE release 4.5.
7182 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7183 remained set. It is now erased.
7185 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7186 times from message ids (which are base 36 rather than the normal 62).
7188 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7189 were being counted as actual protocol errors, and logged if the log
7190 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7191 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7192 and DATA following a set of rejected RCPTs do not count as protocol errors.
7193 In other words, Exim assumes they were pipelined, though this may not
7194 actually be the case. Of course, in all cases the client gets an
7195 appropriate error code.
7197 15. If a lookup fails in an ACL condition, a message about the failure may
7198 be available; it is used if testing the ACL cannot continue, because most
7199 such messages specify what the cause of the deferral is. However, some
7200 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7201 that caused an old message to be retained and used if a later statement
7202 caused a defer, replacing the real cause of the deferral.
7204 16. If an IP address had so many PTR records that the DNS lookup buffer
7205 was not large enough to hold them, Exim could crash while trying to process
7206 the truncated data. It now detects and logs this case.
7208 17. Further to 4.21/58, another change has been made: if (and only if) the
7209 first line of a message (the first header line) ends with CRLF, a bare LF
7210 in a subsequent header line has a space inserted after it, so as not to
7211 terminate the header.
7213 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7214 unnecessarily, used atoi() instead of strtol(), and didn't check the
7215 termination when getting file sizes from file names by regex.
7217 19. Completely re-implemented the support for maildirsize files, in the light
7218 of a number of problems with the previous contributed implementation
7219 (4.30/29). In particular:
7221 . If the quota is zero, the maildirsize file is maintained, but no quota is
7224 . If the maildir directory does not exist, it is created before any attempt
7225 to write a maildirsize file.
7227 . The quota value in the file is just a cache; if the quota is changed in
7228 the transport, the new value overrides.
7230 . A regular expression is available for excluding directories from the
7233 20. The autoreply transport checks the characters in options that define the
7234 message's headers; it allows continued headers, but it was checking with
7235 isspace() after an embedded newline instead of explicitly looking for a
7238 21. If all the "regular" hosts to which an address was routed had passed their
7239 expiry times, and had not reached their retry times, the address was
7240 bounced, even if fallback hosts were defined. Now Exim should go on to try
7243 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7244 equivalent code in the SMTP transport. Some hosts send humungous responses
7245 to HELO/EHLO, more than 1024 it seems.
7247 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7248 gives compiler warnings in some environments. I've now done it "properly",
7251 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7252 (because of problems with the built-in one) was declared to return uschar *
7253 instead of char *, causing compiler failure.
7255 25. Fixed a file descriptor leak when processing alias/forward files.
7257 26. Fixed a minor format string issue in dbfn.c.
7259 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7261 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7262 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7265 29. When a sender address is verified, it is cached, to save repeating the test
7266 when there is more than one recipient in a message. However, when the
7267 verification involves a callout, it is possible for different callout
7268 options to be set for different recipients. It is too complicated to keep
7269 track of this in the cache, so now Exim always runs a verification when a
7270 callout is required, relying on the callout cache for the optimization.
7271 The overhead is duplication of the address routing, but this should not be
7274 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7275 to be verified with callout=postmaster, and the main callout worked but the
7276 postmaster check failed, the verification correctly failed. However, if a
7277 subsequent RCPT command asked for sender verification *without* the
7278 postmaster check, incorrect caching caused this verification also to fail,
7281 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7282 it was not caching the DNS options (qualify_single, search_parents) that
7283 were used when the lookup failed. A subsequent lookup with different
7284 options therefore always gave the same answer, though there were cases
7285 where it should not have. (Example: a "domains = !$mx_any" option on a
7286 dnslookup router: the "domains" option is always processed without any
7287 widening, but the router might have qualify_single set.) Now Exim uses the
7288 cached value only when the same options are set.
7290 32. Added John Jetmore's "exipick" utility to the distribution.
7292 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7293 than a timeout (e.g. a certificate is required, and is not provided), an
7294 Exim server now closes the connection immediately. Previously it waited for
7295 the client to close - but if the client is SSL, it seems that they each
7296 wait for each other, leading to a delay before one of them times out.
7298 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7299 maintained 0.8.x compatibility because I don't think many are using it, and
7300 it is clearly obsolete.
7302 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7305 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7306 client certificate was expired. A simple patch fixes this, though I don't
7307 understand the full logic of why the verify callback is called multiple
7310 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7311 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7312 which causes problems with some clients (such as the Certicom SSL Plus
7313 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7314 disables the coutermeasure allowing Eudora to connect."
7316 38. Exim was not checking that a write() to a log file succeeded. This could
7317 lead to Bad Things if a log got too big, in particular if it hit a file
7318 size limit. Exim now panics and dies if it cannot write to a log file, just
7319 as it does if it cannot open a log file.
7321 39. Modified OS/Makefile-Linux so that it now contains
7323 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7325 The two -D definitions ensure that Exim is compiled with large file
7326 support, which makes it possible to handle log files that are bigger than
7329 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7330 instance) a domain was checked against a named list that involved a lookup,
7331 causing $domain_data to be set, then another domain was checked against the
7332 same list, then the first domain was re-checked, the value of $domain_data
7333 after the final check could be wrong. In particular, if the second check
7334 failed, it could be set empty. This bug probably also applied to
7337 41. The strip_trailing_dot option was not being applied to the address given
7338 with the -f command-line option.
7340 42. The code for reading a message's header from the spool was incrementing
7341 $received_count, but never initializing it. This meant that the value was
7342 incorrect (doubled) while delivering a message in the same process in which
7343 it was received. In the most common configuration of Exim, this never
7344 happens - a fresh exec is done - but it can happen when
7345 deliver_drop_privilege is set.
7347 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7348 it now includes up to 150 characters of the unexpected data in the log
7351 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7352 and building data strings. The size of both of these buffers was 10 000
7353 bytes - far larger than anybody would *ever* want, thought I. Needless to
7354 say, somebody hit the limit. I have increased the maximum line length to
7355 20 000 and the maximum data length of concatenated lines to 100 000. I have
7356 also fixed two bugs, because there was no checking on these buffers. Tsk,
7357 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7358 buffer is too small.
7360 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7361 lsearch lookups. Now it does.
7363 46. When parsing a route_list item in a manualroute router, a fixed-length
7364 buffer was used for the list of hosts. I made this 1024 bytes long,
7365 thinking that nobody would ever have a list of hosts that long. Wrong.
7366 Somebody had a whole pile of complicated expansion conditions, and the
7367 string was silently truncated, leading to an expansion error. It turns out
7368 that it is easier to change to an unlimited length (owing to other changes
7369 that have happened since this code was originally written) than to build
7370 structure for giving a limitation error. The length of the item that
7371 expands into the list of hosts is now unlimited.
7373 47. The lsearch lookup could not handle data where the length of text line was
7374 more than 4095 characters. Such lines were truncated, leading to shortened
7375 data being returned. It should now handle lines of any length.
7377 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7378 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7381 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7382 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7383 no longer happens in comment lines. A list of replacements is now placed
7384 at the head of all of the source files, except those whose only change is
7385 to replace PERL_COMMAND in the very first #! line.
7387 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7388 messages on the queue, with a bottom-up merge sort, using code contributed
7389 by Michael Haardt. This should make operations like -bp somewhat faster on
7390 large queues. It won't affect queue runners, except when queue_run_in_order
7393 51. Installed eximstats 1.31 in the distribution.
7395 52. Added support for SRV lookups to the dnslookup router.
7397 53. If an ACL referred to $message_body or $message_body_end, the value was not
7398 reset for any messages that followed in the same SMTP session.
7400 54. The store-handling optimization for building very long strings was not
7401 differentiating between the different store pools. I don't think this
7402 actually made any difference in practice, but I've tidied it.
7404 55. While running the routers to verify a sender address, $sender_address
7405 was still set to the sender address. This is wrong, because when routing to
7406 send a bounce to the sender, it would be empty. Therefore, I have changed
7407 it so that, while verifying a sender address, $sender_address is set to <>.
7408 (There is no change to what happens when verifying a recipient address.)
7410 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7411 target A or AAAA records (if not already returned) without resetting the
7412 qualify_single or search_parents options of the DNS resolver. These are
7413 inappropriate in this case because the targets of MX and SRV records must
7414 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7415 target that, when qualified, matched something in the local domain. These
7416 two options are now turned off when doing these lookups.
7418 57. It seems that at least some releases of Reiserfs (which does not have the
7419 concept of a fixed number of inodes) returns zero and not -1 for the
7420 number of available inodes. This interacted badly with check_spool_inodes,
7421 which assumed that -1 was the "no such thing" setting. What I have done is
7422 to check that the total number of inodes is greater than zero before doing
7423 the test of how many are available.
7425 58. When a "warn" ACL statement has a log_message modifier, the message is
7426 remembered, and not repeated. This is to avoid a lot of repetition when a
7427 message has many recipients that cause the same warning to be written.
7428 However, Exim was preserving the list of already written lines for an
7429 entire SMTP session, which doesn't seem right. The memory is now reset if a
7430 new message is started.
7432 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7433 debugging output unless log_rewrite was also set.
7435 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7436 of (int)(handle) when we know that handle contains (void *)(-1).
7438 61. The Exim daemon panic-logs an error return when it closes the incoming
7439 connection. However "connection reset by peer" seems to be common, and
7440 isn't really an error worthy of noting specially, so that particular error
7443 62. When Exim is trying to find all the local interfaces, it used to panic and
7444 die if the ioctl to get the interface flags failed. However, it seems that
7445 on at least one OS (Solaris 9) it is possible to have an interface that is
7446 included in the list of interfaces, but for which you get a failure error
7447 for this call. This happens when the interface is not "plumbed" into a
7448 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7449 failure of the "get flags" call assumes that the interface is down.
7451 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7452 makes life easier for people who are doing arithmetic on fields extracted
7453 from dates, where you often get leading zeros that should not be
7454 interpreted as octal.
7456 64. Added qualify_domain to the redirect router, to override the global
7459 65. If a pathologically long header line contained very many addresses (the
7460 report of this problem mentioned 10 000) and each of them was rewritten,
7461 Exim could use up a very large amount of memory. (It kept on making new
7462 copies of the header line as it rewrote, and never released the old ones.)
7463 At the expense of a bit more processing, the header rewriting function has
7464 been changed so that it no longer eats memory in this way.
7466 66. The generation of the Received: header has been moved from the time that a
7467 message starts to be received, to the time that it finishes. The timestamp
7468 in the Received: header should now be very close to that of the <= log
7469 line. There are two side-effects of this change:
7471 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7472 logged header lines no longer include the local Received: line, because
7473 it has not yet been created. The same applies to a copy of the message
7474 that is returned to a non-SMTP sender when a message is rejected.
7476 (b) When a filter file is tested using -bf, no additional Received: header
7477 is added to the test message. After some thought, I decided that this
7480 This change does not affect the value of $received_for. It is still set
7481 after address rewriting, but before local_scan() is called.
7483 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7485 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7486 gave an unhelpful panic error message, and a defer error. I have managed to
7487 change this behaviour so that it now rejects any supplied certificate,
7488 which seems right, as the list of acceptable certificates is empty.
7490 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7491 gave an unhelpful defer error. I have not managed to make this reject any
7492 supplied certificates, but the error message it gives is "no certificate
7493 supplied", which is not helpful.
7495 70. exigrep's output now also includes lines that are not associated with any
7496 message, but which match the given pattern. Implemented by a patch from
7497 Martin Sluka, which also tidied up the Perl a bit.
7499 71. Recipient callout verification, like sender verification, was using <> in
7500 the MAIL FROM command. This isn't really the right thing, since the actual
7501 sender may affect whether the remote host accepts the recipient or not. I
7502 have changed it to use the actual sender in the callout; this means that
7503 the cache record is now keyed on a recipient/sender pair, not just the
7504 recipient address. There doesn't seem to be a real danger of callout loops,
7505 since a callout by the remote host to check the sender would use <>.
7506 [SEE ABOVE: changed after hitting problems.]
7508 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7509 temporary errors. However, in the case of such a code being given after
7510 the end of a data transmission (i.e. after ".") Exim was failing to write
7511 a retry record for the message. (Yes, there was some broken host that was
7512 actually sending 8xx at this point.)
7514 73. An unknown lookup type in a host list could cause Exim to panic-die when
7515 the list was checked. (An example that provoked this was putting <; in the
7516 middle of a list instead of at the start.) If this happened during a DATA
7517 ACL check, a -D file could be left lying around. This kind of configuration
7518 error no longer causes Exim to die; instead it causes a defer error. The
7519 incident is still logged to the main and panic logs.
7521 74. Buglet left over from Exim 3 conversion. The message "too many messages
7522 in one connection" was written to the rejectlog but not the mainlog, except
7523 when address rewriting (yes!) was being logged.
7525 75. Added write_rejectlog option.
7527 76. When a system filter was run not as root (that is, when system_filter_user
7528 was set), the values of the $n variables were not being returned to the
7529 main process; thus, they were not subsequently available in the $sn
7532 77. Added +return_path_on_delivery log selector.
7534 78. A connection timeout was being treated differently from recipients deferred
7535 when testing hosts_max_try with a message that was older than the host's
7536 retry timeout. (The host should not be counted, thus allowing all hosts to
7537 be tried at least once before bouncing.) This may have been the cause of an
7538 occasionally reported bug whereby a message would remain on the queue
7539 longer than the retry timeout, but would be bounced if a delivery was
7540 forced. I say "may" because I never totally pinned down the problem;
7541 setting up timeout/retry tests is difficult. See also the next item.
7543 79. The ultimate address timeout was not being applied to errors that involved
7544 a combination of host plus message (for example, a timeout on a MAIL
7545 command). When an address resolved to a number of possible hosts, and they
7546 were not all tried for each delivery (e.g. because of hosts_max_try), a
7547 message could remain on the queue longer than the retry timeout.
7549 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7552 81. Fixed an obscure SMTP outgoing bug which required at least the following
7553 conditions: (a) there was another message waiting for the same server;
7554 (b) the server returned 5xx to all RCPT commands in the first message so
7555 that the message was not completed; (c) the server dropped the connection
7556 or gave a negative response to the RSET that Exim sends to abort the
7557 transaction. The observed case was a dropped connection after DATA that had
7558 been sent in pipelining mode. That is, the server had advertised PIPELINING
7559 but was not implementing it correctly. The effect of the bug was incorrect
7560 behaviour, such as trying another host, and this could lead to a crash.
7566 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7567 and daemon.c were passed as pointers to ints; they should have been
7568 pointers to socklen_t variables (which are typically unsigned ints).
7570 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7573 3. Fixed a really odd bug that affected only the testing scheme; patching a
7574 certain fixed string in the binary changed the value of another string that
7575 happened to be identical to the end of the original first string.
7577 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7578 name", it returns that address as the IP address. On some operating
7579 systems (e.g. Solaris), it also passes back the IP address string as the
7580 "host name". However, on others (e.g. Linux), it passes back an empty
7581 string. Exim wasn't checking for this, and was changing the host name to an
7582 empty string, assuming it had been canonicalized.
7584 5. Although rare, it is permitted to have more than one PTR record for a given
7585 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7586 all the names associated with an address, because they do in Solaris.
7587 However, it seems that they do not in Linux for data that comes from the
7588 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7589 I found this out when I moved to a new Linux workstation and tried to run
7590 the Exim test suite.
7592 To get round this problem I have changed the code so that it now does its
7593 own call to the DNS to look up PTR records when searching for a host name.
7594 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7595 addresses that are only in /etc/hosts are still found.
7597 This behaviour is, however, controlled by an option called host_lookup_
7598 order, which defaults to "bydns:byaddr". If people want to use the other
7599 order, or indeed, just use one or the other means of lookup, they can
7600 specify it in this variable.
7602 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7603 some operating systems, this comes back from gethostbyaddr() as an empty
7604 string, and this is what Exim used to test for. However, it seems that in
7605 other systems, "." is yielded. Exim now tests for this case too.
7607 7. The values of check_spool_space and check_log_space are now held internally
7608 as a number of kilobytes instead of an absolute number of bytes. If a
7609 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7610 kilobyte. This means that much larger values can be stored.
7612 8. Exim monitor: an attempt to get the action menu when not actually pointing
7613 at a message produces an empty menu entitled "No message selected". This
7614 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7615 no entries in it ("Shell widget menu has zero width and/or height"). So I
7616 have added a single, blank menu entry in this case.
7618 9. Added ${quote_local_part.
7620 10. MIME decoding is now applied to the contents of Subject: header lines when
7623 11. Now that a reference to $sender_host_address automatically causes a reverse
7624 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7625 host lookup before query-style lookups in lists that might use this
7626 variable. This has therefore been abolished, and the "net-" prefix is no
7627 longer necessary for query-style lookups.
7629 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7630 have been a typo for LFLAGS, so it has been changed.
7632 13. The install script calls Exim with "-C /dev/null" in order to find the
7633 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7634 to be output. However, since Exim outputs its version number before the
7635 error, it didn't break the script. It just looked ugly. I fixed this by
7636 always allowing "-C /dev/null" if the caller is root.
7638 14. Ignore overlarge ACL variable number when reading spool file - insurance
7639 against a later release with more variables having written the file.
7641 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7642 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7643 this, as well as the form without the "IPv6" on the front.
7645 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7646 OS/Makefile-Darwin file.
7648 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7649 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7651 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7652 variable. However, some OS use ints for inodes. I've added cast to long int
7653 to get rid of the compiler warning.
7655 19. I had forgotten to lock out "/../" in configuration file names when
7656 ALT_CONFIG_PREFIX was set.
7658 20. Routers used for verification do not need to specify transports. However,
7659 if such a router generated a host list, and callout was configured, Exim
7660 crashed, because it could not find a port number from the (non-existent)
7661 transport. It now assumes port 25 in this circumstance.
7663 21. Added the -t option to exigrep.
7665 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7666 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7667 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7668 them is set, LOOKUP_LSEARCH is forced.
7670 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7672 24. Added sender and host information to the "rejected by local_scan()" log
7673 line; previously there was no indication of these.
7675 25. Added .include_if_exists.
7677 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7678 Linux. It turns out that not all file systems support this. Apparently some
7679 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7680 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7681 not supported on the file descriptor, is now ignored when Exim is trying to
7682 sync a directory. This applies only to Linux.
7684 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7686 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7687 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7690 29. Added support for maildirsize files from supplied patch (modified a bit).
7692 30. The use of :fail: followed by an empty string could lead Exim to respond to
7693 sender verification failures with (e.g.):
7695 550 Verification failed for <xxx>
7696 550 Sender verify failed
7698 where the first response line was missing the '-' that indicates it is not
7699 the final line of the response.
7701 31. The loop for finding the name of the user that called Exim had a hardwired
7702 limit of 10; it now uses the value of finduser_retries, which is used for
7703 all other user lookups.
7705 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7708 33. Exim was neglecting to zero errno before one call of strtol() when
7709 expanding a string and expecting an integer value. On some systems this
7710 resulted in spurious "integer overflow" errors. Also, it was casting the
7711 result into an int without checking.
7713 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7714 did not work. The code looks as if it has *never* worked, though it appears
7715 to have been documented since at least release 1.62. I have made it work.
7717 35. The "timeout_DNS" error in retry rules, also documented since at least
7718 1.62, also never worked. As it isn't clear exactly what this means, and
7719 clearly it isn't a major issue, I have abolished the feature by treating it
7720 as "timeout", and writing a warning to the main and panic logs.
7722 36. The display of retry rules for -brt wasn't always showing the error code
7725 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7726 timeout_connect_A, timeout_connect_MX.
7728 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7729 to the empty sender.
7731 39. The daemon was not analysing the content of -oX till after it had closed
7732 stderr and disconnected from the controlling terminal. This meant that any
7733 syntax errors were only noted on the panic log, and the return code from
7734 the command was 0. By re-arranging the code a little, I've made the
7735 decoding happen first, so such errors now appear on stderr, and the return
7736 code is 1. However, the actual setting up of the sockets still happens in
7737 the disconnected process, so errors there are still only recorded on the
7740 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7741 connections (as happens on some IP stacks) was logged at start up time as
7742 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7743 it from "IPv6 and IPv4", which means that two separate sockets are being
7746 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7747 says whether AF_INET or AF_INET6 was passed as an argument.
7749 42. Exiwhat output was messed up when time zones were included in log
7752 43. Exiwhat now gives more information about the daemon's listening ports,
7753 and whether -tls-on-connect was used.
7755 44. The "port" option of the smtp transport is now expanded.
7757 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7758 silently ignored. Now an error message is written to the main and panic
7761 46. There's a new ACL modifier called "logwrite" which writes to a log file
7762 as soon as it is encountered.
7764 47. Added $local_user_uid and $local_user_gid at routing time.
7766 48. Exim crashed when trying to verify a sender address that was being
7769 49. Exim was recognizing only a space character after ".include". It now also
7770 recognizes a tab character.
7772 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7773 extracting the relevant information from the specification. The man page no
7774 longer contains scrambled data for the -d option, and I've added a section
7775 at the front about calling Exim under different names.
7777 51. Added "extra_headers" argument to the "mail" command in filter files.
7779 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7782 53. Installed eximstats 1.29.
7784 54. Added transport_filter_timeout as a generic transport option.
7786 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7787 Cc: header lines. This was required by RFC 822, but it not required by RFC
7790 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7791 incoming messages that did not have them. Now it does so only if the
7792 message originates locally, that is, if there is no associated remote host
7793 address. When Resent- header lines are present, this applies to the Resent-
7794 lines rather than the non-Resent- lines.
7796 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7797 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7799 58. Messages for configuration errors now include the name of the main
7800 configuration file - useful now that there may be more than one file in a
7801 list (.included file names were always shown).
7803 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7804 for those rare installations that do not start the daemon as root or run it
7805 setuid root. I've cut out the call to initgroups() if the daemon is not
7808 60. The Exim user and group can now be bound into the binary as text strings
7809 that are looked up at the start of Exim's processing.
7811 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7813 62. Added $mailstore_basename variable.
7815 63. Installed patch to sieve.c from Michael Haardt.
7817 64. When Exim failed to open the panic log after failing to open the main log,
7818 the original message it was trying to log was written to stderr and debug
7819 output, but if they were not available (the usual case in production), it
7820 was lost. Now it is written to syslog before the two lines that record the
7821 failures to open the logs.
7823 65. Users' Exim filters run in subprocesses under the user's uid. It is
7824 possible for a "deliver" command or an alias in a "personal" command to
7825 provoke an address rewrite. If logging of address rewriting is configured,
7826 this fails because the process is not running as root or exim. There may be
7827 a better way of dealing with this, but for the moment (because 4.30 needs
7828 to be released), I have disabled address rewrite logging when running a
7829 filter in a non-root, non-exim process.
7835 1. The buildconfig auxiliary program wasn't quoting the value set for
7836 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7837 not defined. This bug was masked in 4.22 by the effect that was fixed in
7840 2. Some messages that were rejected after a message id was allocated were
7841 shown as "incomplete" by exigrep. It no longer does this for messages that
7842 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7844 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7845 have allow_domain_literals set, the ID did not get logged in the <= line.
7846 Domain literals are now always recognized in Message-ID: header lines.
7848 4. The first argument for a ${extract expansion item is the key name or field
7849 number. Leading and trailing spaces in this item were not being ignored,
7850 causing some misleading effects.
7852 5. When deliver_drop_privilege was set, single queue runner processes started
7853 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7854 same command in the process it spins off) were not dropping privilege.
7856 6. When the daemon running as "exim" started a queue runner, it always
7857 re-executed Exim in the spun-off process. This is a waste of effort when
7858 deliver_drop_privilege is set. The new process now just calls the
7859 queue-runner function directly.
7865 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7868 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7869 ignored. Though the use of -oP was forcing the writing of a pid file, it
7870 was always written to the default place.
7872 3. If the message "no IP address found for host xxxx" is generated during
7873 incoming verification, it is now followed by identification of the incoming
7874 connection (so you can more easily find what provoked it).
7876 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7878 5. Added some features to "harden" Exim a bit more against certain attacks:
7880 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7881 be put in Local/Makefile. This is like the never_users runtime option,
7882 but it cannot be overridden. The default setting is "root".
7884 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7885 prefix string with which any file named in a -C command line option
7888 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7889 is retained for -C and -D only if the caller of Exim is root. Without
7890 it, the exim user may also use -C and -D and retain privilege.
7892 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7893 command line option is disabled.
7895 6. Macro names set by the -D option must start with an upper case letter, just
7896 like macro names defined in the configuration file.
7898 7. Added "dereference=" facility to LDAP.
7900 8. Two instances of the typo "uknown" in the source files are fixed.
7902 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7903 the Configure-Makefile script screwed up while processing it.
7905 10. Incorporated PCRE 4.4.
7907 11. The SMTP synchronization check was not operating right at the start of an
7908 SMTP session. For example, it could not catch a HELO sent before the client
7909 waited for the greeting. There is now a check for outstanding input at the
7910 point when the greeting is written. Because of the duplex, asynchronous
7911 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7912 way, but not yet received, when the check is performed.
7914 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7915 on TCP/IP sockets, because this apparently causes some broken clients to
7918 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7919 unchanged) from the Cygwin maintainer.
7921 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7922 was supported instead of testing for mbx. Effectively a typo.
7924 15. The spa authenticator server code was not checking that the input it
7925 received was valid base64.
7927 16. The debug output line for the "set" modifier in ACLs was not showing the
7928 name of the variable that was being set.
7930 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7932 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7933 lookup on its own. Something else was needed to trigger the lookup. For
7934 example, a match in host_lookup or the need for a host name in a host list.
7935 Now, if $sender_host_name is referenced and the host name has not yet been
7936 looked up, a lookup is performed. If the lookup fails, the variable remains
7937 empty, and $host_lookup_failed is set to "1".
7939 19. Added "eqi" as a case-independent comparison operator.
7941 20. The saslauthd authentication condition could segfault if neither service
7942 nor realm was specified.
7944 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7945 error message that was logged was misleading, and incoming SMTP
7946 connections were dropped. The message is now more accurate, and temporary
7947 errors are given to SMTP connections.
7949 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7950 (except RSET) with a 421 error, until QUIT is received. However, it was
7951 failing to send a response to QUIT.
7953 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7954 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7955 ACL is now run after the helo_try_verify_hosts code.
7957 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7958 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7959 case-independent, so other case variants are also recognized). Apparently
7960 some systems use these upper case variants.
7962 25. If more than two messages were waiting for the same host, and a transport
7963 filter was specified for the transport, Exim sent two messages over the
7964 same TCP/IP connection, and then failed with "socket operation on non-
7965 socket" when it tried to send the third.
7967 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7969 27. The extern definition of crypt16() in expand.c was not being excluded when
7970 the OS had its own crypt16() function.
7972 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7973 as a preferred synonym for return_size_limit, both as an option and as an
7976 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7978 30. Changed the default configuration ACL to relax the local part checking rule
7979 for addresses that are not in any local domains. For these addresses,
7980 slashes and pipe symbols are allowed within local parts, but the sequence
7981 /../ is explicitly forbidden.
7983 31. SPA server authentication was not clearing the challenge buffer before
7986 32. log_message in a "warn" ACL statement was writing to the reject log as
7987 well as to the main log, which contradicts the documentation and doesn't
7988 seem right (because no rejection is happening). So I have stopped it.
7990 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7991 However, I am unable to do any testing of this.
7993 34. Fixed an infelicity in the appendfile transport. When checking directories
7994 for a mailbox, to see if any needed to be created, it was accidentally
7995 using path names with one or more superfluous leading slashes; tracing
7996 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7998 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7999 contents are added to the log line that is written for every discarded
8000 recipient. (Previously a log_message setting was ignored.)
8002 36. The ${quote: operator now quotes the string if it is empty.
8004 37. The install script runs exim in order to find its version number. If for
8005 some reason other than non-existence or emptiness, which it checks, it
8006 could not run './exim', it was installing it with an empty version number,
8007 i.e. as "exim-". This error state is now caught, and the installation is
8010 38. An argument was missing from the function that creates an error message
8011 when Exim fails to connect to the socket for saslauthd authentication.
8012 This could cause Exim to crash, or give a corrupted message.
8014 39. Added isip, isip4, and isip6 to ${if conditions.
8016 40. The ACL variables $acl_xx are now saved with the message, and can be
8017 accessed later in routers, transports, and filters.
8019 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8020 strings in the file are not string-expanded.
8022 42. If a MAIL command specified a SIZE value that was too large to fit into an
8023 int variable, the check against message_size_limit failed. Such values are
8024 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8025 day this will have to be increased, but I don't think I want to be around
8026 when emails are that large.
8033 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8034 iconv() is not standard in FreeBSD.
8036 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8037 IPv6 enabled. The observed symptom was a segmentation fault on return from
8038 the function os_common_find_running_interfaces() in src/os.c.
8040 3. In the check_special_case() function in daemon.c I had used "errno" as an
8041 argument name, which causes warnings on some systems. This was basically a
8042 typo, since it was named "eno" in the comments!
8044 4. The code that waits for the clock to tick (at a resolution of some fraction
8045 of a second) so as to ensure message-id uniqueness was always waiting for
8046 at least one whole tick, when it could have waited for less. [This is
8047 almost certainly not relevant at current processor speeds, where it is
8048 unlikely to ever wait at all. But we try to future-proof.]
8050 5. The function that sleeps for a time interval that includes fractions of a
8051 second contained a race. It did not block SIGALRM between setting the
8052 timer, and suspending (a couple of lines later). If the interval was short
8053 and the sigsuspend() was delayed until after it had expired, the suspension
8054 never ended. On busy systems this could lead to processes getting stuck for
8057 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8058 process, before it forks any delivery processes. The open lookup caching
8059 mechanism meant that the open file or database connection was passed into
8060 the delivery process. The problem was that delivery processes always tidy
8061 up cached lookup data. This could cause a problem for the next delivery
8062 process started by the queue runner, because the external queue runner
8063 process does not know about the closure. So the next delivery process
8064 still has data in the lookup cache. In the case of a file lookup, there was
8065 no problem because closing a file descriptor in a subprocess doesn't affect
8066 the parent. However, if the lookup was caching a connection to a database,
8067 the connection was closed, and the second delivery process was likely to
8068 see errors such as "PGSQL: query failed: server closed the connection
8069 unexpectedly". The problem has been fixed by closing all cached lookups
8070 in a queue runner before running a delivery process.
8072 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8073 seem to have the "const" qualifier which it has on other OS. I've
8076 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8077 *of the same type* (client or server) with the same public name that an
8078 error should be diagnosed.
8080 9. When Exim looked up a host name for an IP address, but failed to find the
8081 original IP address when looking up the host name (a safety check), it
8082 output the message "<ip address> does not match any IP for NULL", which was
8083 confusing, to say the least. The bug was that the host name should have
8084 appeared instead of "NULL".
8086 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8087 user that is built into the binary, and the -C or -D options is used, root
8088 privilege is dropped before the configuration file is read. In addition,
8089 logging is switched to stderr instead of the normal log files. If the
8090 configuration then re-defines the Exim user, the unprivileged environment
8091 is probably not what is expected, so Exim logs a panic warning message (but
8094 However, if deliver_drop_privilege is set, the unprivileged state may well
8095 be exactly what is intended, so the warning has been cut out in that case,
8096 and Exim is allowed to try to write to its normal log files.
8102 1. smtp_return_error_details was not giving details for temporary sender
8103 or receiver verification errors.
8105 2. Diagnose a configuration error if two authenticators have the same public
8108 3. Exim used not to create the message log file for a message until the first
8109 delivery attempt. This could be confusing when incoming messages were held
8110 for policy or load reasons. The message log file is now created at the time
8111 the message is received, and an initial "Received" line is written to it.
8113 4. The automatically generated man page for command line options had a minor
8114 bug that caused no ill effects; however, a more serious problem was that
8115 the procedure for building the man page automatically didn't always
8116 operate. Consequently, release 4.20 contains an out-of-date version. This
8117 shouldn't happen again.
8119 5. When building Exim with embedded Perl support, the script that builds the
8120 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8121 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8123 6. The freeze_tell option was not being used for messages that were frozen on
8124 arrival, either by an ACL or by local_scan().
8126 7. Added the smtp_incomplete_transaction log selector.
8128 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8129 was accepting AUTH without a new EHLO.
8131 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8132 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8135 10. Logging of TCP/IP connections (when configured) now happens in the main
8136 daemon process instead of the child process, so that the TCP/IP connection
8137 count is more accurate (but it can never be perfect).
8139 11. The use of "drop" in a nested ACL was not being handled correctly in the
8140 outer ACL. Now, if condition failure induced by the nested "drop" causes
8141 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8142 or "require"), the connection is dropped.
8144 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8145 that yield "discard" can now be used with an "accept" or a "discard" verb,
8146 but an error is generated for any others (because I can't see a useful way
8147 to define what should happen).
8149 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8150 are now processed in the same way as lines in the Exim configuration file.
8151 In particular, continuation lines are supported.
8153 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8155 15. Added -ti meaning -t -i.
8157 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8158 domains, and warn by logging if others are found.
8160 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8161 returned from GIFCONF when Exim is trying to find the list of interfaces on
8162 a host. The code in os.c has been modified to copy each ifreq to an aligned
8163 structure in all cases.
8165 Also, in some cases, the returned ifreq's were being copied to a 'struct
8166 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8167 means the last couple of bytes of an IPv6 address could be chopped if the
8168 ifreq contained only a normal sockaddr (14 bytes storage).
8170 18. Named domain lists were not supported in the hosts_treat_as_local option.
8171 An entry such as +xxxx was not recognized, and was treated as a literal
8174 19. Ensure that header lines added by a DATA ACL are included in the reject log
8175 if the ACL subsequently rejects the message.
8177 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8178 MD5 (which is deprecated).
8180 21. When testing a filter file using -bf, Exim was writing a message when it
8181 took the sender from a "From " line in the message, but it was not doing so
8182 when it took $return_path from a Return-Path: header line. It now does.
8184 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8185 with a valid header line field name (a series of printing characters
8186 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8188 23. Changed "disc" in the source to "disk" to conform to the documentation and
8189 the book and for uniformity.
8191 24. Ignore Sendmail's -Ooption=value command line item.
8193 25. When execve() failed while trying to run a command in a pipe transport,
8194 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8195 could be confused with a return value of 69 from the command itself. This
8196 has been changed to 127, the value the shell returns if it is asked to run
8197 a non-existent command. The wording for the related log line suggests a
8198 non-existent command as the problem.
8200 26. If received_header_text expands to an empty string, do not add a Received:
8201 header line to the message. (Well, it adds a token one on the spool, but
8202 marks it "old" so that it doesn't get used or transmitted.)
8204 27. Installed eximstats 1.28 (addition of -nt option).
8206 28. There was no check for failure on the call to getsockname() in the daemon
8207 code. This can fail if there is a shortage of resources on the system, with
8208 ENOMEM, for example. A temporary error is now given on failure.
8210 29. Contrary to the C standard, it seems that in some environments, the
8211 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8212 program. Exim now does this explicitly; it affects the formatting of
8213 timestamps using strftime().
8215 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8216 complaints. I've now initialized all the variables, to avoid this.
8218 32. Header lines added by a system filter were not being "seen" during
8219 transport-time rewrites.
8221 33. The info_callback() function passed to OpenSSL is set up with type void
8222 (*)(SSL *, int, int), as described somewhere. However, when calling the
8223 function (actually a macro) that sets it up, the type void(*)() is
8224 expected. I've put in a cast to prevent warnings from picky compilers.
8226 34. If a DNS black list lookup found a CNAME record, but there were no A
8227 records associated with the domain it pointed at, Exim crashed.
8229 35. If a DNS black list lookup returned more than one A record, Exim ignored
8230 all but the first. It now scans all returned addresses if a particular IP
8231 value is being sought. In this situation, the contents of the
8232 $dnslist_value variable are a list of all the addresses, separated by a
8235 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8236 to accept a host name and all its aliases if the forward lookup for any of
8237 them yielded the IP address of the incoming connection. Now it accepts only
8238 those names whose forward lookup yields the correct IP address. Any other
8239 names are discarded. This closes a loophole whereby a rogue DNS
8240 administrator could create reverse DNS records to break through a
8241 wildcarded host restriction in an ACL.
8243 37. If a user filter or a system filter that ran in a subprocess used any of
8244 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8245 the wrong values were passed to the pipe command ($thisaddress had the
8246 value of $0, $0 had the value of $1, etc). This bug was introduced by
8247 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8249 38. Improved the line breaking for long SMTP error messages from ACLs.
8250 Previously, if there was no break point between 40 and 75 characters, Exim
8251 left the rest of the message alone. Two changes have been made: (a) I've
8252 reduced the minimum length to 35 characters; (b) if it can't find a break
8253 point between 35 and 75 characters, it looks ahead and uses the first one
8254 that it finds. This may give the occasional overlong line, but at least the
8255 remaining text gets split now.
8257 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8258 file descriptors might be low, and that setting 1000 would always raise it.
8259 It turns out that in some environments, the limit is already over 1000 and
8260 that lowering it causes trouble. So now Exim takes care not to decrease it.
8262 40. When delivering a message, the value of $return_path is set to $sender_
8263 address at the start of routing (routers may change the value). By an
8264 oversight, this default was not being set up when an address was tested by
8265 -bt or -bv, which affected the outcome if any router or filter referred to
8268 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8269 log only when "message" or "log_message" are set. However, if one of the
8270 conditions was an address verification, or a call to a nested ACL, the
8271 messages generated by the underlying test were being passed through. This
8272 no longer happens. The underlying message is available in $acl_verify_
8273 message for both "message" and "log_message" expansions, so it can be
8274 passed through if needed.
8276 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8277 new expansion $bh_ to give the encoded byte string without charset
8278 translation. Translation happens only if iconv() is available; HAVE_ICONV
8279 indicates this at build time. HEADERS_CHARSET gives the charset to
8280 translate to; headers_charset can change it in the configuration, and
8281 "headers charset" can change it in an individual filter file.
8283 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8284 that creates RFC 2047 encoded "words" labels them as that charset instead
8285 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8286 expansion operator; (ii) when Exim creates a From: line for a local
8287 message; (iii) when a header line is rewritten to include a "phrase" part.
8289 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8290 buggy, causing it to skip the first lines of messages whose message ID
8291 ended in 'D'. This would not have bitten before Exim release 4.14, because
8292 message IDs were unlikely to end in 'D' before then. The effect was to have
8293 incorrect size information for certain domains.
8295 45. #include "config.h" was missing at the start of the crypt16.c module. This
8296 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8299 46. If there was a timeout during a "random" callout check, Exim treated it as
8300 a failure of the random address, and carried on sending RSET and the real
8301 address. If the delay was just some slowness somewhere, the response to the
8302 original RCPT would be taken as a response to RSET and so on, causing
8303 mayhem of various kinds.
8305 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8306 when I implemented it. It didn't allow for the fact that some option values
8307 may legitimately be negative (e.g. size_addition), and it didn't even do
8308 the right test for positive values.
8310 48. Domain names in DNS records are case-independent. Exim always looks them up
8311 in lower case. Some resolvers return domain names in exactly the case they
8312 appear in the zone file, that is, they may contain uppercase letters. Not
8313 all resolvers do this - some return always lower case. Exim was treating a
8314 change of case by a resolver as a change of domain, similar to a widening
8315 of a domain abbreviation. This triggered its re-routing code and so it was
8316 trying to route what was effectively the same domain again. This normally
8317 caused routing to fail (because the router wouldn't handle the domain
8318 twice). Now Exim checks for this case specially, and just changes the
8319 casing of the domain that it ultimately uses when it transmits the message
8322 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8325 50. If a filter generated a file delivery with a non-absolute name (possible if
8326 no home directory exists for the router), the forbid_file option was not
8329 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8330 the existing equality matching.
8332 52. Exim was using ints instead of ino_t variables in some places where it was
8333 dealing with inode numbers.
8335 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8336 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8337 called TMPDIR, and if it finds it is different, it changes its value.
8339 54. The smtp_printf() function is now made available to local_scan() so
8340 additional output lines can be written before returning. There is also an
8341 smtp_fflush() function to enable the detection of a dropped connection.
8342 The variables smtp_input and smtp_batched_input are exported to
8345 55. Changed the default runtime configuration: the message "Unknown user"
8346 has been removed from the ACL, and instead placed on the localuser router,
8347 using the cannot_route_message feature. This means that any verification
8348 failures that generate their own messages won't get overridden. Similarly,
8349 the "Unrouteable address" message that was in the ACL for unverifiable
8350 relay addresses has also been removed.
8352 56. Added hosts_avoid_esmtp to the smtp transport.
8354 57. The exicyclog script was not checking for the esoteric option
8355 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8356 will work only if exicyclog is run under the appropriate euid.
8358 58. Following a discussion on the list, the rules by which Exim recognises line
8359 endings on incoming messages have been changed. The -dropcr and drop_cr
8360 options are now no-ops, retained only for backwards compatibility. The
8361 following line terminators are recognized: LF CRLF CR. However, special
8362 processing applies to CR:
8364 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8365 nor a local message in the state where . is a terminator.
8367 (ii) If a bare CR is encountered in a header line, an extra space is added
8368 after the line terminator so as not to end the header. The reasoning
8369 behind this is that bare CRs in header lines are most likely either
8370 to be mistakes, or people trying to play silly games.
8372 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8373 was being incorrectly given as 18 bytes larger than it should have been.
8374 This is a VOB (very old bug).
8376 60. This may never have affected anything current, but just in case it has:
8377 When the local host is found other than at the start of a list of hosts,
8378 the local host, those with the same MX, and any that follow, are discarded.
8379 When the list in question was part of a longer list of hosts, the following
8380 hosts (not currently being processed) were also being discarded. This no
8381 longer happens. I'm not sure if this situation could ever has previously
8384 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8387 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8388 header. This is something that is recommended in a new Internet Draft, and
8389 is something that is documented as being done by Sendmail. There are two
8390 possible values. For messages generated by the autoreply transport, Exim
8393 Auto-Submitted: auto-replied
8395 whereas for all other generated messages (e.g. bounces) it adds
8397 Auto-Submitted: auto-generated
8399 63. The "personal" condition in filters now includes a test for the
8400 Auto-Submitted: header. If it contains the string "auto-" the message it
8401 not considered personal.
8403 64. Added rcpt_include_affixes as a generic transport option.
8405 65. Added queue_only_override (default true).
8407 66. Added the syslog_duplication option.
8409 67. If what should have been the first header line of a message consisted of
8410 a space followed by a colon, Exim was mis-interpreting it as a header line.
8411 It isn't of course - it is syntactically invalid and should therefore be
8412 treated as the start of the message body. The misbehaviour could have
8413 caused a number of strange effects, including loss of data in subsequent
8414 header lines, and spool format errors.
8416 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8417 client host had authenticated. This control can now be exercised by an ACL
8418 for more flexibility.
8420 69. By default, callouts do not happen when testing with -bh. There is now a
8421 variant, -bhc, which does actually run the callout code, including
8422 consulting and updating the callout cache.
8424 70. Added support for saslauthd authentication, courtesy of Alexander
8427 71. If statvfs() failed on the spool or log directories while checking their
8428 size for availability, Exim confusingly gave the error "space shortage".
8429 Furthermore, in debugging mode it crashed with a floating point exception.
8430 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8431 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8432 serious problem, Exim now writes to the main and panic logs when this
8433 happens, with details of the failure. It then refuses to accept the
8434 incoming message, giving the message "spool directory problem" or "log
8435 directory problem" with a 421 code for SMTP messages.
8437 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8438 0, 1, and 2 exist, because some OS complain for execs without them (see
8439 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8440 descriptors. However, the code omitted to check that the open succeeded,
8441 causing mysterious errors if for some reason the permissions on /dev/null
8442 got screwed. Now Exim writes a message to the main and panic logs, and
8443 bombs out if it can't open /dev/null.
8445 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8446 interact so that it is all more flexible. It is supposed to remain
8447 backwards compatible. Also added extra_local_interfaces.
8449 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8450 to bomb out with an assertion failure - to the client this appears as a
8451 connection drop. This problem occurs in the part of the code that was taken
8452 from the Samba project. Fortunately, the assertion is in a very simple
8453 function, so I have fixed this by reproducing the function inline in the
8454 one place where it is called, and arranging for authentication to fail
8455 instead of killing the process with assert().
8457 75. The SPA client code was not working when the server requested OEM rather
8458 than Unicode encoding.
8460 76. Added code to make require_files with a specific uid setting more usable in
8461 the case where statting the file as root fails - usually a non-root-mounted
8462 NFS file system. When this happens and the failure is EACCES, Exim now
8463 forks a subprocess and does the per-uid checking as the relevant uid.
8465 77. Added process_log_path.
8467 78. If log_file_path was not explicitly set, a setting of check_log_space or
8468 check_log_inodes was ignored.
8470 79. If a space check for the spool or log partitions fails, the incident is now
8471 logged. Of course, in the latter case the data may get lost...
8473 80. Added the %p formatting code to string_format() so that it can be used to
8474 print addresses in debug_print(). Adjusted all the address printing in the
8475 debugging in store.c to use %p rather than %d.
8477 81. There was a concern that a line of code in smtp_in.c could overflow a
8478 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8479 initially expressed, the concern was not well-founded, because trailing
8480 spaces are removed early. However, if the trailing spaces were followed by
8481 a NULL, they did not get removed, so the overflow was possible. Two fixes
8484 (a) I re-wrote the offending code in a cleaner fashion.
8485 (b) If an incoming SMTP command contains a NULL character, it is rejected
8488 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8489 runs initgroups(), so that if the Exim user is in any additional groups,
8490 they will be used during message reception.
8496 The change log for 4.20 and earlier releases has been archived.